AMD Radeon RX 5700 XT vs. Nvidia GeForce RTX 2060 Super: $400 GPU Throwdown
(Image credit: Tom’s Hardware)
Two of the top contenders for the best graphics card right now are AMD’s Radeon RX 5700 XT and Nvidia’s GeForce RTX 2060 Super. They’re not the fastest GPUs around—or at least, Nvidia’s GPU isn’t the company’s fastest option, as the RTX 2070 through RTX 2080 Ti are all faster—but they’re two of the best overall values. Both cards are available for around $400, representing the lower portion of the high-end GPU market, and we’re here to declare a winner.
Both the RTX 2060 Super and RX 5700 XT were introduced in the summer of 2019. AMD’s card features a Navi 10 GPU with a new RDNA architecture, while Nvidia’s card is effectively a refresh of 2018’s RTX 2070—the same TU106 GPU with slightly fewer CUDA cores enabled, but with a lower price that makes it the better bargain. $400 is an important price point as well, as going higher up the performance ladder typically brings diminishing returns. You can get great performance at 1440p with high settings, or 1080p ultra, without blowing up your build budget. Similar previous-generation cards like the GTX 970, GTX 1070, R9 390 and RX Vega 56 have all sold well.
You can also step down one rung and look at the Radeon RX 5700 and GeForce RTX 2060, or the Radeon RX 5600 XT if you’re going to stick to 1080p ultra or 1440p medium settings. As for stepping up, the only faster card from AMD right now is the Radeon VII, which is expensive, power hungry and not actually much faster than the RX 5700 XT. If you want something more potent, you’ll need to either wait for AMD to release its next-gen Navi 20/21 GPUs — which are supposed to include hardware ray tracing functionality as well — or look at Nvidia’s more expensive RTX models. Right now AMD has nothing that can beat the RTX 2070 Super and above.
But let’s get back on topic. Which card is actually better, the RTX 2060 Super or the RX 5700 XT? We’ve put the reference models from both companies against each other in this head-to-heat battle. We’ll look at features and specifications, drivers and software, performance in games, power and thermals, and finally overall value. Cue the high noon music as the rivals prepare to draw.
Featured Technology
AMD’s Radeon RX 5700 XT is the current halo product for the company, using the fully-enabled Navi 10 XT GPU and sporting 2560 processing cores. The reference model has a game clock of 1755 MHz, with a maximum boost clock of 1905 MHz, and factory overclocked cards typically increase the game clock to around 1815 MHz (give or take). Nvidia’s GeForce RTX 2060 Super uses the TU106-410-A1 GPU, with 2176 CUDA cores and a conservative boost clock of 1650 MHz. It also includes 34 RT cores for ray tracing calculations and 272 Tensor cores for deep learning computations (ie, DLSS).
1. Hardware Specifications
Swipe to scroll horizontally
RX 5700 XT | RTX 2060 Super | ||
---|---|---|---|
Architecture | Navi 10 | TU106 | |
Process (nm) | 7 | 12 | |
Transistors (Billion) | 10. 2) | 251 | 445 |
SMMs / CUs | 40 | 34 | |
GPU Cores | 2560 | 2176 | |
Tensor Cores | N/A | 272 | |
RT Cores | N/A | 34 | |
Turbo Clock (MHz) | 1755 | 1650 | |
VRAM Speed (Gbps) | 14 | 14 | |
VRAM (GB) | 8 | 8 | |
VRAM Bus Width | 256 | 256 | |
ROPs | 64 | 64 | |
TMUs | 160 | 136 | |
GFLOPS (Boost) | 8986 | 7181 | |
Bandwidth (GBps) | 448 | 448 | |
TDP (watts) | 225 | 175 | |
Launch Date | Jul-19 | Jul-19 |
Comparing theoretical performance based on paper specs is typically a fool’s errand, as architectural details come into play. Still, on paper the RX 5700 XT has 9.0 TFLOPS and 448 GBps of bandwidth going up against the RTX 2060 Super’s 7.2 TFLOPS and 448 GBps. That’s a 25 percent advantage for AMD, but (spoiler!) it almost never shows up to that level in games.
