Rx vega 64 vs radeon 7: AMD Radeon-VII vs RX Vega-64

AMD Radeon RX Vega 64 vs AMD Radeon VII: What is the difference?

53points

AMD Radeon RX Vega 64

73points

AMD Radeon VII

Comparison winner

vs

54 facts in comparison

AMD Radeon RX Vega 64

AMD Radeon VII

Why is AMD Radeon RX Vega 64 better than AMD Radeon VII?

• 256 more shading units?
  4096vs3840
• 16 more texture mapping units (TMUs)?
  256vs240

Why is AMD Radeon VII better than AMD Radeon RX Vega 64?

• 203MHz faster GPU clock speed?
  1450MHzvs1247MHz
• 1.22 TFLOPS higher floating-point performance?
  13.82 TFLOPSvs12.6 TFLOPS
• 131.46 GPixel/s higher pixel rate?
  230.4 GPixel/svs98.94 GPixel/s
• 55MHz faster memory clock speed?
  1000MHzvs945MHz
• 110MHz higher effective memory clock speed?
  2000MHzvs1890MHz
• 36. 2 GTexels/s higher texture rate?
  432 GTexels/svs395.8 GTexels/s
• 1.98x more VRAM?
  16GBvs8.1GB
• 540.2GB/s more memory bandwidth?
  1024GB/svs483.8GB/s

Which are the most popular comparisons?

AMD Radeon RX Vega 64

vs

AMD Radeon R9 290X

AMD Radeon VII

vs

AMD Radeon Vega 8

AMD Radeon RX Vega 64

vs

Nvidia GeForce RTX 2060

AMD Radeon VII

vs

Nvidia GeForce GTX 1650

AMD Radeon RX Vega 64

vs

Nvidia GeForce GTX 1080 Ti

AMD Radeon VII

vs

Nvidia GeForce GTX 1080 Ti

AMD Radeon RX Vega 64

vs

Asus Radeon RX 6900 XT

AMD Radeon VII

vs

AMD Radeon RX 6800 XT

AMD Radeon RX Vega 64

vs

AMD Radeon Vega 8

AMD Radeon VII

vs

Nvidia GeForce RTX 3050 Laptop

AMD Radeon RX Vega 64

vs

AMD Radeon Vega Frontier Edition

AMD Radeon VII

vs

Nvidia Titan Xp

AMD Radeon RX Vega 64

vs

Nvidia GeForce GTX 1650

AMD Radeon VII

vs

Sapphire Radeon RX Vega 64

AMD Radeon RX Vega 64

vs

Nvidia GeForce RTX 3060 Ti

AMD Radeon VII

vs

Gigabyte Radeon RX 6600 XT Eagle

AMD Radeon RX Vega 64

vs

AMD Radeon RX 6800 XT

AMD Radeon VII

vs

AMD Radeon RX Vega 8

Price comparison

User reviews

Performance

1. GPU clock speed

1247MHz

1450MHz

The graphics processing unit (GPU) has a higher clock speed.

2.GPU turbo

1546MHz

1800MHz

When the GPU is running below its limitations, it can boost to a higher clock speed in order to give increased performance.

3.pixel rate

98.94 GPixel/s

230.4 GPixel/s

The number of pixels that can be rendered to the screen every second.

4.floating-point performance

12.6 TFLOPS

13.82 TFLOPS

Floating-point performance is a measurement of the raw processing power of the GPU.

5.texture rate

395.8 GTexels/s

432 GTexels/s

The number of textured pixels that can be rendered to the screen every second.

6.GPU memory speed

945MHz

1000MHz

The memory clock speed is one aspect that determines the memory bandwidth.

7.shading units

Shading units (or stream processors) are small processors within the graphics card that are responsible for processing different aspects of the image.

8.texture mapping units (TMUs)

TMUs take textures and map them to the geometry of a 3D scene. More TMUs will typically mean that texture information is processed faster.

9.render output units (ROPs)

The ROPs are responsible for some of the final steps of the rendering process, writing the final pixel data to memory and carrying out other tasks such as anti-aliasing to improve the look of graphics.

Memory

1.effective memory speed

1890MHz

2000MHz

The effective memory clock speed is calculated from the size and data rate of the memory. Higher clock speeds can give increased performance in games and other apps.

2. maximum memory bandwidth

483.8GB/s

1024GB/s

This is the maximum rate that data can be read from or stored into memory.

3.VRAM

VRAM (video RAM) is the dedicated memory of a graphics card. More VRAM generally allows you to run games at higher settings, especially for things like texture resolution.

4.memory bus width

2048bit

4096bit

A wider bus width means that it can carry more data per cycle. It is an important factor of memory performance, and therefore the general performance of the graphics card.

5.version of GDDR memory

Unknown. Help us by suggesting a value. (AMD Radeon RX Vega 64)

Unknown. Help us by suggesting a value. (AMD Radeon VII)

Newer versions of GDDR memory offer improvements such as higher transfer rates that give increased performance.

6.Supports ECC memory

✖AMD Radeon RX Vega 64

✖AMD Radeon VII

Error-correcting code memory can detect and correct data corruption. It is used when is it essential to avoid corruption, such as scientific computing or when running a server.

Features

1.DirectX version

DirectX is used in games, with newer versions supporting better graphics.

2.OpenGL version

OpenGL is used in games, with newer versions supporting better graphics.

3.OpenCL version

Some apps use OpenCL to apply the power of the graphics processing unit (GPU) for non-graphical computing. Newer versions introduce more functionality and better performance.

4.Supports multi-display technology

✔AMD Radeon RX Vega 64

✔AMD Radeon VII

The graphics card supports multi-display technology. This allows you to configure multiple monitors in order to create a more immersive gaming experience, such as having a wider field of view.

5.load GPU temperature

Unknown. Help us by suggesting a value. (AMD Radeon RX Vega 64)

Unknown. Help us by suggesting a value. (AMD Radeon VII)

A lower load temperature means that the card produces less heat and its cooling system performs better.

6.supports ray tracing

✖AMD Radeon RX Vega 64

✖AMD Radeon VII

Ray tracing is an advanced light rendering technique that provides more realistic lighting, shadows, and reflections in games.

7.Supports 3D

✖AMD Radeon RX Vega 64

✖AMD Radeon VII

Allows you to view in 3D (if you have a 3D display and glasses).

8.supports DLSS

✖AMD Radeon RX Vega 64

✖AMD Radeon VII

DLSS (Deep Learning Super Sampling) is an upscaling technology powered by AI. It allows the graphics card to render games at a lower resolution and upscale them to a higher resolution with near-native visual quality and increased performance. DLSS is only available on select games.

9.PassMark (G3D) result

Unknown. Help us by suggesting a value. (AMD Radeon RX Vega 64)

This benchmark measures the graphics performance of a video card. Source: PassMark.

Ports

1.has an HDMI output

✔AMD Radeon RX Vega 64

✔AMD Radeon VII

Devices with a HDMI or mini HDMI port can transfer high definition video and audio to a display.

2.HDMI ports

More HDMI ports mean that you can simultaneously connect numerous devices, such as video game consoles and set-top boxes.

3.HDMI version

Unknown. Help us by suggesting a value. (AMD Radeon RX Vega 64)

Unknown. Help us by suggesting a value. (AMD Radeon VII)

Newer versions of HDMI support higher bandwidth, which allows for higher resolutions and frame rates.

4.DisplayPort outputs

Allows you to connect to a display using DisplayPort.

5.DVI outputs

Allows you to connect to a display using DVI.

6.mini DisplayPort outputs

Allows you to connect to a display using mini-DisplayPort.

Price comparison

Cancel

Which are the best graphics cards?

Radeon 7 vs Vega 64 and Vega 56: is AMD’s new card worth the upgrade?

When the Radeon 7 was first announced in January 2019, AMD promised the new card would be a significant step up over their previous flagship card, the Vega 64. Team Red suggested that most games would run at 25 to 40 per cent faster on the Radeon 7, even at 4K resolution and max settings. That’s a bold claim, and now that the Radeon 7 has officially been released we can put it to the test.

We have already covered how AMD’s new top card compares to the Nvidia GPUs in our Radeon 7 vs RTX 2080 and Radeon 7 vs RTX 2080 Ti articles, so now it’s time to see how the Radeon 7 stacks up against the previous AMD performance champions: the Vega 56 and Vega 64. If you’re already running one of these cards and want to see how much of a performance boost you’d get for your $700/£600 investment, this is the article for you.

---

Thinking about upgrading your gaming PC? Take a look at our recommendations for the best graphics cards on the market. Pairing your graphics card with one of the best gaming monitors is also important, so check out our latest picks!

---

Our greatest focus will be showing how the Radeon 7 compares to its Vega forebears in a gamut of recent AAA games. We’ll also look at the few feature differences between the first and second-generation Vega cards and show how much you could expect to pay for the Radeon 7. Let’s move quickly through the features, then move onto the performance benchmarks and pricing.

How much better is the Radeon 7 than the Vega 56 or 64?

Radeon 7 vs Vega 64 and Vega 56: feature comparison

Although the Radeon 7 is a new card made using an advanced 7nm manufacturing process, it doesn’t offer any substantial feature additions over the first-generation Vega 56 and Vega 64 GPUs. You’ll get the same ReLive software for streaming or recording video, Eyefinity for creating multi-monitor gaming setups and the AMD Link app for controlling your graphics card remotely. AMD’s excellent WattMan software returns as well, with the ability to overclock your card, adjust power usage and enable various features. Finally, both cards will work well with FreeSync monitors, which can smooth out uneven frame-rates without introducing the input lag that can make traditional v-sync such a pain.

While it isn’t a feature per se, the Radeon 7 does have significantly more VRAM than its predecessor — 16GB versus 8GB. The greater capacity may make a difference to gaming performance in some titles, but it could have a greater effect on creative workloads. For example, editing 4K or 8K footage in Adobe Premiere can use upwards of 10GB of video memory, so having 16GB on tap could speed up workflows significantly. That makes the Radeon 7 a better choice for content creators than the mere improvement in compute power would suggest.

Radeon 7 vs Vega 64 and Vega 56: game benchmarks

So: how much faster is the Radeon 7 than the Vega 56 and Vega 64? We are going to answer that question by looking at ten recent games at the full range of resolutions: 1920 x 1080 (1080p), 2560 x 1440 (1440p) and 3840 x 2160 (4K). No matter which monitor you have on your desk, you should get a good idea of the performance you can expect with one of these cards in your system — as long as the rest of your setup is up to par.

Note that while we are including 1080p benchmark results for completeness, we do not recommend using the Radeon 7 or similarly powerful hardware at this resolution. We ran into CPU bottlenecking in many tests, even with a powerful test rig with a water-cooled and overclocked Core i7-8700K processor, indicating that we are leaving GPU power on the table. CPU bottlenecks can also result in greater run-to-run variance in our results. In general, we recommending using a high-end Core i7 or Ryzen 7 processor if you are targeting high refresh rates at 1080p or 1440p.

To give you a better idea of these cards’ raw performance, you can also see a selection of their specifications in the table below. All three AMD cards are so-called reference models that operate at stock speeds, while the RTX 2080 FE we’re using as a point of comparison does come with a healthy factory overclock. That means entry-level RTX 2080 cards will perform slightly worse than our results indicate.

	Radeon 7	RX Vega 64	RX Vega 56	RTX 2080 FE
Manufacturing process	7nm	14nm	14nm	12nm
GPU cores	3840	4096	3584	2944
VRAM	16GB HBM2	8GB HBM2	8GB HBM2	8GB GDDR6
Memory bandwidth	1024GB/s	484GB/s	410GB/s	448GB/s
Base clock	1400MHz	1200MHz	1138MHz	1515MHz
Boost clock	1800MHz	1536MHz	1474MHz	1800MHz
TDP	300W	295W	210W	225W
RRP	$700/£650	$600/£560	$500/£460	$800/£750

Our testing results are shown using a unique Digital Foundry benchmarking system — as long as you’re viewing the desktop version of this page, anyway. A YouTube video will show you the scene that we tested each card on, with live frame-rate and frame time data embedded below. The advantage of this added complexity is that you can use the controls to the right of the video to add or remove different cards and resolutions.

Below the real-time telemetry, you can find quick summaries for the entire run, including the handy lowest one per cent and lowest five per cent figures which give you an idea of each card’s worst-case performance and overall stability. It’s worth remembering that to see these different figures, you need to mouse over the image. You can also click the chart to toggle between absolute figures and percentages.

Without further ado, let’s get into the benchmarks!

Assassin’s Creed Odyssey

We begin with Assassin’s Creed Odyssey, which was released in late 2018. The Radeon 7 is 22 per cent faster than the Vega 56 and 13 per cent faster than the Vega 64 in this first test at 1080p with the latest AMD drivers installed — beforehand, the gap was closer to 40 per cent. It is often interesting to see how these percentages change as resolution is adjusted, and we do see those numbers shift at 4K: the Radeon 7 leads the Vega 56 by 38 per cent and the Vega 64 by 28 per cent. Meanwhile, the RTX 2080 remains the overall winner in this game at all resolutions.

AC Odyssey: Ultra High, TAA

Please enable JavaScript to use our comparison tools.

Assassin’s Creed Unity

Assassin’s Creed Unity includes a tough depth of field effect that has a tremendous impact on Vega performance, and sadly that bugbear hasn’t been shaken off with the Radeon 7. All three cards exhibit the issue, with frame-rates dropping to the low thirties on the older Vega cards and the high thirties on the Radeon 7 at 1080p. Regardless, the new AMD card does manage to carve out a 27 percent lead over the Vega 64 at the lower resolutions and 33 per cent at 4K.

AC Unity: Ultra High, FXAA

Please enable JavaScript to use our comparison tools.

Battlefield 1

Battlefield 1 is the best showing for the Radeon 7, and the first time we see the Radeon 7 ahead of our factory-overclocked RTX 2080 at 177fps. The older Vega cards perform well in our roll through no man’s land too, averaging 133fps and 144fps at 1080p. The advantage that the Radeon 7 has over the Vega 64 is 23 per cent, moving to around 35 per cent at 1440p and 4K. Still, all but the Vega 56 manage to deliver at least 60fps average at 4K.

Battlefield 1: Ultra, TAA

Please enable JavaScript to use our comparison tools.

Crysis 3

More than half a decade after its release, Crysis 3 remains a famously difficult game to run at very high settings — particularly at 4K. The two Vega cards only manage just over 30fps at this resolution, delivering a console-quality feel to the proceedings, but the Radeon 7 does better to manage 49fps on average. That’s again a 35 per cent advantage over the Vega 64 and 49 per cent over the Vega 56 — not a bad generational leap.

Crysis 3: very high, SMAA T2X

Please enable JavaScript to use our comparison tools.

Far Cry 5

2018 release Far Cry 5 has one of the most chilled out integrated benchmarks we’ve ever seen, with calming music and a pretty Montana lake scene to look at. It’s a little more stressful if you’re a graphics card, with none of the cards in our comparison managing to average over 60fps at 4K. The Radeon 7 and RTX 2080 come closest, rounding up to 60fps on the dot, while the older Vega cards deliver 41fps and 46fps at this resolution.

Far Cry 5: Ultra, TAA

Please enable JavaScript to use our comparison tools.

Far Cry Primal

Far Cry Primal is a rework of the Far Cry 4 map into something a little more rough and ready, a prehistoric take on the familiar open world formula. The game was released in 2016 but remains a challenging task for even the newly released Radeon 7. AMD’s latest GPU does manage a comfortable 134fps at 1080p and 101fps at 1440p, but averages just under 60fps at 4K. That’s 46 per cent faster than the Vega 56 and 35 per cent faster than the Vega 56, if you were wondering.

Far Cry Primal: Ultra, SMAA

Please enable JavaScript to use our comparison tools.

Ghost Recon Wildlands

Ghost Recon Wildlands was released in 2017, but its punishing ultra preset means that it will be menacing graphics cards for years to come. None of the four graphics cards we’re comparing today manage to even get close to 60fps at 4K, with the Radeon 7 and RTX 2080 sitting at 38fps and the older Vega cards at sub-30fps. Once again, the advantage the Radeon 7 has over the Vega 64 at 4K is around 29 per cent.

Ghost Recon Wildlands: Ultra, TAA

Please enable JavaScript to use our comparison tools.

Rise of the Tomb Raider

Rise of the Tomb Raider’s integrated benchmark is a nice way to test graphics hardware, even though it is a little easier to run than the full game. The Radeon 7 isclose behind the RTX 2080 here, with a 24 per cent advantage over the Vega 64 and a 31 per cent lead over the Vega 56. The numbers do oscillate up and down as we move from 1080p to higher resolutions, but the trend remains the same.

Rise of the Tomb Raider: Very High, SMAA

Please enable JavaScript to use our comparison tools.

Shadow of the Tomb Raider

Our penultimate game to test is Shadow of the Tomb Raider, which comes with a longer and more realistic benchmark. The Vega 56 and Vega 64 show their age here, managing not more than 30fps at 4K, while the Radeon 7 ties the RTX 2080 at 46fps at the same resolution. The gap from the Vega 64 to the Radeon 7 is about 32 per cent.

Shadow of the Tomb Raider DX12: Highest, TAA

Please enable JavaScript to use our comparison tools.

The Witcher 3

Our last unique game is one of the best to be released in 2015, The Witcher 3. Our horseback ride through Novigrad reveals a 28 per cent lead for the Radeon 7 over the Vega 64 at 1080p, growing to 33 per cent at 1440p and 36 per cent at 4K. The Radeon 7 and RTX 2080 are very closely matched at 4K, with both cards getting within a few frames of 60fps, while the Vega cards are in the low 40s.

The Witcher 3: Ultra, POST-AA, No Hairworks

Please enable JavaScript to use our comparison tools.

Cross-generational comparison

We’ll conclude with a look at how the Radeon 7 compares to past AMD graphics cards, all the way up to the R9 Fury X. You can see that the Radeon 7 pulls out one of the biggest inter-generational leads we’ve ever seen, with a 27 per cent gap over the outgoing Vega 64 at 1080p and 28 per cent at 1440p. Overall, it’s a strong statement from AMD — and with Navi rumoured to be on the horizon for a potential launch this year, we may yet see even greater leaps from the American company.

Assassin’s Creed Unity: Ultra High, FXAA

Please enable JavaScript to use our comparison tools.

Radeon 7 vs Vega 64 and Vega 56: price and availability

The Radeon 7 is offered with three free games when purchased from select vendors: The Division 2, Resident Evil 2 and Devil May Cry 5. You’ll get precisely the same games with the Vega 56 and Vega 64 when buying from participating retailers, so there’s no difference here. However, the Radeon 7 is only being available in reference designs at present, while there are a few custom Vega 56 and Vega 64 cards. If having a particularly quiet setup is important to you, one of these custom designs could be worth waiting for.

So how much can you expect to pay for each card and how easy is it to find them? The Radeon 7 is the most expensive of the three, with an RRP of $700/£600, but worries surrounding availability haven’t materialised since launch. The Vega 64 is the middle child, costing around $400/£310, and is relatively easy to find in most territories. Finally, the Vega 56 is the cheapest Vega card thus far, with entry-level models going for around $300/£250. Unfortunately, the Vega 56 is also the hardest to find at present, although the used market can be worth checking. If you snag a bargain, a Vega card can still be worthwhile — but the Radeon 7 offers significantly better performance. The RTX 2080 and other Nvidia cards are also worth considering if you’re open to joining Team Green.

That concludes our comparison between the new Radeon 7 and AMD’s last-generation Vega 56 and Vega 64 products. Want to learn more about the Radeon 7? We encourage you to check out our written Radeon 7 review or our Radeon 7 review video for a more detailed investigation into AMD’s new top card.

Which GPUs are worth buying? We’ve made our picks for the best graphics cards available, updated with the latest graphics cards as they’re released. As well as an overall performance champ, we name the best value graphics card and best cheap graphics card to guide your next upgrade.

Radeon Pro Vega II vs Radeon RX Vega 64 Graphics cards Comparison

When comparing Radeon Pro Vega II and Radeon RX Vega 64, we look primarily at benchmarks and game tests. But it is not only about the numbers. Often you can find third-party models with higher clock speeds, better cooling, or a customizable RGB lighting. Not all of them will have all the features you need. Another thing to consider is the port selection. Most graphics cards have at least one DisplayPort and HDMI interface, but some monitors require DVI. Before you buy, check the TDP of the graphics card — this characteristic will help you estimate the consumption of the graphics card. You may even have to upgrade your PSU to meet its requirements. An important factor when choosing between Radeon Pro Vega II and Radeon RX Vega 64 is the price. Does the additional cost justify the performance hit? Our comparison should help you make the right decision.

Radeon Pro Vega II

Radeon RX Vega 64

Check Price

Main Specs

	Radeon Pro Vega II	Radeon RX Vega 64
Power consumption (TDP)	475 Watt	295 Watt
Interface	PCIe 3.0 x16	PCIe 3.0 x16
Supplementary power connectors	None	2x 8-pin
Memory type	HBM2	HBM2
Maximum RAM amount	32 GB	8 GB
Display Connectors	1x HDMI, 4x USB Type-C	1x HDMI, 3x DisplayPort
		Check Price

• Radeon Pro Vega II has 61% more power consumption, than Radeon RX Vega 64.
• Both video cards are using PCIe 3.0 x16 interface connection to a motherboard.
• Radeon Pro Vega II has 24 GB more memory, than Radeon RX Vega 64.
• Both cards are used in Desktops.
• Radeon Pro Vega II is build with GCN 5.1 architecture, and Radeon RX Vega 64 — with Vega.
• Core clock speed of Radeon RX Vega 64 is 56 MHz higher, than Radeon Pro Vega II.
• Radeon Pro Vega II is manufactured by 7 nm process technology, and Radeon RX Vega 64 — by 14 nm process technology.
• Memory clock speed of Radeon Pro Vega II is 667 MHz higher, than Radeon RX Vega 64.

Game benchmarks

Assassin’s Creed OdysseyBattlefield 5Call of Duty: WarzoneCounter-Strike: Global OffensiveCyberpunk 2077Dota 2Far Cry 5FortniteForza Horizon 4Grand Theft Auto VMetro ExodusMinecraftPLAYERUNKNOWN’S BATTLEGROUNDSRed Dead Redemption 2The Witcher 3: Wild HuntWorld of Tanks
high / 1080p	65−70	65−70
ultra / 1080p	50−55	50−55
QHD / 1440p	40−45	40−45
4K / 2160p	24−27	24−27
low / 720p	90−95	90−95
medium / 1080p	80−85	80−85
Radeon Pro Vega II and Radeon RX Vega 64 have the same average FPS in Assassin’s Creed Odyssey.
high / 1080p	100−110	100−110
ultra / 1080p	100−105	95−100
QHD / 1440p	75−80	75−80
4K / 2160p	40−45	40−45
low / 720p	150−160	150−160
medium / 1080p	120−130	120−130
The average gaming FPS of Radeon Pro Vega II in Battlefield 5 is 1% more, than Radeon RX Vega 64.
low / 768p	45−50	45−50
Radeon Pro Vega II and Radeon RX Vega 64 have the same average FPS in Call of Duty: Warzone.
low / 768p	250−260	250−260
medium / 768p	230−240	230−240
QHD / 1440p	150−160	150−160
4K / 2160p	100−110	100−110
high / 768p	220−230	220−230
Radeon Pro Vega II and Radeon RX Vega 64 have the same average FPS in Counter-Strike: Global Offensive.
low / 768p	70−75	70−75
medium / 1080p	50−55	50−55
Radeon Pro Vega II and Radeon RX Vega 64 have the same average FPS in Cyberpunk 2077.
low / 768p	120−130	120−130
medium / 768p	110−120	110−120
Radeon Pro Vega II and Radeon RX Vega 64 have the same average FPS in Dota 2.
high / 1080p	90−95	90−95
ultra / 1080p	85−90	85−90
QHD / 1440p	65−70	65−70
4K / 2160p	35−40	35−40
low / 720p	120−130	120−130
medium / 1080p	95−100	95−100
Radeon Pro Vega II and Radeon RX Vega 64 have the same average FPS in Far Cry 5.
high / 1080p	140−150	130−140
ultra / 1080p	110−120	100−110
QHD / 1440p	75−80	70−75
4K / 2160p	21−24	21−24
low / 720p	260−270	250−260
medium / 1080p	170−180	170−180
The average gaming FPS of Radeon Pro Vega II in Fortnite is 4% more, than Radeon RX Vega 64.
high / 1080p	120−130	110−120
ultra / 1080p	95−100	95−100
QHD / 1440p	75−80	75−80
4K / 2160p	50−55	50−55
low / 720p	160−170	160−170
medium / 1080p	120−130	120−130
The average gaming FPS of Radeon Pro Vega II in Forza Horizon 4 is 1% more, than Radeon RX Vega 64.
low / 768p	170−180	170−180
medium / 768p	160−170	160−170
high / 1080p	120−130	120−130
ultra / 1080p	70−75	70−75
QHD / 1440p	65−70	65−70
4K / 2160p	65−70	−
Radeon Pro Vega II and Radeon RX Vega 64 have the same average FPS in Grand Theft Auto V.
high / 1080p	60−65	55−60
ultra / 1080p	45−50	45−50
QHD / 1440p	35−40	35−40
4K / 2160p	24−27	21−24
low / 720p	120−130	120−130
medium / 1080p	80−85	80−85
The average gaming FPS of Radeon Pro Vega II in Metro Exodus is 1% more, than Radeon RX Vega 64.
low / 768p	120−130	120−130
Radeon Pro Vega II and Radeon RX Vega 64 have the same average FPS in Minecraft.
high / 1080p	120−130	120−130
ultra / 1080p	95−100	90−95
4K / 2160p	18−20	18−20
low / 720p	150−160	150−160
medium / 1080p	140−150	130−140
The average gaming FPS of Radeon Pro Vega II in PLAYERUNKNOWN’S BATTLEGROUNDS is 2% more, than Radeon RX Vega 64.
high / 1080p	55−60	55−60
ultra / 1080p	35−40	35−40
QHD / 1440p	30−33	27−30
4K / 2160p	18−20	18−20
low / 720p	120−130	120−130
medium / 1080p	85−90	85−90
Radeon Pro Vega II and Radeon RX Vega 64 have the same average FPS in Red Dead Redemption 2.
low / 768p	260−270	250−260
medium / 768p	180−190	180−190
high / 1080p	110−120	110−120
ultra / 1080p	60−65	60−65
4K / 2160p	40−45	40−45
The average gaming FPS of Radeon Pro Vega II in The Witcher 3: Wild Hunt is 1% more, than Radeon RX Vega 64.
low / 768p	220−230	210−220
ultra / 1080p	100−110	100−110
The average gaming FPS of Radeon Pro Vega II in World of Tanks is 3% more, than Radeon RX Vega 64.