AMD’s RX 5700 XT is also the first ‘real’ 7nm GPU—we don’t really count the Radeon VII, since it was late to the party, based around an old architecture, and overall couldn’t match Nvidia’s existing cards. TSMC’s 7nm finFET process allows AMD to stuff approximately 10.3 billion transistors into the Navi 10 die that measures just 251 mm2.
In contrast, the RTX 2060 Super uses TSMC’s older 12nm finFET process. It only has a ‘few’ more transistors,10.8 billion (that’s a half billion, but not a big difference percentage-wise), but the die size is 445 mm2. That’s a 77% larger die, again for only 5% more transistors—not that die size or process technology matter to the end user, but they can affect power use, cost and even performance.
On the memory side of things, both GPUs feature 8GB of GDDR6 RAM clocked at 14 Gbps on a 256-bit bus, giving them the same bandwidth. That doesn’t mean the memory subsystems of the GPUs are the same, however. Both companies use various forms of lossless delta color compression to extract more performance from the memory. Details are somewhat limited, though Nvidia’s GPUs seem to either compress data better or simply extract more performance from the same memory bandwidth in other ways.
The bigger difference between the GPUs is in support for rendering features, specifically DirectX Raytracing (DXR) and Vulkan-RT. How important is ray tracing in games? Right now, there are only a small number of supported titles, and the ray tracing effects are often relatively small.
Here’s the complete list of ray tracing games that were available when we wrote this: Battlefield V, Call of Duty: Modern Warfare, Control, Deliver Us the Moon, Metro Exodus, Quake II RTX, Shadow of the Tomb Raider and Wolfenstein: Youngblood. That’s eight games, and another eight should launch in 2020 (Cyberpunk 2077, Doom Eternal, Dying Light 2, MechWarrior 5: Mercenaries, Minecraft RTX, Synced: Off-Planet, Vampire: The Masquerade — Bloodlines 2 and Watch Dogs: Legion). The problem is that so far, we’d only rate the ray tracing effects as truly desirable in Control and maybe Deliver Us the Moon—and even then the performance hit can be large.
That’s why Nvidia also has deep learning super-sample (DLSS) technology, which uses machine learning and massive amounts of training on Nvidia’s supercomputers to allow games to render at a lower resolution and then upscale with anti-aliasing detection and removal. DLSS works okay but it’s not perfect. It’s also very useful if you’re not using the RTX 2080 Ti and playing titles with ray tracing at 1080p.
The two competitors also offer adaptive refresh rate technology: AMD FreeSync and Nvidia G-Sync. Nvidia has an advantage now, as it has started to support select ‘compatible’ adaptive sync displays (aka, FreeSync but without the AMD branding), whereas AMD only works with FreeSync/Adaptive Sync monitors. Nvidia says it has tested hundreds of FreeSync displays, 90 of which are now ‘G-Sync Compatible ‘—including LG’s latest OLED 120Hz TVs. Also, Nvidia has added HDMI 2.1 VRR (variable refresh rate) support to its RTX series cards.
Ultimately, despite AMD having an advantage in manufacturing technology, Nvidia’s hardware features win out. You may not love the way ray tracing effects look in every game, or the potential performance impact, but at least you have the option to turn them on and try them out. Plus you have the choice between G-Sync and FreeSync / G-Sync Compatible displays.
Winner: Nvidia
Image 1 of 3
(Image credit: Tom’s Hardware)(Image credit: Tom’s Hardware)(Image credit: Tom’s Hardware)
2. Drivers and Software
Both AMD and Nvidia offer full suites of drivers and software. To put things in perspective, the driver download size currently tips the scale at 529 MB for Nvidia and 482 MB for AMD. That’s nearly as large as the entirety of Windows XP, just for your GPU drivers and associated software.
AMD’s latest Radeon Adrenalin 2020 drivers added a bunch of new features: integer scaling, Game Boost, and a completely revamped UI. AMD claims performance is «up to 12% better» relative to its Adrenalin 2019 drivers, though having tested both it’s usually not that big of a jump. There are loads of pre-existing features as well, like Radeon Anti-Lag, Radeon Image Sharpening, Radeon Chill, Enhanced Sync, and the ability to record or stream your gaming sessions. It’s almost too much to dig through, but AMD basically gives you everything you need in one package, including the ability to tweak and overclock your graphics card.