Full Specs

	Radeon Pro Vega II	Radeon RX Vega 64
Architecture	GCN 5.1	Vega
Code name	Vega 20	Vega
Type	Workstation	Desktop
Release date	3 June 2019	14 August 2017
Pipelines	4096	4096
Core clock speed	1574 MHz	1630 MHz
Boost Clock	1720 MHz	1546 MHz
Transistor count	13,230 million	12,500 million
Manufacturing process technology	7 nm	14 nm
Texture fill rate	440. 3	395.8
Floating-point performance		13,353 gflops
Length		279 mm
Memory bus width	4096 Bit	2048 Bit
Memory clock speed	1612 MHz	945 MHz
Memory bandwidth	825.3 GB/s	483.8 GB/s
Shared memory		—
DirectX	12 (12_1)	12 (12_1)
Shader Model	6.4	6.4
OpenGL	4.6	4.6
OpenCL	2.0	2.0
Vulkan	1.2.131	1.1.125
Monero / XMR (CryptoNight)		1. 9 kh/s
Ethereum / ETH (DaggerHashimoto)		31. 5 Mh/s
		Check Price

Similar compares

• Radeon Pro Vega II vs GeForce RTX 2060
• Radeon Pro Vega II vs GeForce RTX 2060
• Radeon RX Vega 64 vs GeForce RTX 2060
• Radeon RX Vega 64 vs GeForce RTX 2060
• Radeon Pro Vega II vs Quadro RTX 5000 Max Q
• Radeon Pro Vega II vs Radeon RX 5700
• Radeon RX Vega 64 vs Quadro RTX 5000 Max Q
• Radeon RX Vega 64 vs Radeon RX 5700

Radeon VII vs Radeon RX Vega 64

1. Home
2. VGA Benchmarks
3. Radeon VII vs Radeon RX Vega 64

• Radeon VII

  123%

• Radeon RX Vega 64

  100%

Relative performance

Reasons to consider Radeon VII

23% higher gaming performance.

This is a much newer product, it might have better long term support.

Supports Direct3D 12 Async Compute

Supports FreeSync

Supports ReLive (allows game streaming/recording with minimum performance penalty)

Supports TrueAudio

Based on an outdated architecture (AMD GCN), there may be no performance optimizations for current games and applications

Reasons to consider Radeon RX Vega 64

Supports Direct3D 12 Async Compute

Supports FreeSync

Supports ReLive (allows game streaming/recording with minimum performance penalty)

Supports TrueAudio

Based on an outdated architecture (AMD GCN), there may be no performance optimizations for current games and applications

---

HWBench recommends Radeon VII

The Radeon VII is the better performing card based on the game benchmark suite used (72 combinations of games and resolutions).

Core Configuration

	Radeon VII		Radeon RX Vega 64
GPU Name	Vega 20 (Vega 20 XT)	vs	Vega 10 (Vega 10 XT)
Fab Process	7 nm	vs	14 nm
Die Size	331 mm²	vs	484 mm²
Transistors	13,230 million	vs	12,500 million
Shaders	3840	vs	4096
Compute Units	60	vs	64
Core clock	1450 MHz	vs	1250 MHz
ROPs	64	vs	64
TMUs	240	vs	256

Memory Configuration

	Radeon VII		Radeon RX Vega 64
Memory Type	HBM2	vs	HBM2
Bus Width	4096 bit	vs	2048 bit
Memory Speed	1000 MHz 2000 MHz effective	vs	945 MHz 1890 MHz effective
Memory Size	16384 Mb	vs	8192 Mb

Additional details

	Radeon VII		Radeon RX Vega 64
TDP	295 watts	vs	295 watts
Release Date	7 Feb 2019	vs	8 Aug 2017

• Radeon VII

  115. 20 GP/s

• Radeon RX Vega 64

  96.00 GP/s

GigaPixels — higher is better

• Radeon VII

  432.00 GT/s

• Radeon RX Vega 64

  384.00 GT/s

GigaTexels — higher is better

• Radeon VII

  1024.00 GB/s

• Radeon RX Vega 64

  483.80 GB/s

GB/s — higher is better

• Radeon VII

  13824.00 GFLOPs

• Radeon RX Vega 64

  12288.00 GFLOPs

GFLOPs — higher is better

Ultra Quality, DirectX12, Windows 10×64

• Radeon VII

  70

• Radeon RX Vega 64

  60

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10×64

• Radeon VII

  146

• Radeon RX Vega 64

  115

FPS (higher is better)

Ultra Quality, DirectX11, Windows 10×64

• Radeon VII

  84

• Radeon RX Vega 64

  66

FPS (higher is better)

Ultra quality TTA DirectX11 Windows10 x64

• Radeon VII

  135

• Radeon RX Vega 64

  108

FPS (higher is better)

Ultra High Quality, TAA, DirectX12, Windows 10 x64

• Radeon VII

  145

• Radeon RX Vega 64

  101

FPS (higher is better)

Windows 10 x64, Ultra quality, DirectX12

• Radeon VII

  56

• Radeon RX Vega 64

  48

FPS (higher is better)

Very High Quality TAA, DirectX12, Windows 10 x64

• Radeon VII

  144

• Radeon RX Vega 64

  121

FPS (higher is better)

Highest quality DirectX11 Windows10 x64

• Radeon VII

  134

• Radeon RX Vega 64

  115

FPS (higher is better)

Ultra quality DirectX12 Windows10 x64

• Radeon VII

  104

• Radeon RX Vega 64

  100

FPS (higher is better)

Ultra quality DirectX11 Windows10 x64

• Radeon VII

  109

• Radeon RX Vega 64

  106

FPS (higher is better)

Ultra Quality, DirectX11, Windows 10×64

• Radeon VII

  119

• Radeon RX Vega 64

  114

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10 x64

• Radeon VII

  111

• Radeon RX Vega 64

  88

FPS (higher is better)

DX11, Max Details, 16:1 AF, 2xMSAA

• Radeon VII

  161

• Radeon RX Vega 64

  126

FPS (higher is better)

Very high quality DirectX11 Windows10 x64

• Radeon VII

  100

• Radeon RX Vega 64

  78

FPS (higher is better)

Ultra Detail,16:1 AF, DirectX12, Windows 10×64

• Radeon VII

  71

• Radeon RX Vega 64

  61

FPS (higher is better)

Ultra Quality, DirectX11, Windows 10×64

• Radeon VII

  44

• Radeon RX Vega 64

  36

FPS (higher is better)

Ultra Quality, HR Textures, DirectX11, Windows 10 x64

• Radeon VII

  115

• Radeon RX Vega 64

  100

FPS (higher is better)

UltraTX Max Quality, Vulkan, Windows 10 x64

• Radeon VII

  71

• Radeon RX Vega 64

  60

FPS (higher is better)

Max Quality, DirectX12, Windows 10×64

• Radeon VII

  149

• Radeon RX Vega 64

  134

FPS (higher is better)

Highest Details, Pure hair, HBAO+, DirectX12, Windows 10 x64

• Radeon VII

  117

• Radeon RX Vega 64

  90

FPS (higher is better)

Ultra quality DirectX12 Windows10 x64

• Radeon VII

  168

• Radeon RX Vega 64

  132

FPS (higher is better)

Ultra Quality, DirectX12, Async Compute ,Windows 10×64

• Radeon VII

  169

• Radeon RX Vega 64

  137

FPS (higher is better)

DX11,Max Details, 16:1 HQ-AF, +AA

• Radeon VII

  129

• Radeon RX Vega 64

  103

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10×64

• Radeon VII

  67

• Radeon RX Vega 64

  58

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10×64

• Radeon VII

  56

• Radeon RX Vega 64

  48

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10×64

• Radeon VII

  117

• Radeon RX Vega 64

  89

FPS (higher is better)

Ultra Quality, DirectX11, Windows 10×64

• Radeon VII

  60

• Radeon RX Vega 64

  49

FPS (higher is better)

Ultra quality TTA DirectX11 Windows10 x64

• Radeon VII

  107

• Radeon RX Vega 64

  84

FPS (higher is better)

Ultra High Quality, TAA, DirectX12, Windows 10 x64

• Radeon VII

  112

• Radeon RX Vega 64

  81

FPS (higher is better)

Windows 10 x64, Ultra quality, DirectX12

• Radeon VII

  38

• Radeon RX Vega 64

  32

FPS (higher is better)

Very High Quality TAA, DirectX12, Windows 10 x64

• Radeon VII

  104

• Radeon RX Vega 64

  86

FPS (higher is better)

Highest quality DirectX11 Windows10 x64

• Radeon VII

  102

• Radeon RX Vega 64

  84

FPS (higher is better)

Ultra quality DirectX12 Windows10 x64

• Radeon VII

  97

• Radeon RX Vega 64

  74

FPS (higher is better)

Ultra quality DirectX11 Windows10 x64

• Radeon VII

  99

• Radeon RX Vega 64

  87

FPS (higher is better)

Ultra Quality, DirectX11, Windows 10×64

• Radeon VII

  108

• Radeon RX Vega 64

  89

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10 x64

• Radeon VII

  79

• Radeon RX Vega 64

  62

FPS (higher is better)

DX11, Max Details, 16:1 AF, 2xMSAA

• Radeon VII

  116

• Radeon RX Vega 64

  91

FPS (higher is better)

Very high quality DirectX11 Windows10 x64

• Radeon VII

  74

• Radeon RX Vega 64

  60

FPS (higher is better)

Ultra Detail,16:1 AF, DirectX12, Windows 10×64

• Radeon VII

  58

• Radeon RX Vega 64

  49

FPS (higher is better)

Ultra Quality, DirectX11, Windows 10×64

• Radeon VII

  37

• Radeon RX Vega 64

  28

FPS (higher is better)

Ultra Quality, HR Textures, DirectX11, Windows 10 x64

• Radeon VII

  85

• Radeon RX Vega 64

  72

FPS (higher is better)

UltraTX Max Quality, Vulkan, Windows 10 x64

• Radeon VII

  58

• Radeon RX Vega 64

  48

FPS (higher is better)

Max Quality, DirectX12, Windows 10×64

• Radeon VII

  97

• Radeon RX Vega 64

  92

FPS (higher is better)

Highest Details, Pure hair, HBAO+, DirectX12, Windows 10 x64

• Radeon VII

  82

• Radeon RX Vega 64

  61

FPS (higher is better)

Ultra quality DirectX12 Windows10 x64

• Radeon VII

  125

• Radeon RX Vega 64

  102

FPS (higher is better)

Ultra Quality, DirectX12, Async Compute ,Windows 10×64

• Radeon VII

  128

• Radeon RX Vega 64

  100

FPS (higher is better)

DX11,Max Details, 16:1 HQ-AF, +AA

• Radeon VII

  93

• Radeon RX Vega 64

  79

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10×64

• Radeon VII

  50

• Radeon RX Vega 64

  43

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10×64

• Radeon VII

  35

• Radeon RX Vega 64

  29

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10×64

• Radeon VII

  68

• Radeon RX Vega 64

  52

FPS (higher is better)

Ultra Quality, DirectX11, Windows 10×64

• Radeon VII

  35

• Radeon RX Vega 64

  28

FPS (higher is better)

Ultra quality TTA DirectX11 Windows10 x64

• Radeon VII

  63

• Radeon RX Vega 64

  50

FPS (higher is better)

Ultra High Quality, TAA, DirectX12, Windows 10 x64

• Radeon VII

  66

• Radeon RX Vega 64

  51

FPS (higher is better)

Windows 10 x64, Ultra quality, DirectX12

• Radeon VII

  19

• Radeon RX Vega 64

  16

FPS (higher is better)

Very High Quality TAA, DirectX12, Windows 10 x64

• Radeon VII

  59

• Radeon RX Vega 64

  47

FPS (higher is better)

Highest quality DirectX11 Windows10 x64

• Radeon VII

  61

• Radeon RX Vega 64

  50

FPS (higher is better)

Ultra quality DirectX12 Windows10 x64

• Radeon VII

  52

• Radeon RX Vega 64

  38

FPS (higher is better)

Ultra quality DirectX11 Windows10 x64

• Radeon VII

  58

• Radeon RX Vega 64

  46

FPS (higher is better)

Ultra Quality, DirectX11, Windows 10×64

• Radeon VII

  63

• Radeon RX Vega 64

  49

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10 x64

• Radeon VII

  40

• Radeon RX Vega 64

  34

FPS (higher is better)

DX11, Max Details, 16:1 AF, 2xMSAA

• Radeon VII

  66

• Radeon RX Vega 64

  49

FPS (higher is better)

Very high quality DirectX11 Windows10 x64

• Radeon VII

  37

• Radeon RX Vega 64

  30

FPS (higher is better)

Ultra Detail,16:1 AF, DirectX12, Windows 10×64

• Radeon VII

  38

• Radeon RX Vega 64

  32

FPS (higher is better)

Ultra Quality, DirectX11, Windows 10×64

• Radeon VII

  24

• Radeon RX Vega 64

  16

FPS (higher is better)

Ultra Quality, HR Textures, DirectX11, Windows 10 x64

• Radeon VII

  52

• Radeon RX Vega 64

  43

FPS (higher is better)

UltraTX Max Quality, Vulkan, Windows 10 x64

• Radeon VII

  38

• Radeon RX Vega 64

  31

FPS (higher is better)

Max Quality, DirectX12, Windows 10×64

• Radeon VII

  51

• Radeon RX Vega 64

  46

FPS (higher is better)

Highest Details, Pure hair, HBAO+, DirectX12, Windows 10 x64

• Radeon VII

  44

• Radeon RX Vega 64

  31

FPS (higher is better)

Ultra quality DirectX12 Windows10 x64

• Radeon VII

  69

• Radeon RX Vega 64

  55

FPS (higher is better)

Ultra Quality, DirectX12, Async Compute ,Windows 10×64

• Radeon VII

  83

• Radeon RX Vega 64

  61

FPS (higher is better)

DX11,Max Details, 16:1 HQ-AF, +AA

• Radeon VII

  53

• Radeon RX Vega 64

  44

FPS (higher is better)

Ultra Quality, DirectX12, Windows 10×64

• Radeon VII

  30

• Radeon RX Vega 64

  25

FPS (higher is better)

	VS
Radeon VII		GeForce RTX 2070 Super

	VS
Radeon VII		Radeon RX 5700 XT

	VS
Radeon RX Vega 64		Radeon RX 6650 XT

	VS
Radeon RX Vega 64		GeForce RTX 2060 Super

	VS
GeForce RTX 3060 Ti		GeForce RTX 2080 Super

	VS
GeForce RTX 2080 Super		TITAN V

Please enable JavaScript to view the comments powered by Disqus.

Radeon VII vs RX Vega 64 ⚙️

Argon2d-ninja238KH/s259W0.919KH/W

Astralhash37.186MH/s257W0.145MH/W

Autolykos2229.1MH/s142W1.613MH/W

BCD26.286MH/s240W0.11MH/W

BMW5121507.901MH/s225W6.702MH/W

BeamHashII29.62H/s157W0.189H/W

BeamHashIII46.79H/s243W0.193H/W

Blake (2b)3.024GH/s233W0.013GH/W

Blake (2b-BTCC)3028.794MH/s—W—MH/W

Blake (2s)6.879GH/s245W0. 028GH/W

Blake (2s-Kadena)0.902GH/s128W0.007GH/W

C1131.205MH/s188W0.166MH/W

CNReverseWaltz2044.128H/s177W11.549H/W

Chukwa162.512KH/s246W0.661KH/W

Chukwa264.422KH/s250W0.258KH/W

CryptoNightAlloy1082.72H/s147W7.365H/W

CryptoNightArto2123.47H/s152W13.97H/W

CryptoNightB2N3802H/s—W—H/W

CryptoNightConceal3791.8H/s144W26.332H/W

CryptoNightFast3999. 725H/s168W23.808H/W

CryptoNightFastV24897.5H/s190W25.776H/W

CryptoNightGPU1387.62H/s192W7.227H/W

CryptoNightHaven2243.6H/s143W15.69H/W

CryptoNightHeavy2250.603H/s171W13.161H/W

CryptoNightHeavyX1021.285H/s223W4.58H/W

CryptoNightItalo2572H/s—W—H/W

CryptoNightLiteV74.343KH/s154W0.028KH/W

CryptoNightR1712.43H/s160W10.703H/W

CryptoNightSaber2224.988H/s189W11. 772H/W

CryptoNightStelliteV42152.9H/s137W15.715H/W

CryptoNightStelliteV52938.98H/s162W18.142H/W

CryptoNightTalleo0.015MH/s158W0MH/W

CryptoNightTurtle22.245KH/s228W0.098KH/W

CryptoNightUPX215.41KH/s—W—KH/W

CryptoNightV72.188KH/s151W0.014KH/W

CryptoNightV81923.55H/s205W9.383H/W

CryptoNightWOW1493.58H/s190W7.861H/W

CryptoNightZLS1911.45H/s223W8.572H/W

Cuckaroo29S4. 65H/s125W0.037H/W

Cuckaroo29b6.62H/s163W0.041H/W

CuckooBFC116.31H/s143W0.813H/W

CuckooCycle4.97H/s200W0.025H/W

Darkcoin2.34GH/s162W0.014GH/W

Dedal23.812MH/s228W0.104MH/W

Eaglesong0.867GH/s253W0.003GH/W

Equihash0.534KH/s199W0.003KH/W

Equihash(125,4)31.8H/s140W0.227H/W

Equihash(144,5)47.2H/s192W0.246H/W

Equihash(150,5)23. 177H/s150W0.155H/W

Equihash(192,7)29.305H/s176W0.167H/W

Equihash(210,9)96.2H/s78W1.233H/W

Equihash(96,5)27.519KH/s96W0.287KH/W

Etchash104.6MH/s195W0.536MH/W

Ethash104.6MH/s195W0.536MH/W

FiroPoW37.31MH/s276W0.135MH/W

Globalhash84.565MH/s138W0.613MH/W

HMQ172513.54MH/s220W0.062MH/W

Handshake0.353GH/s253W0.001GH/W

HeavyHash389. 789MH/s249W1.565MH/W

Hex21.38MH/s183W0.117MH/W

HoneyComb57.649MH/s135W0.427MH/W

Jeonghash18.006MH/s149W0.121MH/W

KAWPOW42.73MH/s220W0.194MH/W

KangarooTwelve2.235GH/s250W0.009GH/W

Lyra2REv289.098MH/s250W0.356MH/W

Lyra2REv3103.746MH/s249W0.417MH/W

Lyra2vc0ban87.182MH/s252W0.346MH/W

Lyra2z6.573MH/s218W0.03MH/W

MTP4. 01MH/s166W0.024MH/W

NeoScrypt2.573MH/s238W0.011MH/W

PHI161242.471MH/s225W0.189MH/W

PHI210.13MH/s56W0.181MH/W

Padihash19.401MH/s227W0.085MH/W

Pawelhash16.899MH/s231W0.073MH/W

Phi546599750.664H/s246W189429.881H/W

ProgPow24.262MH/s263W0.092MH/W

ProgPowSERO31.084MH/s261W0.119MH/W

ProgPowZ33.618MH/s247W0.136MH/W

RandomSFX1581. 755H/s163W9.704H/W

RandomX1592.305H/s161W9.89H/W

SHA-256csm1836.461MH/s239W7.684MH/W

Skein2687.688MH/s182W3.779MH/W

Skunkhash61.487MH/s193W0.319MH/W

SonoA3.507MH/s123W0.029MH/W

Tellor0.625GH/s—W—GH/W

TimeTravel1052.919MH/s263W0.201MH/W

Tribus120.957MH/s252W0.48MH/W

Ubqhash86.288MH/s196W0.44MH/W

Verthash0. 327MH/s196W0.002MH/W

VerusHash10.468MH/s144W0.073MH/W

X11k4.74MH/s212W0.022MH/W

X16R22.592MH/s160W0.141MH/W

X16RT23.542MH/s258W0.091MH/W

X16RTVEIL23.523MH/s207W0.114MH/W

X16Rv219.977MH/s232W0.086MH/W

X16S23.294MH/s241W0.097MH/W

X1723.051MH/s222W0.104MH/W

X17R20.908MH/s186W0.112MH/W

X1818. 002MH/s221W0.081MH/W

X21S16.552MH/s200W0.083MH/W

X22i13.298MH/s257W0.052MH/W

X25X2.348MH/s182W0.013MH/W

Xevan7.474MH/s227W0.033MH/W

Zhash48.67H/s145W0.336H/W

cuckARoo294.9H/s162W0.03H/W

cuckARood295.18H/s138W0.038H/W

cuckARoom297.55H/s220W0.034H/W

cuckAToo311.217H/s220W0.006H/W

cuckAToo320. 68H/s195W0.003H/W

vProgPow15.808MH/s250W0.063MH/W

AMD Radeon Vega 64 Revisited One Last Time: Did It Age Better than GTX 10 Series?

Today we’re dusting off our old Radeon Vega 64 graphics card and comparing it with our recently dusted off GeForce GTX 1080, because why not, it could be interesting, so here’s what’s probably going to be our last big revisit to the Radeon Vega GPU on a 51 game benchmark that includes many modern titles.

We weren’t huge fans of the Radeon Vega series, especially Vega 64. It wasn’t a terrible product, but in our opinion it was also far from a good one. In a nutshell, Vega 64 arrived a year late to compete with the well established GeForce GTX 1080. It tried to do so at the same price, while generally offering less performance. It also chugged power — more than an overclocked GTX 1080 Ti.

The AMD reference design was also pretty bad. It was hot, loud, and often throttled. There were very few partner models to speak of, and almost none of them were any good.

Meanwhile there were dozens of quality GTX 1080’s to pick from and many of them had been readily available for quite some time. As a result, we often recommended the GTX 1080 over Vega 64 and we think time has proven that was the right choice.

We should mention however that there were times when Vega 64 was selling at heavily discounted prices, so if you got one cheap, then it probably worked out nicely for you.

All that aside, AMD Vega 64 is now 5 years old and we haven’t revisited it since 2018, so how does it stack up head-to-head with the GTX 1080 in 2022?

We recently saw that the old Pascal GeForce GPU tends to struggle a bit in modern titles, either due to its aging architecture that lacks hardware support for technologies such as async compute, or a lack of driver optimization, which no doubt plays a key role.

Vega 64 was a more forward-thinking architecture that was more compute heavy, so it should be interesting to see how this match up has turned out half a decade later.

We should note that our 27 game comparison back in 2018 saw the GTX 1080 win this rivalry by a 5% margin at 1440p, so overall performance was similar, but as we’ve explained, there were other factors at play (or rather, shortcomings) that ultimately saw us go with Nvidia for this battle.

For testing we’re using our Ryzen 7 5800X3D test system using 32GB of DDR4-3200 CL14 memory along with the latest available display drivers. Resizable BAR has been enabled for all of the 51 games tested at 1080p and 1440p, and we’ll take a close look at the data for about a dozen of the titles tested before jumping into the usual breakdown graphs.

Gaming Benchmarks

First up we have Fortnite and this is an old but also new title as Epic continually updates the game and has recently moved it to Unreal Engine 5. Unfortunately though, the DirectX 12 version of the game is broken, as it suffers from constant stuttering using either AMD or Nvidia GPUs, forcing us to test using DX11.

This is a problem for AMD as their DX11 support in Fortnite has always been a little sketchy, so no surprises here as the the GTX 1080 comfortably beats Vega 64 at 1080p.

The 1440p results are a bit more interesting. If we were to just look at the average frame rate we’d conclude that Vega 64 does better at the higher resolution, but that would be ignoring that 1% lows are still much lower, which does lead to noticeable stuttering for the Radeon GPU in Fortnite.

When we recently compared the Radeon RX 6600 and GTX 1080 in a recent head to head, we noted that the GeForce GPU was heavily underperforming in Hitman, certainly well below where you’d expect it to be.

AMD’s Vega 64 doesn’t suffer to nearly the same degree and was able to break the 60 fps barrier at 1440p, delivering almost 40% more performance with a 30% boost seen at 1080p. This is a good result for the old Radeon.

Next we have F1 22, a newly released title that supports ray tracing, but of course neither of these older GPUs support it. It’s still a new game that will require some level of driver optimization and it appears Nvidia has done a much better job in this instance as the GTX 1080 was 42% faster at 1080p and 19% faster at 1440p.

Flipping the results the other way is Warhammer III, using the medium quality preset. Vega 64 was miles faster than the GTX 1080 here, serving up just over 60% more frames at 1080p and a little over 50% more at 1440p. The game played noticeably better with the Radeon, particularly at 1440p.

Moving on to Cyberpunk 2077, we find another title where Vega 64 is able to outpace the GTX 1080, offering 17% better performance at 1080p, and a 24% more at 1440p. Another big win for the Radeon in a modern and very demanding title, though we were forced turn down quality settings to medium for smooth playable performance.

Resident Evil Village is yet another game where Vega was dominant. The GTX 1080 was able to deliver perfectly playable performance with 72 fps on average at 1440p, while Vega 64 was good for an additional 24% frames at the same resolution.

We also saw a strong 17% performance uplift at 1080p, which put the Vega graphics card well into high refresh rate territory.

War Thunder is a game that heavily favors the GeForce GTX, and while it’s an older game, it’s one that plays at significantly higher frame rates with the GTX 1080. For example, the GeForce hit 361 fps at 1080p making it 63% faster than Vega, and we also saw a notable 46% margin at 1440p in favor of the Nvidia graphics card.

Despite getting destroyed in this test, the positive note for AMD is that Vega 64 still managed to deliver an average of 174 fps at 1440p, which we believe is ample for this title.

God of War is a new title on PC that appears to play quite well using either Vega 64 or the GTX 1080. Both averaged just over 90 fps at 1080p, with over 60 fps at 1440p, so a good result for both GPUs and it’s good to see 5-6 year old products delivering highly playable performance in modern titles using reasonable quality settings.

Vega 64 was jumped back ahead in Red Dead Redemption, hitting just 60 fps at 1080p using high quality settings, though that meant it was 15% faster than the GTX 1080, which managed just 52 fps. The margin remained much the same at 1440p as Vega was 18% faster.

Vega 64 also performed well in Dying Light 2, delivering 67 fps at 1080p to make it 12% faster than the GTX 1080, not bad given it was generally slower than the GeForce GPU back in our 2018 testing.

We’re also looking at a 15% performance advantage at 1440p, though neither GPU was particularly impressive here.

The last game we’re going to look at is Watch Dogs Legion. Despite being an Nvidia sponsored title, AMD landed another win as Vega 64 pulled ahead by 12% at 1080p and 1440p.

Certainly not a massive margin but convincing enough considering back in the day Vega 64 was typically slower than the GTX 1080.

Performance Summary

Those were some interesting results and it certainly appeared from the 10+ games we just looked at that Vega 64 got the better of the GTX 1080. Of course, we did test 51 games in total, so we shouldn’t jump to any conclusions just yet

Let’s check out the full breakdowns at 1080p and 1440p…

As suspected, Vega 64 did get the better of the GTX 1080 at 1080p. Although we’re talking about a slim 3% margin, which we’d prefer to call a tie, it’s not a bad result for AMD as this GPU was at least 5% slower back in 2018.

While the results have been somewhat reversed, the truth is the GTX 1080 and Vega 64 were always evenly matched and that still holds true today. Good news for Vega, but of course, it’s too little too late and gamers should avoid both of these GPUs in 2022 as there are simply better second hand deals.

The margins are very similar at 1440p for the most part and now the GeForce GTX 1080 is 5% slower overall. It’s interesting to note that back in 2018, Dawn of War III was one of just two titles where Vega 64 could beat the GTX 1080 by double-digit margins, and today Warhammer III is the worst title test for the GeForce GPU. Also in that previous test, we saw 6 games where Vega 64 was faster by 5% or more, whereas the updated 2022 testing found 28 examples (of a larger game sample, of course).

A Blast from Radeon’s Past

We always find it interesting to look back at these old battles, not because we can give any useful or relevant consumer advice as these products are now too old for that, but because we actually get to see how it all played out. Many gamers who haven’t upgraded in the past few years are likely still running one of these cards, too.