The latest Nvidia 442.19 drivers don’t have any major new highlights to discuss, but like AMD the company has been adding features and enhancements for years. Integer scaling is available (on Turing GPUs), and you can adjust 3D settings and video enhancements. But a lot of the extras are within GeForce Experience. That’s where you’ll find ShadowPlay for streaming and recording your gameplay, hear about driver updates, get recommended settings profiles for most major games, and there’s the in-game overlay that allows you to apply filters—including sharpening, select ReShade filters, and more.
There’s also no overclocking support built into Nvidia’s drivers, but it’s not a major loss. Tools like MSI Afterburner and EVGA Precision X1 are readily available for those that want to tinker, and are arguably better than anything AMD or Nvidia are likely to build into their drivers.
We prefer AMD’s unified approach, but there’s something else we like about AMD’s drivers. Generally speaking, the software and drivers discussion ends up being nitpicky. However, there’s one thing that I’ve come to despise about GeForce Experience: The required login and captcha to make things work. We upgrade GPU drivers more frequently than most users, and with a constant stream of new GPUs to test we’re more aware of login requirements and captchas than most. Still, it’s extremely annoying how frequently GeForce Experience seems to forget login credentials. Please, just make all of that go away, or at least make it optional. We’ve been data mined enough, thanks.
Frequency and quality of driver updates is a bit harder to quantify. Both companies put out new drivers on a regular basis—AMD had 33 releases in 2019 by our count, and Nvidia had 26 releases (not including non-public drivers). That’s 2-3 driver updates per month on average, which some would argue is too many. Both companies also delivered ‘game ready’ drivers in a timely fashion for all of the major launches, and both had a few hiccups along the way.
Winner: AMD, because logging in to use drivers is a terrible idea
3. Gaming Performance
Unless you’re folding protein strands, searching for ET, building an AI to take over the world, mining cryptocurrency or some other work, chances are good that you’re looking at a graphics card upgrade to play games. We’ve tested 12 games across multiple resolutions and settings to determine which of these two GPU ends up being the fastest. Here are the results:
Average / Minimum FPS
Swipe to scroll horizontally
Game | Setting | RX 5700 XT | RTX 2060 Super | RX 5700 XT vs. RTX 2060 Super |
---|---|---|---|---|
12 Game Average | 1080p Medium | 161.1 / 115.3 | 147.7 / 104.8 | 9% / 10% |
1080p Ultra | 111.1 / 79.9 | 102.1 / 73.5 | 8.8% / 8.8% | |
1440p Ultra | 80.6 / 60.9 | 74.5 / 57.2 | 8.2% / 6.4% | |
4k Ultra | 45.4 / 35.5 | 42.3 / 33.7 | 7.1% / 5.3% | |
Assassin’s Creed Odyssey | 1080p Medium | 101.3 / 77.4 | 105.7 / 77.9 | -4.2% / -0.6% |
1080p Ultra | 73.3 / 55.8 | 66.2 / 50.4 | 10.7% / 10.7% | |
1440p Ultra | 56. 2 / 44.0 | 53.7 / 41.8 | 4.7% / 5.3% | |
4k Ultra | 36.2 / 29.1 | 35.9 / 30.4 | 0.8% / -4.3% | |
Borderlands 3 | 1080p Medium | 168.1 / 121.2 | 138.2 / 97.5 | 21.6% / 24.3% |
1080p Ultra | 95.3 / 74.8 | 77.4 / 62.9 | 23.1% / 18.9% | |