AMD’s Vega GPU range never impressed us much. Vega 56 was decent, Vega 64 was just too inefficient, and the liquid cooled version was too expensive. Many AMD fans back in the day would argue Vega’s case, claiming undervolting fixed the efficiency issues and that the more «future proof» architecture would eventually reign supreme.

The undervolting thing was largely hit-and-miss on our testing, but they were technically right; the Vega or GCN 5th generation architecture was set to stand the test of time better, however irrelevant that point might be 5 years later.

The real shame is that it took 4+ years before Vega was truly competitive and while we just acknowledged that the overall margins haven’t changed much, there are several examples where Vega 64 was massively faster such as Warhammer III, Hitman 3, Cyberpunk 2077, Resident Evil Village and another half a dozen titles.

Back in 2018 the only big win we could find for Vega was DiRT 4 when using CMAA, other than that performance was very similar or swung heavily in favor of the GTX 1080.

Looking back, we think that for Vega 64 to be a true GTX 1080 competitor at the same $500 price point it would have needed to be at least 10-15% faster. We say this because there’s nothing that could be done about the massive power usage. Also if there had been a wider range of high quality boards to choose from, that would have helped Vega’s case, but there wasn’t.

There’s no question in our mind, for those paying MSRP pricing around 2017 and 2018 — prior to the release of Nvidia’s Turing generation (RTX 20 series) — the GTX 1080 was the better buy. Since then Radeon GPUs have become significantly more competitive and we’re excited to see what the upcoming GPU battle has in store for us. Hopefully we’ll have some answers for you before year’s end.

Shopping Shortcuts:

• GeForce RTX 3070 Ti on Amazon
• GeForce RTX 3080 on Amazon
• Radeon RX 6800 on Amazon
• Radeon RX 6600 XT on Amazon
• GeForce RTX 3060 on Amazon
• Intel Core i5-12600K on Amazon
• Intel Core i7-12700K on Amazon
• AMD Ryzen 7 5800X on Amazon
• AMD Ryzen 9 5900X on Amazon

AMD Radeon RX Vega 64 vs AMD Radeon VII

: What is the difference?

53 BALLLA

AMD Radeon RX VEGA 64

73 BALLLA

AMD Radeon VII

Winter when comparing

VS

54 Facts compared to

AMD RADEON RX VEGA 64

AMD RADEON VII000 AMD Is Vega 64 better than AMD Radeon VII?

• 256 more stream processors?
  4096 vs 3840
• 16 more texture units (TMUs)?
  256 vs 240

• GPU frequency 203MHz higher?
  1450MHz vs 1247MHz
• 1. 22 TFLOPS higher than FLOPS?
  13.82 TFLOPS vs 12.6 TFLOPS
• 131.46 GPixel/s higher pixel rate?
  230.4 GPixel/s vs 98.94 GPixel/s
• 55MHz faster memory speed?
  1000MHz vs 945MHz
• 110MHz higher effective clock speed?
  2000MHz vs 1890MHz
• 36.2 GTexels/s higher number of textured pixels?
  432 GTexels/s vs 395.8 GTexels/s
• 1.98x more VRAM?
  16GB vs 8.1GB
• 540.2GB/s more memory bandwidth?
  1024GB/s vs 483.8GB/s

Which comparisons are the most popular?

AMD Radeon RX Vega 64

vs

AMD Radeon R9 290X

AMD Radeon VII

vs

AMD Radeon Vega 8

AMD Radeon RX Vega 64

vs

Nvidia GeForce RTX 2060

AMD Radeon VII

VS

NVIDIA GeForce GTX 16500003

AMD Radeon RX VEGA 64

VS

ASUS Radeon RX 6900 XT

AMD Radeon VII

AMD Radeon RX 6800 XT

VS 9000 9000 AMD RADEN AMD Radeon RX Vega 64

vs

AMD Radeon Vega Frontier Titan Edition

AMD Radeon VII

Nvidia 9 vs

0003

AMD Radeon RX Vega 64

vs

Nvidia GeForce GTX 1650

AMD Radeon VII

vs

Sapphire Radeon RX Vega 64

AMD Radeon RX Vega 64

vs

Nvidia GeForce RTX 3060 Ti

AMD Radeon VII

VS

GIGABYTE RADEON RX 6600 XT Eagle

AMD Radeon RX Vega 64

AMD Radeon RX 6800 XT

AMD Radeon VII

VS

000 AMD RADAR

000 AMD RADEN0003

Price Match

User Reviews

Performance

1. GPU clock speed

1247MHz

1450MHz

The graphics processing unit (GPU) has a higher clock speed.

2.turbo GPU

1546MHz

1800MHz

When the GPU is running below its limits, it can jump to a higher clock speed to increase performance.

3.pixel speed

98.94 GPixel/s

230.4 GPixel/s

The number of pixels that can be displayed on the screen every second.

4.flops

12.6 TFLOPS

13.82 TFLOPS

FLOPS is a measure of GPU processing power.

5.texture size

395.8 GTexels/s

432 GTexels/s

Number of textured pixels that can be displayed on the screen every second.

6.GPU memory speed

945MHz

1000MHz

Memory speed is one aspect that determines memory bandwidth.

7. shading patterns

Shading units (or stream processors) are small processors in a video card that are responsible for processing various aspects of an image.

8.textured units (TMUs)

TMUs accept textured units and bind them to the geometric layout of the 3D scene. More TMUs generally means texture information is processed faster.

9 ROPs

ROPs are responsible for some of the final steps of the rendering process, such as writing the final pixel data to memory and for performing other tasks such as anti-aliasing to improve the appearance of graphics.

Memory

1.memory effective speed

1890MHz

2000MHz

The effective memory clock frequency is calculated from the memory size and data transfer rate. A higher clock speed can give better performance in games and other applications.

2. max memory bandwidth

483.8GB/s

1024GB/s

This is the maximum rate at which data can be read from or stored in memory.

3.VRAM

VRAM (video RAM) is the dedicated memory of the graphics card. More VRAM usually allows you to run games at higher settings, especially for things like texture resolution.

4.memory bus width

2048bit

4096bit

Wider memory bus means it can carry more data per cycle. This is an important factor in memory performance, and therefore the overall performance of the graphics card.

5.GDDR memory versions

Unknown. Help us offer a price. (AMD Radeon RX Vega 64)

Unknown. Help us offer a price. (AMD Radeon VII)

Later versions of GDDR memory offer improvements such as faster data transfer rates, which improves performance.

6. Supports memory debug code

✖AMD Radeon RX Vega 64

✖AMD Radeon VII

Memory debug code can detect and fix data corruption. It is used when necessary to avoid distortion, such as in scientific computing or when starting a server.

Functions

1.DirectX version

DirectX is used in games with a new version that supports better graphics.

2nd version of OpenGL

The newer version of OpenGL, the better graphics quality in games.

OpenCL version 3.

Some applications use OpenCL to use the power of the graphics processing unit (GPU) for non-graphical computing. Newer versions are more functional and better quality.

4. Supports multi-monitor technology

✔AMD Radeon RX Vega 64

✔AMD Radeon VII

The video card has the ability to connect multiple displays. This allows you to set up multiple monitors at the same time to create a more immersive gaming experience, such as a wider field of view.

5. GPU temperature at boot

Unknown. Help us offer a price. (AMD Radeon RX Vega 64)

Unknown. Help us offer a price. (AMD Radeon VII)

Lower boot temperature — this means that the card generates less heat and the cooling system works better.

6.supports ray tracing

✖AMD Radeon RX Vega 64

✖AMD Radeon VII

Ray tracing is an advanced light rendering technique that provides more realistic lighting, shadows and reflections in games.

7. Supports 3D

✖AMD Radeon RX Vega 64

✖AMD Radeon VII

Allows you to view in 3D (if you have a 3D screen and glasses).

8.supports DLSS

✖AMD Radeon RX Vega 64

✖AMD Radeon VII

DLSS (Deep Learning Super Sampling) is an AI based scaling technology. This allows the graphics card to render games at lower resolutions and upscale them to higher resolutions with near-native visual quality and improved performance. DLSS is only available in some games.

9. PassMark result (G3D)

Unknown. Help us offer a price. (AMD Radeon RX Vega 64)

This test measures the graphics performance of a graphics card. Source: Pass Mark.

Ports

1.has HDMI output

✔AMD Radeon RX Vega 64

✔AMD Radeon VII

Devices with HDMI or mini HDMI ports can stream HD video and audio to the connected display.

2.HDMI connectors

More HDMI connectors make it possible to connect multiple devices at the same time, such as game consoles and TV sets.

3rd HDMI version

Unknown. Help us offer a price. (AMD Radeon RX Vega 64)

Unknown. Help us offer a price. (AMD Radeon VII)

Newer versions of HDMI support higher bandwidth, resulting in higher resolutions and frame rates.

4. DisplayPort 9 outputs0003

Allows you to connect to a display using DisplayPort.

5.DVI outputs

Allows connection to a display using DVI.

Mini DisplayPort 6.outs

Allows connection to a display using Mini DisplayPort.

Price Match

Cancel

Which graphics cards are better?

Which is better AMD Radeon RX 580 or Sapphire Radeon RX Vega 64

Sapphire Radeon RX Vega 64

AMD Radeon RX 580

GPU base clock

The graphics processing unit (GPU) has a high clock speed.

1247MHz

max 2457

Average: 938 MHz

1257MHz

max 2457

Average: 938 MHz

GPU memory frequency

This is an important aspect calculating memory bandwidth

945MHz

max 16000

Average: 1326. 6 MHz

2000MHz

max 16000

Average: 1326.6 MHz

FLOPS

The measurement of processing power of a processor is called FLOPS.

12.45TFLOPS

max 1142.32

Average: 92.5 TFLOPS

6.4TFLOPS

max 1142.32

Average: 92.5TFLOPS

Turbo GPU

If the GPU speed drops below its limit, it can switch to a high clock speed to improve performance.
Show all

1546 MHz

max 2903

Average: 1375.8 MHz

1340MHz

max 2903

Average: 1375. 8 MHz

Texture size

A certain number of textured pixels are displayed on the screen every second.
Show all

395.8 GTexels/s

max 756.8

Average: 145.4 GTexels/s

193 GTexels/s

max 756.8

Average: 145.4 GTexels/s

Architecture name

VEGA

GCN 4.0

Graphic processor name

VEGA

Polaris 20

DECARY memory 9000

NO

No data no data no data0003

Memory bandwidth

This is the speed at which the device stores or reads information.

483.8GB/s

max 2656

Average: 198. 3 GB/s

256GB/s

max 2656

Average: 198.3 GB/s

Effective memory speed

The effective memory clock speed is calculated from the size and information transfer rate of the memory. The performance of the device in applications depends on the clock frequency. The higher it is, the better.
Show all

1890MHz

max 19500

Average: 6984.5 MHz

8000MHz

max 19500

Average: 6984.5 MHz

RAM

8GB

max 128

Average: 4.6 GB

8GB

max 128

Average: 4. 6 GB

GDDR Memory Versions

Latest GDDR memory versions provide high data transfer rates to improve overall performance
Show all

5

Average: 4.5

5

Average: 4.5

Memory bus width

A wide memory bus means that it can transfer more information in one cycle. This property affects the performance of the memory as well as the overall performance of the device’s graphics card.
Show all

2048bit

max 8192

Average: 290.1bit

256bit

max 8192

Average: 290.1bit

Heat dissipation (TDP)

Heat dissipation requirement (TDP) is the maximum amount of energy that can be dissipated by the cooling system. The lower the TDP, the less power will be consumed.
Show all

295W

Average: 140.4W

185W

Average: 140.4W

Process

The small size of the semiconductor means it is a new generation chip.

14 nm

Average: 47.5 nm

14 nm

Average: 47.5 nm

Number of transistors

The higher their number, the more processor power it indicates

12500 million

max 80000

Average: 5043 million

5700 million

max 80000

Average: 5043 million

PCIe version

Considerable speed is provided by the expansion card used to connect the computer to peripherals. The updated versions have impressive throughput and provide high performance.
Show all

3

Mean: 2.8

3

Mean: 2.8

Width

272mm

max 421.7

Average: 242.6mm

241mm

max 421.7

Average: 242.6mm

DirectX

Used in demanding games providing enhanced graphics

12

max 12.2

Average: 11.1

12

max 12.2

Average: 11.1

OpenCL version

Used by some applications to enable GPU power for non-graphical calculations. The newer the version, the more functional it will be
Show all

2

max 4.6

Average: 1.7

2.1

max 4.6

Average: 1.7

opengl version

Later versions provide better game graphics

4.5

max 4.6

Average: 4

4.6

max 4.6

Average: 4

Shader model version

6.4

max 6.6

Average: 5.5

6.4

max 6.6

Average: 5. 5

Vulkan version

1.1

1.2

Has hdmi output

HDMI output allows you to connect devices with HDMI or mini HDMI ports. They can transmit video and audio to the display.

Yes

Yes

HDMI version

The latest version provides a wide signal transmission channel due to the increased number of audio channels, frames per second, etc.
Show all

2

max 2.1

Average: 2

2

max 2.1

Average: 2

DisplayPort

Allows connection to a display using DisplayPort

3

Average: 2

3

Average: 2

Number of HDMI sockets

The more there are, the more devices can be connected at the same time (for example, game/TV type consoles)
Show in full

one

Average: 1. 1

one

Average: 1.1

HDMI

Yes

Yes

Passmark score

14327

max 29325

Average: 7628.6

8680

max 29325

Average: 7628.6

3DMark Cloud Gate GPU test score

124827

max 1

Average: 80042.3

80866

max 1

Average: 80042.3

3DMark Fire Strike Score

18001

max 38276

Average: 12463

11746

max 38276

Average: 12463

3DMark Fire Strike Graphics test score

22051

max 49575

Average: 11859. 1

13648

max 49575

Average: 11859.1

3DMark 11 Performance GPU score

30208

max 57937

Average: 18799.9

18889

max 57937

Average: 18799.9

3DMark Vantage Performance Score

54157

max 97887

Average: 37830.6

43457

max 97887

Average: 37830.6

3DMark Ice Storm GPU score

384458

max 533357

Average: 372425. 7

341973

max 533357

Average: 372425.7

SPECviewperf 12 test score — Solidworks

78

max 202

Average: 62.4

max 202

Average: 62.4

SPECviewperf 12 test score — specvp12 sw-03

79

max 202

Average: 64

max 202

Average: 64

SPECviewperf 12 evaluation — Siemens NX

23

max 212

Average: 14

max 212

Average: 14

SPECviewperf 12 test score — specvp12 showcase-01

109

max 232

Average: 121. 3

max 232

Average: 121.3

SPECviewperf 12 score — Showcase

109

max 175

Average: 108.4

max 175

Average: 108.4

SPECviewperf 12 test score — Medical

49

max 107

Average: 39.6

max 107

Average: 39.6

SPECviewperf 12 test score — specvp12 mediacal-01

49

max 107

Average: 39

max 107

Average: 39

SPECviewperf 12 test score — Maya

80

max 177

Average: 129. 8

max 177

Average: 129.8

SPECviewperf 12 test score — specvp12 maya-04

82

max 180

Average: 132.8

max 180

Average: 132.8

SPECviewperf 12 test score — Energy

12

max 25

Average: 9.7

max 25

Average: 9.7

SPECviewperf 12 test score — specvp12 energy-01

12

max 21

Average: 10.7

max 21

Average: 10.7

SPECviewperf 12 — Creo 9 test score0422

57

max 153

Average: 49. 5

max 153

Average: 49.5

SPECviewperf 12 test score — specvp12 creo-01

57

max 153

Average: 52.5

max 153

Average: 52.5

SPECviewperf 12 test score — specvp12 catia-04

154

max 189

Average: 91.5

max 189

Average: 91.5

SPECviewperf 12 evaluation — Catia

155

max 189

Average: 88.6

max 189

Average: 88.6

SPECviewperf 12 test score — specvp12 3dsmax-05

142

max 316

Average: 189. 5

max 316

Average: 189.5

SPECviewperf 12 test score — 3ds Max

138

max 269

Average: 169.8

max 269

Average: 169.8

Vega 64 OC vs Radeon VII OC in 4K

This content has been written by a website visitor and has been rewarded.

Hello dear readers! In today’s article, two champions will come together in a confrontation: the world’s best in 3DMark Time Spy Extreme by graphics score Vega 64, and the world’s best Radeon VII in 3DMark Time Spy Extreme among Radeon VII + Ryzen 7 5800X bundles. Both video cards are close relatives based on the GCN architecture: Radeon VII is the same Vega 64 on a more modern technical process, with an increased amount of video memory, but a smaller number of blocks by 8%.

Image source — Hardware Battle resource

1. Test configurations

recommendations

Test configuration 1:

Processor: AMD Ryzen 7 5800X (Vermeer, L3 32 MB), 8 cores / 160 MHz 50.0 threads , limits motherboard

Video card: PowerColor Radeon VII 2125/1200 MHz, HBCC on.

Motherboard: ASRock X470 Taichi, AM4

CPU cooling system: custom CBO

Video card cooling system: custom CBO

RAM: G. Skill Trident Z 4×8 GB 3667CL14

Storage: 1TB NVMe M.2 SSD Samsung 970 PRO

Power Supply: Super Flower Leadex Platinum 1200 Watt

Chassis: Phanteks Enthoo Evolv0 ATX 9003 Monitor Acer Nitro XV273KP 27″ IPS 3840×2160 144Hz

Software: Windows 10 Pro Version 2004

Graphics Driver: AMD Radeon Software Adrenalin 21.2.3

Test Configuration 2:

CPU:

CPU:

MB), 6 cores/12 threads, 3600@4550 MHz

Video card: Sapphire RX Vega 64 Liquid Cooling 1680/1108 MHz

Motherboard: Asus TUF B450 Pro Gaming, AM4

CPU cooling system: CBO Celsius S36

Video card cooling system: stock memory Crucial

Sport LT 2×8 GB 3800CL15

Storage: 512 Gb NVMe M. 2 SSD Samsung 970 Evo+

PSU: Zalman ZM1000 EBT 1000 Watt

Chassis: Fractal Design Define R6 black 9Ol000 Games, settings, test results

All games were tested at a resolution of 3840×2160 to minimize the impact of the processor on the result.To compare the results, we used the benchmarks built into the games.0003

Assassin´s Creed: Odyssey

version of the game — 1.05, API — DX11, Preset — the highest

Result VEGA 64

Result Radeon VII

in AC: ODYSSEY RADEON VII by 17%.

Deus Ex: Mankind Divided

Game version — license with all updates, API — DX12, preset — Ultra, MSAA — off.

Result Vega 64

Result Radeon VII

Deus Ex: MD Radeon VII is 21% faster than Vega 64.

Far Cry 5

Game version — license with all updates, API — DX11, preset — Max.

Vega 64 result

Radeon VII result

Far Cry V Radeon VII is 18% faster than Vega 64.

Strange Brigade

Game version: 1.47.22.14, API — Vulkan, preset — Ultra, Async — on.

Result Vega 64

Result Radeon VII

In Strange Brigade in the Vulkan API, Radeon VII is 28% faster than Vega 64 in average fps, but exhibits abnormally low minimum fps. To check the capabilities of Radeon VII, I tested the game in the DX12 API — the average fps result remained practically unchanged, but the minimum fps increased by 60%.

Shadow of The Tomb Raider

Game version — license with all updates, API — DX12, preset — Max

Vega 64 result

Radeon VII result

In SoTR, Radeon VII is 15% faster than Vega 64.

World War Z

Game version — 1.70, API — Vulkan, preset — Ultra.

Vega 64 result

Radeon VII result

In WWZ, Radeon VII is 14% faster than Vega 64.

Total War Saga: Troy

Game version — 1. 2.0, API — DX11. Since the preset ultra settings have been changed, I give a screen of graphics settings

Vega 64 result (32 fps)

Radeon VII result

In Total War Saga: Troy Radeon VII is 21% faster than Vega 64.

Metro: Exodus

Game version — 1.07.16, API — DX12, preset — custom, based on Ultra. Disabled settings are on the screenshot of the benchmark result.

Vega 64 result

Radeon VII result

Metro: Exodus Radeon VII is 14% faster than Vega 64.

Horizon: Zero Dawn

Game version — 1.0.10.5, API — DX12, preset — Ultra.

Vega 64 result

Radeon VII result

Horizon: Zero Dawn Radeon VII is 23% faster than Vega 64.

The final table of results:

3. Conclusion

Vega 64 is still a relevant video card, which can comfortably play in 1440p, and, by lowering a number of settings, in 2160p. Nevertheless, it is not a competitor to the Radeon VII with high-quality cooling, which today’s article showed.

Thanks to all overclockers and gamers for your attention, special thanks to Mr. RadeoForce for providing Vega 64 test results.

Comparison of AMD Radeon VII and AMD Radeon RX Vega 64 Liquid

Comparative analysis of AMD Radeon VII and AMD Radeon RX Vega 64 Liquid video cards by all known characteristics in the categories: General information, Specifications, Video outputs and ports, Compatibility, dimensions, requirements, API support, Memory, Technology support.
Analysis of video card performance by benchmarks: PassMark — G3D Mark, PassMark — G2D Mark, Geekbench — OpenCL, CompuBench 1.5 Desktop — Face Detection (mPixels/s), CompuBench 1.5 Desktop — Ocean Surface Simulation (Frames/s), CompuBench 1.5 Desktop — T -Rex (Frames/s), CompuBench 1.5 Desktop — Video Composition (Frames/s), CompuBench 1.5 Desktop — Bitcoin Mining (mHash/s), GFXBench 4. 0 — Car Chase Offscreen (Frames), GFXBench 4.0 — Manhattan (Frames), GFXBench 4.0 — T-Rex (Frames), GFXBench 4.0 — Car Chase Offscreen (Fps), GFXBench 4.0 — Manhattan (Fps), GFXBench 4.0 — T-Rex (Fps), 3DMark Fire Strike — Graphics Score.

AMD Radeon VII

versus

AMD Radeon RX Vega 64 Liquid

Benefits

Reasons to choose AMD Radeon VII

• Newer graphics card, release date difference 1 year(s) 5 month(s)
• % more: 1750 MHz vs 1677 MHz
• 1% more texturing speed: 432.24 GT/s vs 429.3 GTexel / s
• A newer manufacturing process for making the video card more powerful, but with lower power consumption: 7 nm vs 14 nm
• Max memory size is 2x larger: 16 GB vs 8 GB
• Memory clock is 6% faster: 2000 MHz vs 1890 MHz
• Performance in 3DMark Fire Strike — Graphics Score is about 18% faster: 9129 vs 7765

Issue date	7 February 2019 vs 8 August 2017
Boost Core Clock	1750 MHz vs 1677 MHz
Texturing speed	432. 24 GT/s vs 429.3 GTexel/s
Process	7nm vs 14nm
Maximum memory size	16 GB vs 8 GB
Memory frequency	2000 MHz vs 1890 MHz
3DMark Fire Strike — Graphics Score	9129 vs 7765

Reasons to choose AMD Radeon RX Vega 64 Liquid

• 7% more shader processors: 4096 vs 3840
• About 33% less power consumption: 225 Watt vs 300 Watt

Core frequency	1406 MHz vs 1400 MHz
Number of shaders	4096 vs 3840
Power Demand (TDP)	225 Watt vs 300 Watt

Benchmark comparison

GPU 1: AMD Radeon VII
GPU 2: AMD Radeon RX Vega 64 Liquid

3DMark Fire Strike — Graphics Score

GPU 1 GPU 2

Name	AMD Radeon VII	AMD Radeon RX Vega 64 Liquid
PassMark — G3D Mark	16825
PassMark — G2D Mark	789
Geekbench — OpenCL	86533
CompuBench 1. 5 Desktop — Face Detection (mPixels/s)	202.389
CompuBench 1.5 Desktop — Ocean Surface Simulation (Frames/s)	4610.698
CompuBench 1.5 Desktop — T-Rex (Frames/s)	21.511
CompuBench 1.5 Desktop — Video Composition (Frames/s)	69.204
CompuBench 1.5 Desktop — Bitcoin Mining (mHash/s)	481.759
GFXBench 4.0 — Car Chase Offscreen (Frames)	11724
GFXBench 4. 0 — Manhattan (Frames)	3717
GFXBench 4.0 — T-Rex (Frames)	3360
GFXBench 4.0 — Car Chase Offscreen (Fps)	11724
GFXBench 4.0 — Manhattan (Fps)	3717
GFXBench 4.0 — T-Rex (Fps)	3360
3DMark Fire Strike — Graphics Score	9129	7765