1440p Ultra | 67.1 / 55.6 | 55.0 / 46.4 | 22% / 19.8% | |
4k Ultra | 36.1 / 30.7 | 30.2 / 26.3 | 19.5% / 16.7% | |
The Division 2 | 1080p Medium | 199.6 / 152.5 | 164.5 / 127.0 | 21.3% / 20.1% |
1080p Ultra | 104. 4 / 83.2 | 94.8 / 75.5 | 10.1% / 10.2% | |
1440p Ultra | 68.3 / 57.6 | 63.8 / 53.5 | 7.1% / 7.7% | |
4k Ultra | 35.3 / 30.6 | 33.4 / 29.4 | 5.7% / 4.1% | |
Far Cry 5 | 1080p Medium | 152.5 / 121.0 | 141.5 / 104.1 | 7.8% / 16.2% |
1080p Epic | 133.8 / 110.1 | 124.6 / 87.6 | 7.4% / 25.7% | |
1440p Epic | 100.3 / 85.1 | 88.8 / 73.5 | 13% / 15.8% | |
4k Epic | 51.8 / 44.7 | 47.1 / 40.8 | 10% / 9.6% | |
Forza Horizon 4 | 1080p Medium | 218. 6 / 172.6 | 181.4 / 143.2 | 20.5% / 20.5% |
1080p Ultra | 166.5 / 141.5 | 137.5 / 116.3 | 21.1% / 21.7% | |
1440p Ultra | 134.1 / 116.5 | 111.4 / 94.8 | 20.4% / 22.9% | |
4k Ultra | 84.3 / 72.7 | 73.6 / 63.6 | 14.5% / 14.3% | |
Hitman 2 | 1080p ‘Medium’ | 162.1 / 101.2 | 144.3 / 90.4 | 12.3% / 11.9% |
1080p ‘Max’ | 121.6 / 68.9 | 115.5 / 65.2 | 5.3% / 5.7% | |
1440p ‘Max’ | 85.3 / 50.9 | 79.3 / 49.5 | 7.6% / 2.8% | |
4k ‘Max’ | 44. 9 / 28.3 | 41.2 / 25.6 | 9% / 10.5% | |
Metro Exodus | 1080p Normal | 114.1 / 56.8 | 98.8 / 50.7 | 15.5% / 12% |
1080p High | 81.8 / 43.0 | 74.5 / 40.1 | 9.8% / 7.2% | |
1440p High | 63.6 / 36.8 | 56.2 / 34.0 | 13.2% / 8.2% | |
4k High | 39.4 / 25.0 | 34.7 / 22.4 | 13.5% / 11.6% | |
The Outer Worlds | 1080p Medium | 155.8 / 93.6 | 166.3 / 102.1 | -6.3% / -8.3% |
1080p Ultra | 106.4 / 71.4 | 112.1 / 79.4 | -5.1% / -10.1% | |
1440p Ultra | 71. 0 / 50.5 | 76.0 / 58.0 | -6.6% / -12.9% | |
4k Ultra | 37.0 / 28.2 | 39.9 / 32.5 | -7.3% / -13.2% | |
Red Dead Redemption 2 | 1080p Low | 159.8 / 117.9 | 145.2 / 107.6 | 10.1% / 9.6% |
1080p High | 106.2 / 80.0 | 96.8 / 73.4 | 9.7% / 9% | |
1440p High | 80.5 / 60.4 | 72.3 / 55.1 | 11.3% / 9.6% | |
4k High | 47.9 / 37.3 | 41.1 / 29.8 | 16.5% / 25.2% | |
Shadow of the Tomb Raider | 1080p Medium | 153.1 / 120.3 | 134.7 / 101.6 | 13.7% / 18.4% |
1080p Highest | 125. 9 / 97.0 | 117.7 / 89.6 | 7% / 8.3% | |
1440p Highest | 82.5 / 66.8 | 78.9 / 63.2 | 4.6% / 5.7% | |
4k Highest | 42.1 / 34.1 | 41.5 / 35.1 | 1.4% / -2.8% | |
Strange Brigade | 1080p Medium | 222.9 / 189.6 | 216.2 / 190.0 | 3.1% / -0.2% |
1080p Ultra | 162.1 / 119.7 | 142.7 / 110.0 | 13.6% / 8.8% | |
1440p Ultra | 116.5 / 90.7 | 104.9 / 85.4 | 11.1% / 6.2% | |
4k Ultra | 69.4 / 59.7 | 60.4 / 52.5 | 14.9% / 13.7% | |
Total War: Warhammer 2 | 1080p Medium | 171. 3 / 127.7 | 174.8 / 126.2 | -2% / 1.2% |
1080p Ultra | 94.5 / 63.8 | 98.3 / 67.5 | -3.9% / -5.5% | |
1440p Ultra | 73.3 / 53.5 | 77.7 / 58.6 | -5.7% / -8.7% | |
4k Ultra | 41.1 / 30.1 | 44.9 / 34.8 | -8.5% / -13.5% |
AMD’s RX 5700 XT just about sweeps the results, coming out 7-10% faster overall at the tested resolutions. The only games that favored the Nvidia RTX 2060 Super are The Outer Worlds and Total War: Warhammer 2—both of which are, oddly enough, AMD promotional titles. Both games also support the DirectX 11 API (we tested DX12 in Warhammer 2 on AMD GPUs, DX11 on Nvidia), where eight of the ten remaining games were tested with either the DX12 or Vulkan API. We also didn’t test with ray tracing enabled in the two games that support it, since AMD has so far elected not to add driver support for DXR.
This isn’t an insurmountable lead, and factory overclocked cards could potentially narrow the gap, but both AMD and Nvidia partner cards are typically overclocked to similar levels so that shouldn’t shake up the standings much. The good news is that either GPU should easily handle 1080p ultra gaming at 60 fps, and 60 fps is even doable at 1440p ultra on many of the games—Assassin’s Creed Odyssey, Borderlands 3 and Metro Exodus being three where one or both cards come up short.
Winner: AMD
4. Power Consumption and Heat Output
These are high-end GPUs and depending on the specific model, PCI Express Graphics (PEG) power connector requirements will vary. Most cards will require both 8-pin and 6-pin PEG connectors. Some factory overclocked models might include dual 8-pin PEG connectors, though almost certainly not because it’s strictly necessary. Only the RTX 2060 Super Founders Edition uses a single 8-pin PEG connector.
AMD’s RX 5700 XT has an official Typical Board Power (TBP) rating of 225W, while Nvidia’s RTX 2060 Super is rated at 175W. Those are both for reference designs, however, and factory overclocked partner cards are very likely to be a bit higher. AMD recommends a minimum 600W power supply, and Nvidia recommends at least a 550W PSU.
In actual use, power requirements with a Core i9-9900K test system are far lower than those PSU recommendations. We measured around 325W at the outlet with the RTX 2060 Super, and 370W with the RX 5700 XT. We’ll be refining our power testing with specialized hardware in the near future, as reading power use via software utilities depends on the drivers, but AMD and Nvidia don’t report the same metrics. AMD gives GPU-only power use, while Nvidia gives total board power use. GPU-Z as an example showed average power draws of 173.2W for the 2060 Super vs. 179.0W for the 5700 XT—even though outlet power use was 48W higher on the AMD card.
For most PCs, the 30-50W difference in power use isn’t going to matter a lot—if your PSU can handle the 2060 Super, it should be able to handle the 5700 XT as well. However, higher power use leads directly to more heat output, which potentially means more fan noise. It’s possible for custom cards to use larger coolers and better fans to get around that, but in general the Nvidia RTX 2060 Super will use less power, generate less heat and keep noise levels down.
The fact that Nvidia offers similar (though slightly lower) performance while using the older 12nm manufacturing node is impressive, and we can only look forward to seeing what Nvidia’s 7nm Ampere GPUs do later this year.
Winner: Nvidia
5. Value Proposition
Determining which card is the better value is tricky, since all of the above elements are factors. Both cards have a recommended price of $400. Right now, the least expensive RTX 2060 Super cards cost $400, though there’s an EVGA RTX 2060 Super SC Ultra Gaming at Newegg that drops to $380 after rebates if you’re willing to deal with those. There are rebates on AMD’s RX 5700 XT as well, but the least expensive model is the ASRock RX 5700 XT Challenger D 8G OC , which is currently on sale for $360 without any mail-in rebate hoops to jump through.
Individual GPU sales may come and go, but for the past several months at least, the RX 5700 XT has been selling for less than the RTX 2060 Super. AMD also sweetens the deal by offering some free games with RX 5700 XT cards (you’ll want to verify the deal is available from the reseller before buying, however). AMD’s current Raise the Game bundle will get you Monster Hunter World: Iceborne, Resident Evil 3 and three months of Xbox Game Pass for PC with the RX 5700 XT. How much those games are worth is debatable, but it’s obviously a much better deal than getting nothing extra.