Feature comparison

conveyors

	AMD Radeon VII	AMD Radeon RX Vega 64 Liquid
Architecture	GCN 5. 1	GCN 5.0
Codename	Vega 20	Vega 10
Design	AMD Radeon 2nd generation Vega
Generation GCN	5th Gen
Production date	February 7, 2019	August 8, 2017
Price at first issue date	$699
Place in the rating	98	66
Price now	$699. 99
Type	Desktop	Desktop
Price/performance ratio (0-100)	32.74
Boost Core Clock	1750 MHz	1677MHz
Number of Compute	60
Core clock	1400MHz	1406 MHz
Process	7nm	14nm
Peak Double Precision (FP64) Performance	3. 46 TFLOPs
Peak Half Precision (FP16) Performance	27.7TFLOPS
Peak Single Precision (FP32) Performance	13.8 TFLOPs
Number of shaders	3840	4096
Pixel fill rate	115.26GP/s
Stream Processors	3840
Texturing speed	432.24 GT/s	429.3 GTexel/s
Power Demand (TDP)	300 Watt	225 Watt
Number of transistors	13,230 million
Floating point performance		13. 738 gflops
Video connectors	1x HDMI, 3x DisplayPort	1x HDMI, 3x DisplayPort
Number of DisplayPort	3
DisplayPort support
HDMI
VGA
Interface	PCIe 3.0 x16	PCIe 3.0 x16
Length	305 mm	267 mm
Recommended Power Supply	750W
Additional power connectors	2x 8-pin	2x 8-pin
DirectX	12. 0
OpenCL	2.0
OpenGL	4.6
Vulcan
High Bandwidth Memory (HBM)
Maximum memory size	16GB	8GB
Memory Bandwidth	1024 GB/s	483.8 GB/s
Memory bus width	4096 Bit	2048 Bit
Memory frequency	2000MHz	1890 MHz
Memory type	HBM2	HBM2
AMD Eyefinity
AMD Radeon™ Chill
AMD Radeon™ ReLive
FreeSync
PowerTune
TressFX
TrueAudio
Unified Video Decoder (UVD)
Video Code Engine (VCE)
Virtual Super Resolution (VSR)
VR Ready
ZeroCore		9 Part 1 >> Part 2>> Part 3 In 2018, the AMD Radeon Technologies Group allowed itself a short break. After launching the Polaris architecture in 2016, and the Vega architecture in 2017, AMD decided to enjoy the first full year on the Vega architecture. Instead of launching a third architecture after 3 years, the company will focus on further expanding the family by bringing variants of Vega laptops and servers to market. While AMD’s notebook endeavors have gone in a strange direction to say the least, work on the Radeon Instinct server has led to positive results. As a result, the company announced the first 7nm GPU. With the latest generation product update in November, in the form of the Radeon RX 590, we expected AMD to focus on the consumer for a while. Instead, AMD made a rather unexpected announcement at CES 2019: the company will release a new high-end consumer card, the Radeon VII (Seven). Based on the aforementioned server GPU, the graphics card is positioned as the latest flagship. Designed for gamers and content creators alike, Radeon VII should be AMD’s decisive step towards gamers. It’s our turn to take a look at AMD’s latest card. At first glance, the Radeon VII seems obvious. The silicon core of the card is the AMD Vega 20 GPU, a derivative of the original Vega 10 that has been enhanced for scientific computing and machine learning, and is built on TSMC’s cutting-edge 7nm process for improved performance. This marks a major milestone for AMD in server GPU development as they released the first high-performance server-class GPU since GPU Hawaii in 2013. AMD was eager to showcase the Vega 20 late in its development, as this GPU is the heart of AMD’s relatively new Radeon Instinct MI50 and MI60 server accelerators. Vega 20, primarily intended for servers, does not belong to the class of GPUs that could find their way to the consumer cheaply and easily. But something unexpected happened: NVIDIA made almost no difference in performance per dollar. Instead, the new Turing-based GeForce RTX cards focus on options that are meant to usher in a new paradigm for rendering games with real-time ray-traced effects, and dedicate more resources to Turing’s GPUs for that purpose. The end result has been relatively high prices for the GeForce RTX 20 series cards, although their performance gains in games are much less than usual when changing generations of graphics cards. Faced with a less hostile pricing environment than initially anticipated, AMD decided to finally bring the Vega to 20 consumers, battling NVIDIA at one of those high price tiers. The Radeon VII hits the mainstream at $699 and the GeForce GTX 2080 becomes Radeon’s new flagship gaming card. Looking at the spec sheet, the Radeon VII comes with a peak clock speed of 1800 MHz, while the official peak frequency is 1750 MHz. This is a nice improvement over the RX Vega 64, which was only clocked at 1630 MHz, meaning AMD still has about a 10% boost in frequency. And with an open cooler and a revised SMU, Radeon VII will be able to boost even faster and better support high clock speeds. So while AMD’s latest card doesn’t add more ROPs or CUs (in fact, there’s a slight drop from the RX Vega 64), it takes bandwidth to the next level. However, the biggest change from the RX Vega 64 is that AMD has doubled the amount of memory, and more than doubled the memory bandwidth. This is possible thanks to the 7nm process technology, due to which the latest AMD GPU has a relatively modest die size of 331mm2. The extra space allowed an AMD interposer and two more HBM2 stacks, allowing for more VRAM and a wider memory bus. AMD was also able to increase the memory frequency slightly, from 1.89Gbps/pin on RX Vega 64 to flat 2 Gbps/pin on Radeon VII. Interestingly, based on the base specs, the Radeon VII is essentially a Radeon Instinct MI50. Thus, AMD has the risk of «cannibalizing» sales of Instinct, if the high performance of the Radeon VII will interest professional computer users. As a result, AMD has cut some features of the chip to better differentiate products. We’ll talk about this a little later, but the main thing is that the card runs at a lower frequency than FP64, loses full support for ECC chips, and, naturally for a consumer product, uses Radeon Software gaming drivers instead of the Instinct professional driver stack. Of course, whenever you talk about connecting a server GPU to a consumer graphics card, you are talking about a powerful card with a lot of potential, and this certainly applies to the Radeon VII. In a nutshell, the advantage of AMD’s latest flagship card is high performance, and the highest in class 16GB of HBM2 memory. As one of the few obvious advantages of AMD’s specification over NVIDIA’s competitors, VRAM capacity is an important part of AMD’s marketing; they will focus heavily on content creation and VRAM-intensive games. New to this card is a triple-fan cooler, replacing the single-fan cooler on the Radeon RX Vega 64/56 cards. Also, another nifty little change: AMD enters the retail market as a board supplier on its own and directly sells the new card at a set price of $699. Given that AIB is also releasing their branded reference cards today, this is an opportunity to avoid inflated prices. to launch. Meanwhile, looking at the competitive situation, there are a couple of issues that need to be addressed. An important part of the offer is (as has become common in recent times) a set of games. The current Raise the Game Fully Loaded pack includes Devil May Cry 5, The Division 2 and Resident Evil 2 for free with Radeon VII, RX Vega and RX 59 cards0. Meanwhile, the RX 580 and RX 570 cards allow you to choose two games out of three. As expected, the game bundle would just be an added value over a direct competitor — in this case the RTX 2080 — but NVIDIA has its own Game On duel bundle with Anthem and Battlefield V. In a scenario where the Radeon VII is expected to fight for consumers with the RTX 2080, with no chance of ultimate victory, these additional benefits are becoming increasingly important. The Radeon VII marks the first product launch since the recent change in competitive landscape due to variable refresh rate monitor technology. Variable refresh rate monitors have already become a must-have for gamers, and since the technology’s launch earlier this decade, there’s been a clear divide between AMD and NVIDIA cards. AMD cards support VESA Adaptive Sync, better known under the AMD FreeSync brand, while NVIDIA desktop cards only support their proprietary G-Sync. But last month, NVIDIA unexpectedly announced that their cards will support VESA Adaptive Sync under the brand name ‘G-Sync Compatibility’. Details on how this program works are scarce, but after all, adaptive sync can be used in NVIDIA drivers even if the FreeSync panel is not certified G-Sync Compatible. The end result is that while NVIDIA’s news doesn’t stop AMD from realizing features, it undermines AMD’s advantage in FreeSync — all cheap VESA Adaptive Sync monitors that were previously only used on AMD cards can now be used on cards as well. NVIDIA. AMD was, of course, very happy to emphasize the «free» FreeSync, but as a weapon against NVIDIA, this marketing ploy has lost its force. AMD’s official line is to consider this a victory for FreeSync. The release of the Radeon VII and its competitive positioning against the GeForce RTX 2080 means that AMD needs to be clear on its position on the current capability gap between their cards and the latest NVIDIA Turing. Therefore, AMD’s position has remained unchanged in DirectX Raytracing (DXR) technologies and AI-based image quality and performance technologies such as DLSS. In short, AMD’s argument is that they don’t think the difference in image quality and performance gains are worth the price of these features. In the meantime, AMD is not standing still, and together with the backup DXR drivers, they are working on support for WinML and DirectML for their cards. AMD Risk: If DXR or NVIDIA’s DLSS efforts start to move quickly, the feature gap will become a much bigger issue. In general, releasing a 7nm gaming GPU to consumers is currently a very aggressive move. Especially at an early stage of the life cycle of this process. But AMD simply doesn’t have time to wait and see. The only major obstacle to product promotion will be pricing: the cost must be acceptable to consumers. This brings us to today’s launch. With a price tag of $699, NVIDIA has already lowered the price bar by talking about dedicated hardware to accelerate ray tracing and machine learning. For the Radeon VII, the terms revolve around 16GB HBM2 and value for semi-professionals and content creators. All that remains is their gaming performance. Under the hood of Vega 20: GCN at 7 nm While we’re already familiar with the Vega 20 through the release of the Radeon Instinct MI50 and MI60 server-class graphics cards, this chip’s entry into the consumer market marks the first 7nm gaming card. Rapid down-node hopping — this time from GlobalFoundries’ 14nm LPP to TSMC’s 7nm (CLN7FF) — has been AMD/ATi’s hallmark before, and here again AMD is in a rush to release, bringing the product to consumers so early. In this case, all the attributes of Vega 20 are clearly intended for professionals and servers, although they also have advantages for games. In the complicated history of the Radeon VII, Vega’s fate was especially mysterious. For starters, the 7nm Vega 20 GPU making its way to consumers so early came as a surprise. And while AMD has only mentioned Vega 20 in relation to Radeon Instinct products — referring to its upcoming Navi architecture when referring to 7nm GPUs for gamers — AMD now claims the release of the Radeon VII to the consumer market was long overdue. Perhaps the same could be said for the Vega at 14nm+/12LP and the Vega 11 GPU (not to be confused with the 11 Ryzen 2400G Vega compute units), although this is hardly unusual given the nature of semiconductor development. To be fair, AMD has been a bit lukewarm about Vega ever since the launch of RX Vega, which didn’t quite reach its goals. Even though the Radeon VII is in some ways a «Schrödinger chip», it raises some interesting questions about 7nm. For example, AMD has already pushed back the sampling and launch of Vega 20 earlier. It turns out that the timing of the launch of Radeon VII is the earliest that could be for the heir to Radeon Instinct. Especially since with a die size of 331 mm2, these are not the small mobile SoCs that we have seen so far on TSMC 7 nm. Designed with compute/ML-centric improvements in mind, equipped with 4 HBM2 stacks, and built on a mature and cutting-edge 7nm node, the Vega 20 GPU doesn’t look like a card ready to make a good consumer price offering. And yet, thanks to a happy coincidence, it happened. Vega 20 combines the updated GCN 5 architecture with 7nm process technology, with 13.2V at 331mm2 transistors (compared to 12.5V at 496mm2 for Vega 10). As a rule, as the size decreases, the space savings are reinvested in more transistors. For a gaming graphics card, this could mean anything from more CUs and feature blocks to redesigned layouts and increased data paths to increase frequency tolerance. The latter, of course, should provide a higher clock speed, and this choice was an important part of the Vega 10 design, where a significant number of transistors were used to achieve the required frequency response. Combined with the power efficiency of a smaller technology node, the new chip can achieve these higher clock speeds without consuming additional power. However, in Vega 20, most of the saved space was left «as is»: there is more free space on the chip. There are several reasons for this, some obvious and some not. To begin with, for a large high-performance GPU on a cutting-edge 7nm node early in the lifecycle, development and manufacturing are very costly. And it is likely that as the size increases, the cost will increase significantly, and the profitability of the product will decrease. And while TSMC’s 7nm process has only been publicly seen in mobile SoCs so far, Vega 20 appears to be using HPC-centric 7.5T libraries, as opposed to using 6T libraries designed for mobile SoCs. But more importantly, the space saved allows two more HBM2 stacks to be placed on an intermediate device of the same size. For current generation HBM2 densities and capabilities, the limit for a dual-stack chip is 16 GB of memory when using an «8-Hi» stack pair. But for a server-grade GPU — especially one focused on machine learning — a four-stack configuration is needed to provide 32 GB of memory and a wider 4096-bit bus. For the Vega 20, AMD has opted for just such a configuration, and is also releasing both 32GB (8-Hi) and 16GB (4-Hi) versions of the chip. Radeon VII, in turn, uses one of these 16 GB chips. It should be noted that this is not the first 16 GB Vega card from AMD — they have already released a card with this amount of memory. This was a dedicated Vega Frontier Edition, but after its retirement, this is the first 16 GB graphics card release in the AMD Vega line. Equipping a consumer card with 16 GB of memory is a gamble on AMD’s part. And I suspect that’s one of the reasons AMD wants to capture some of the professional visualization market with the Radeon VII. When it comes to workstation usage and content creation tasks, a lot of VRAM is easy to sell as there are already tasks that can use all that VRAM and more. But for gaming, this is not the smartest suggestion, because games have more fixed requirements for VRAM, and so far there have been no such large cards, and developers have not yet begun to use the 16 GB VRAM resource. Although on the other hand, the loud statement “this card has more than enough video memory” can have a serious impact on sales, and in 2019A flagship graphics card for gamers should have this much volume anyway. Returning to the Vega 20 design, another step AMD has taken to reduce development complexity and cost at 7nm is to stick to a known architecture. AMD added some small optimizations compared to Vega 10, but they didn’t risk a major redesign. In fact, this is the logic of the tick step of the famous tick-tock strategy of Intel. In fact, Vega 20 is such a direct successor to Vega 10 that, apart from the number of memory controllers, the number of other functional blocks is the same. The GPU combines 64 CUs and 256 texture units divided into 4 Shader Engines, as well as 64 ROPs and an AMD GPU. (I should add that in following their plan, AMD is also carefully avoiding the issue of shader engine scaling. The nature of the 4 Shader Engine limit has remained unclear in recent years, but with Vega there have been hints of going beyond these limits with better load balancing via «intelligent workgroup distributors» (IWD) Regardless, setting up and redesigning a balanced 4+ SE configuration is a difficult task, and doesn’t make sense if AMD plans to make fundamental changes to GCN. ) Thus, on an architectural level, the Vega 20 is a purely evolutionary design. But with that said, there’s a bit more «evolution» to