Winner: AMD
Swipe to scroll horizontally
Round | Nvidia GeForce RTX 2060 Super | AMD Radeon RX 5700 XT |
---|---|---|
Featured Technology | ✗ | |
Drivers and Software | ✗ | |
Gaming Performance | ✗ | |
Power Consumption | ✗ | |
Value Proposition | ✗ | |
Total | 2 | 3 |
Lower prices go a long way toward overcoming any potential feature and power advantages Nvidia might have. If ray tracing in games truly catches on—we want more games that look like Control rather than Battlefield V or Shadow of the Tomb Raider, for example—the RTX 2060 Super could be a better long-term purchase. Will it still be fast enough to handle all the extra effects, or will you need to upgrade to a 3060 or some other future card? We can’t say for certain. Regardless, we’ve been hearing about all the ‘amazing’ ray tracing enabled games coming soon for more than a year now, and most have failed to live up to the hype.
Our advice is to get a card that runs current games better, and save ray tracing for another day. AMD’s RX 5700 XT costs less than the RTX 2060 Super, comes with a gaming bundle as an added bonus, has good driver and software support, and outperforms the competition in ten of the 12 games we tested. If you want the best GPU right now for $400 or less, AMD’s RX 5700 XT, Navi 10 and the RDNA architecture are the winner.
If you’re willing to spend more money, AMD falls behind the RTX 2070 Super and beyond. Alternatively, you can still find a few RTX 2070 cards at clearance pricing, like this Gigabyte RTX 2070 Windforce 8G (also available at B&H and Newegg for the same $400 asking price). The RTX 2070 ends up being 5% faster than the RTX 2060 Super, making performance mostly a tie. It still costs more, but ray tracing use in games should pick up with both the PlayStation 5 and Xbox Series X coming later this year and supporting the feature.
Pricing can also change, and Nvidia dropped the RTX 2060 starting price to $300 when AMD launched the RX 5600 XT in January 2020. The RTX 2060 Super prices didn’t drop, but if they do, Nvidia could reclaim the $350 crown. What’s more likely is that Nvidia’s next-gen Ampere GPUs will shake things up in the graphics card market when they launch later this year.
Overall Winner: AMD
Get instant access to breaking news, in-depth reviews and helpful tips.
Contact me with news and offers from other Future brandsReceive email from us on behalf of our trusted partners or sponsors
Jarred Walton is a senior editor at Tom’s Hardware focusing on everything GPU. He has been working as a tech journalist since 2004, writing for AnandTech, Maximum PC, and PC Gamer. From the first S3 Virge ‘3D decelerators’ to today’s GPUs, Jarred keeps up with all the latest graphics trends and is the one to ask about game performance.
AMD RX 6700 XT vs RX 5700 XT Benchmark [Updated 2023 ]
AMD RX 6700 XT vs RX 5700 XT — General Specs
Sale
26,929 Reviews
AMD Ryzen 7 3700X 8-Core, 16-Thread…
- The world’s most advanced processor in the desktop PC gaming segment
- Can deliver ultra-fast 100 plus FPS performance in the world’s most…
- 8 Cores and 16 processing threads bundled with the AMD Wraith Prism…
AMD RX 6700 XT vs RX 5700 XT — Speed & Performance
Clock Speeds (Base & Boost)
Sale
372 Reviews
GIGABYTE B550 AORUS Master (AM4. ..
- Supports 3rd Gen AMD Ryzen Processors
- Dual Channel ECC/ Non-ECC Unbuffered DDR4, 4 DIMMs
- Direct 16 Phases Digital VRM Solution with 70A Power Stage
Cores (Shader, Tensor & RT)
36,689 Reviews
SAMSUNG (MZ-V7E1T0BW) 970 EVO SSD 1TB -…
- INNOVATIVE V NAND TECHNOLOGY: Powered by Samsung V NAND Technology,…
- BREAKTHROUGH READ WRITE SPEEDS: Sequential read and write performance…
- PERFORMANCE OPTIMIZATION AND DATA SECURITY: Seamless cloning and file…
Raytracing & AI Performance
38,086 Reviews
Corsair Vengeance RGB Pro 32GB (2x16GB)…
- High performance DDR4 memory illuminates your system with vivid,…
- Take control with CORSAIR iCUE software and synchronize lighting…
- Customize lighting profiles with millions of different patterns and. ..
Rasterization (ROPs/TMUs/CU/Texture Rate/Pixel Rate)
Sale
479 Reviews
ASUS ROG Strix 27″ 4K HDR DSC Gaming…
- 27-inch 4K (3840×2160) IPS gaming monitor with 144Hz refresh rate and…
- Supports Display Stream Compression (DSC) Technology for transporting…
- High dynamic range (HDR) technology with DisplayHDR 400 certification…
Theoretical Performance (Integer & Floating Point)
Sale
13,565 Reviews
Corsair Void RGB Elite Wireless Premium…
- Noise Control: Noise Cancellation.Headphones fit type:On-Ear
- Hear everything from the lightest footstep to the most. Thundering…
- Constructed for enduring comfort through long gaming sessions with…
- RTX 3080 has 128KB of L1 Cache, which is 64KB (100%) more than the RTX 2080 Ti’s 64KB L1 Cache.
- RTX 3080 has 5MB of L2 Cache, which is 0.5MB (9%) less than the RTX 2080 Ti’s 5.5MB L2 Cache.
Sale
5,085 Reviews
Corsair Nightsword RGB — Comfort…
- Corsair exclusive software automatically detects the center of gravity…
- A custom PixArt PMW3391 native 18, 000 DPI optical sensor, adjustable…
- Built with a comfortable contoured shape that naturally fits your…
AMD Radeon Pro 5700 XT vs NVIDIA GeForce RTX 2080 Ti comparison which is better?
Home / AMD Radeon Pro 5700 XT VS NVIDIA GEFORCE RTX 2080 TI
AMD Radeon Pro 5700 XT
56%
Devicelist
VS
NVIDIA GeForce RTIS 9000 85000 9000 85000 85000 85000 9000 85000 85000 85000 85000 9000 85000 85000 9000
We compared the specifications of AMD Radeon Pro 5700 XT and NVIDIA GeForce RTX 2080 Ti and compiled a list of benefits and comparison table for you. Find out which one to choose in 2023. nine0003
AMD Radeon Pro 5700 XT Benefits
Process |
7 nm -5 nm (-41.7%) better than vs 12 nm |
Power Demand (TDP) |
130 W -120 W (-48%) better than vs 250 W |
Maximum memory |
16 GB 5 GB (45.5%) better than vs 11 GB |
Benefits NVIDIA GeForce RTX 2080 Ti
Comparison winner
Number of shader processors |
4352 1792 (70%) better than vs 2560 |
Boost frequency |
1545 MHz 46 MHz (3. 1%) better than vs 1499 MHz |
Memory bus width |
352 bit 96 bit (37.5%) better than vs 256 bit |
Memory frequency |
14000 MHz At 13988 MHz (116566.7%) better than vs 12 MHz |
Memory bandwidth |
616 232 (60.4%) better than vs 384 |
General information | |
Price-quality ratio The sum of all the advantages of the device divided by its price. The more%, the better the quality per unit price in comparison with all analogues. |
|
n/a | 32.4% |
Architecture |
|
RDNA 1. 0 | Turing |
Codename |
|
Navi 10 | TU102 |
Type |
|
For workstations | Desktop |
Price at the time of issue |
|
n/a | 999 $ |
Number of shaders |
|
2560 | 4352
1792 (70%) better than |
Core clock |
|
n/a | 1350 MHz |
Boost frequency |
|
1499 MHz | 1545 MHz
46 MHz (3.1%) better than |
Number of transistors |
|
10.300 million | 18.600 million |
Process |
|
7nm
-5 nm (-41.7%) better than nine0027 | 12 nm |
Interface |
|
PCIe 4. 0 x16 | PCIe 3.0 x16 |
Power Demand (TDP) Calculated thermal power shows the average heat dissipation in load operation, |
|
130 W nine0002 -120 W (-48%) better than |
250 W |
Length |
|
n/a | 267 mm |
Additional power connectors |
|
no | 2x 8-pin |
G-SYNC Ready NVIDIA G-SYNC technology delivers a smooth gaming experience with variable refresh rates and the elimination of visual artifacts. nine0003 |
|
Multi Monitor |
|
n/a | + |
VR Ready Technology from NVIDIA that gives manufacturers access to Multi res Shading, Context Priority, and GPU Direct virtual reality technologies. |
|
n/a | |
Vulkan NVIDIA’s Vulkan technology allows developers to gain low-level access to the GPU to optimize graphics commands (better than OpenGL and Direct3D APIs). |
|
1.2 | + |
OpenCL |
|
2.0 | |
Video connectors |