it than just downsizing the tech node. The combination of these improvements means that in practice the Vega 20 should be slightly faster than the Vega 10 when running at the same clock speed. The most notable improvement here is the extra memory bandwidth; twice as much per ROP, texture unit, and ALU than Vega 10. This is especially good for ROPs, which traditionally need a lot of bandwidth. Not stopping there, AMD has also made some improvements to the Core Fabric that connects memory to ROP (among other things). Unfortunately, AMD does not want to divulge what these improvements are, only confirming that there are no cache changes among them. Another piece of the puzzle is that AMD has added new instructions and data types that will speed up machine learning in some cases. AMD hasn’t given us full details, but in general we do know that they have added support for the INT8 and INT4 data types, which are useful for low-precision inference scenarios. AMD also added the FP16 dot product, which is stored as the result of FP32. This is a rather specific scenario that is useful for some machine learning algorithms because it gives a higher accuracy result than the FP16-in / FP16-out dot product. Speaking of data types, AMD has also significantly increased the performance of FP64 for Vega 20. As a base architecture, GCN allows GPUs to be built at 1/2 to 1/16 the speed of FP32. For pure consumer GPUs, this has always been 1/16. However, for GPUs that carried the multifunctional load as server-oriented chips, 1/4 and 1/2 speeds were available. Vega 20, in turn, is the first ½ FP64 processor from AMD since Hawaii in 2013. This means that while its overall FP32 performance gain over Vega 10 cards isn’t as noticeable, its FP64 performance is much better: more than 8x over RX Vega 64 (theoretically). Of course, as a consumer card, the Radeon VII can’t really take advantage of these benefits — it’s limited to 1/4 speed — but more on that later. But when it comes to video blocks and display controllers, AMD has made only minor updates. Officially, the display controller (DCE) has been updated to version 12.1, while the version of the unified video decoder (UVD) is 7.2, and the version of the video coding engine (VCE) is 4.1. No additional encoding or decoding support has been added compared to Vega 10. We also found Display Core Next blocks from Raven Ridge and Video Core Next 1.0, so this may be the last architecture that these ASICs have honored with their presence. Function block wrapping is a new SMU style discussed in recent Linux kernel patches. Another feature is improved thermal monitoring, now the number of temperature diodes has been doubled to 64 sensors. As a consequence, AMD has now completely switched to node temperature monitoring instead of edge monitoring. Node temperature measurements were still used in Vega 10 (displayed as ‘hotspot’ temperature), but Vega 20 completely switched to node temperature measurement for all cooler speeds, turbo frequencies and the like. The result is more accurate reporting, as well as the marginal 1-2% performance gain AMD claims in internal testing. The updated SMU, as a consequence of the above, has the following effect: temperature, clock speed and related GPU metrics are no longer read through registers, but are instead retrieved directly from the SMU. Naturally, this breaks compatibility with third-party utilities (like GPU-Z), and while AMD has already notified some developers of the changes, applications will need to be updated to use AMD’s new API calls. Finally, since this is AMD’s first high-end Vega chip since the original Vega 10 in 2017, there are a few questions surrounding certain promising Vega hardware features. Information from AMD was not always clear, resulting in these features becoming a constant source of confusion for consumers. With Vega 20, AMD doesn’t seem to be changing the situation. That is, there was no further development by AMD regarding primitive shaders. Primitive shaders still have to be explicitly defined by developers, which means that the full potential of Next Generation Geometry (NGG) Vega 20 is not used (although it should be noted that the Intelligent Workgroup Distributor option was always enabled). Meanwhile, AMD Vegas 10 already used the Draw Stream Binning Rasterizer (DSBR), and there were no changes here; the feature is enabled for an unspecified list of games. According to a quick check on Rapid Packed Math (rapid FP16), it is used for two famous games: Far Cry 5 and Wolfenstein II. FP64 performance and difference between Radeon VII and Radeon Instinct MI50 One of the interesting and fun implications of the Radeon VII launch is that for the first time in a long time, AMD has had to seriously think about how they’re going to differentiate their consumer products from workstation and server products. AMD now offers professional hardware in the Radeon Pro and Radeon Instinct series, with the Vega 20 GPU being AMD’s first truly server GPU. Thus, while professional and consumer products are largely differentiated in terms of software features, Radeon VII provides some new options that are not really necessary for consumers. This may seem like a trivial question — it would seem AMD would be better off just leaving all features enabled. But as the company attempts to move into the higher-margin server business, consumer products like the Radeon VII are actually a tough proposition. AMD needs to lock down enough of the server functionality of the Vega 20 GPU to not sell the consumer equivalent of the Radeon Instinct MI50 at a fraction of the price. On the other hand, it’s in their interest to show off some of these features to make the Radeon VII a really valuable card (which could justify the $69 price tag).$9), as well as give developers an idea of ​​what level of server hardware AMD can do. In this case, we are talking about the performance of the FP64. As we noted in our Vega 20 GPU review, the FP64 performance in the Vega 20 is very high: it’s half the speed of the FP32, or 6. 9 TFLOPS. This is one of the premium features of the Vega 20, and ever since the Radeon VII was first announced at CES, the company has struggled to decide how much of that performance to actually make available to the Radeon VII. At the time of its announcement, we were told that the Radeon VII would have unlimited (1/2) FP64 performance, but was later told it would be 1/8. Now, with the actual release of the card, AMD has finally made up their mind: they’ve taken the average and reported 1/4 speed. In an effort to clarify the situation, AMD issued a statement: The Radeon VII was designed for gamers and creators, enthusiasts and trailblazers. Given the broader market that the Radeon VII is targeting, we looked at various levels of FP64 performance. We previously reported that the Radeon VII delivers 0.88 TFLOPS (DP = 1/16 SP). However, based on customer interest and feedback, we would like to announce that we have decided to increase double precision performance to 3. 52 3.46 TFLOPS (DP = 1/4SP). If you paid attention to FP64 performance in your testing, you may have seen this performance boost since the VBIOS drivers and the drivers we shared with reviewers were pre-test drivers that had these values ​​already set. In addition, we’ve updated other values ​​to reflect the achievable peak frequency when calculating Radeon VII performance, as noted in [charts]. The end result is that while the Radeon VII won’t be as fast as the MI60/MI50 when it comes to FP64 computing, AMD delivers a card that’s only a step behind the leaders. At 3.5 TLFLOPS of FP64 theoretical performance, the Radeon VII is in the right price bracket. There are simply no other current-gen boards under $2,000 that attempt this task. All NVIDIA GeForce cards and all other AMD Radeon cards lack the required hardware for fast FP64. The next closest competitor to the Radeon VII in this regard is the NVIDIA Titan V, which is more than 4 times more expensive. As we can see, this is a bit of a niche market, especially when there is so much industry focus on AI and neural network performance. But, nevertheless, some very happy Data Scientist, especially academic ones, will become the owners of the card. Speaking of AI, machine learning performance is another area where AMD has cut the card. Unfortunately, more detailed information is not available at this time. But given the unique needs of the ML market, I wouldn’t be surprised to find that INT8/INT4 performance is slightly reduced on the Radeon VII. Or the performance of some FP16 dot products has been reduced. Full-chip support for ECC also went under the knife. Due to the inherent functionality of HBM2, all Vega cards already have ECC for their DRAMs. However, the Vega 20 was a step ahead with ECC protection for its internal caches, a feature the Radeon VII does not have access to. Finally, the Radeon VII also reduces the I/O capabilities outside of the Vega 20 chip a bit. While AMD hasn’t made a lot of noise about it so far, Vega 20 is actually their first PCI-Express 4.0-capable GPU, and this feature is enabled. on Radeon Instinct cards. However, for Radeon VII this is not enabled and the card is limited to PCIe 3.0 speeds (so future Zen 2 buyers will not have a PCIe 4.0 card to fully work with their new processor). Similarly, Infinity Fabric external links for multi-GPU support have been disabled, and the Radeon VII will only perform solo. Overall, AMD’s choice is not surprising, especially given the Radeon VII’s target market and price point. But these noticeable “cuts” will make a big difference for certain users. And if they don’t push those users towards Radeon Instinct, they will surely lead to the inevitable Radeon Pro on Vega 20. Meet the Radeon VII First things first, design and build. And for the AMD Radeon VII, we’ve already noticed the biggest change: the exposed air cooler. Keeping the matte metallic exterior of the previous RX Vega 64 Limited Edition and Liquid, they ditched the air duct in favor of three fans, AIB’s standard custom configuration for high-end cards. While NVIDIA’s GeForce RTX series went the route of using coolers with two open fans, AMD is accustomed to changing the situation at will. Apart from the RX Vega 64 Liquid, the AIO CLC R9 Fury X also had an impressive design. But as we mentioned with Founders Edition cards, moving away from venting in favor of open coolers means adopting a heat dissipation configuration that can no longer guarantee complete self-cooling. That is, the cooling efficiency is now dependent on the air flow in the chassis. This is usually a problem for large OEMs who configure machines using duct type cards, but this does not apply to the highest level cards that usually come from system integrators for pre-build systems. Switching to open coolers benefits higher TDP, and at 300W TBP Radeon VII is indeed one of the leaders in power consumption. It seemed like only 5 watts more than the RX Vega 64. But maybe the extra heat is generated by the two extra stacks of HBM2, plus the problem is that the same amount of power is consumed on a smaller die area. And the 300W TBP means that all the energy saved from the smaller process has been reinvested in performance. If the higher clock speeds of the Radeon VII are the bulk of the performance gain over the RX Vega 64, then the air duct may not be able to handle the load. Going back to the Radeon VII build, the card naturally has two 8-pin PCIe connectors, but doesn’t have a BIOS switch like the RX Vega to switch the BIOS to low power. And the usual LEDs display the inscription «Radeon» on the side, and the «R» icon in the corner. As for display output, there are no surprises here: 3x DisplayPort and 1x HDMI. A little disassembly of the video card revealed a cooling configuration with flat heat pipes for TIM, instead of the usual paste. Despite poor performance in terms of heat transfer, we know that RX Vega cards ended up being supplied in both molded and unmolded versions, requiring special instructions from manufacturers. Perhaps this is a way to prevent potential ASIC height difference issues. Thank you for staying with us. Do you like our articles? Want to see more interesting content? Support us by placing an order or recommending to your friends, 30% discount for Habr users on a unique analogue of entry-level servers that we created for you: The whole truth about VPS (KVM) E5-2650 v4 (6 Cores) 10GB DDR4 240GB SSD 1Gbps from $20 or how to share a server? (available with RAID1 and RAID10, up to 24 cores and up to 40GB DDR4). VPS (KVM) E5-2650 v4 (6 Cores) 10GB DDR4 240GB SSD 1Gbps until summer free when paying for a period of six months, you can order here. Dell R730xd 2 times cheaper? Only here 2 x Intel Dodeca-Core Xeon E5-2650v4 128GB DDR4 6x480GB SSD 1Gbps 100 TV from $249 in the Netherlands and USA! Read about How to Build Infrastructure Corp. class with the use of Dell R730xd E5-2650 v4 servers worth 9,000 euros for a penny? Radeon VII vs Radeon RX Vega 56 Mobile Radeon VII vs Radeon RX Vega 56 Mobile — Th200 Contents Introduction Features Tests Games Key differences Conclusion Comments Video card Video card Introduction We compared two graphics cards: AMD Radeon VII vs. AMD Radeon RX Vega 56 Mobile. On this page, you will learn about the key differences between them, as well as which one is the best in terms of features and performance. The AMD Radeon VII is a Vega II (Radeon VII) generation GCN 5.1 graphics card released Feb 7th, 2019. It comes with 16GB of HBM2 memory running at 1000MHz, has a 2x 8-pin power connector and consumes up to 165 W. AMD Radeon RX Vega 56 Mobile is a Mobility Radeon (Vega) generation GCN 5.0 graphics card released on Jun 1st, 2018. It comes with 8GB HBM2 memory running at 800MHz and consumes up to 120W. 9Ol000 PCIe 3.0 X16 Graphics Processor Crystal size 331 mm² 495 Mm² memory 1.024 GB/S 9000.6 GB/S 582.8 GFLOPS (1:16) CLOCK Speeds Basic frequency 1400 MHZ 1138 MHZ 1750 MHZ 1301 MHZ MGC frequency 9000.19000 800 MGZ frequency 9000.19000 $ 9000 3,0004 800 MGC Render Config SHADING Units 3840 3584 Texture Units 240 224 Raster Units 64 64 Compte UNITS 003 60 56 Graphics Features DirectX 12 (12_1) 12 (12_1) OpenGL 4. 6 4.6 OpenCL 2.1 2.1 Vulkan 1.2 1.2 Board Design Heating 295W 120W Power Conventors 2x 8-Pin None Width 9000 Dual-Slot MXM Module Benchmarks 3DMark Graphics 3DMark is a benchmarking tool designed and developed by UL to measure the performance of computer hardware. Upon completion, the program gives a score, where a higher value indicates better performance. AMD Radeon VII AMD Radeon RX Vega 56 Mobile Blender bmw27 Blender is the most popular 3D content creation software. It has its own test, which is widely used to determine the rendering speed of processors and video cards. We chose the bmw27 scene. The result of the test is the time taken to render the given scene. AMD Radeon VII AMD Radeon RX Vega 56 Mobile Th200 RP Th200 RP is a test created by Th200. It measures the raw power of the components and gives a score, with a higher value indicating better performance. AMD Radeon VII +65% AMD Radeon RX Vega 56 Mobile Games 1920×1080, Ultra Game Radeon VII Radeon RX Vega 56 Mobile Battlefield V Cyberpunk 2077 DOOM Eternal Far Cry 5 Hitman 3 Metro Exodus Red Dead Redemption 2 Shadow of the Tomb Raider The Witcher 3 Average 109. 74 fps 0.00 fps 2560×1440, Ultra Game Radeon VII Radeon RX Vega 56 Mobile Assassin’s Creed Valhalla Battlefield V Cyberpunk 2077 DOOM Eternal Far Cry 5 Hitman 3 Metro Exodus Red Dead Redemption 2 Shadow of the Tomb Raider The Witcher 3 Average 87. 97 fps 0.00 fps 3840×2160, Ultra Game Radeon VII Radeon RX Vega 56 Mobile Assassin’s Creed Valhalla Battlefield V Cyberpunk 2077 DOOM Eternal Far Cry 5 Hitman 3 Metro Exodus Red Dead Redemption 2 Shadow of the Tomb Raider Average 51. 78 fps 0.00 fps Key differences Why is AMD Radeon VII better than AMD Radeon RX Vega 56 Mobile? Newer — 8 months later release Has 65% better performance More modern manufacturing process — 7 nm vs 14 nm Has 8 GB more memory Has a slightly larger memory bus Has 150% more bandwidth — 1,024 GB/s vs 409.6 GB/s Has 35% faster pixel fill rate — 112.0 GPixel/ s vs 83.26 GPixel/s Has 45% higher texture fill rate — 420.0 GTexel/s vs 291.4 GTexel/s Has more texture units +16 Has 256 more shading units Has 4 more compute units Why is AMD Radeon RX Vega 56 Mobile better than AMD Radeon VII? Uses 60% less power — 120W vs 295W Conclusion Which is better Radeon VII or Radeon RX Vega 56 Mobile? The Radeon VII delivers 65% better performance, consumes up to 146% more energy and holds 8 GB more memory.