|
No outputs | 1x HDMI, 3x DisplayPort, 1x USB Type-C |
DirectX |
|
12 (12_1) | 12 Ultimate (12_1) |
Shader model |
|
6.5 | no data |
Benchmark |
Memory | |
Memory Type | |
GDDR6 | GDDR6 |
Maximum memory Large video memory allows you to run demanding games with lots of textures, |
|
16 GB
5 GB (45.5%) better than |
11 GB |
Memory bus width The larger the video memory bus width, the more data is transferred to the GPU per unit of time and the better performance in demanding games. |
|
256 bit | 352 bit
96 bit (37.5%) better than |
Shared memory |
|
n/a | — |
Memory frequency A high memory frequency has a positive effect on the speed of a video card with a large amount of data. |
|
12 MHz | 14000 MHz
At 13988 MHz (116566.7%) better than |
Memory bandwidth The higher the data transfer bandwidth, the more effective RAM the PC can use. |
|
384 | 616
232 (60.4%) better than |
RX 5700 and XT — Radeon RX 6900XT — Big Navi showed a slight lead over the RX 5700 XT and up to 25% behind the GeForce RTX 2080 Ti
odubo
Forum friend
-
-
#1
The outgoing week was generous with leaks and rumors. Insiders literally bombarded us with news about the next generation of NVIDIA graphics cards. You can learn everything in one place in our material. Apparently, next week will pass in the same vein. It is curious that among all this informational noise, some information about the graphics chip from AMD broke through. Experts have already taken a closer look at several recent leaks, which allowed them to form some impressions about the upcoming Radeon RX 6000 series models. In short, you should not wait for a real flagship. Big Navi will get slow memory and a small number of cores. But these are all unsubstantiated rumors. Let’s focus on the first tests, which speak better than any words about real performance. nine0592 The so-called Radeon RX 6900XT has been spotted on the Internet. In fact, it could be any other AMD card, but experts believe that we can have a flagship in front of us. Ashes of the Singularity, a popular game among fans of complex testing, was chosen as a testing ground. The results were presented by two popular insiders: Rogame and TUM_APISAK. In the first two images, you will see a comparison table showing the success of individual video cards.
nine0003
Well, we clearly did not expect such a result. In 4K resolution at maximum graphics settings, the new product is only 20% faster than the Radeon RX 5700 XT. Here it is worth considering that we have an overclocked model in front of us, but this does not change things, because we were clearly promised something different. Representatives of the red camp spoke about an uncompromising graphics chip that will be able to provide the highest settings in 4K resolution in any game.
No less sad to compare the engineering version of the Radeon RX 6900XT with GeForce RTX 2080 Ti. So, the new product is approximately equal to the reference model from NVIDIA, but loses an incredible 25% to the overclocked version. You can console yourself that we have an engineer in front of us. It is doubtful that AMD will be able to make it run faster in a month and a half. Although, time will tell.
Original article on Overs
it certainly can be a flagship, and the weakest card in the new line, let’s hope so, otherwise amudé with 6000 series will be a failure of the year…
vkushta
Experienced
-
-
#2
well, what did you want… For kefir, the bus and memory decides. The maximum that AMD could do was to slightly improve the core, which we, the miners, don’t care …
If cards were released on HBM, it would be another matter
nek29
Experienced
-
-
#3
Well, here from which side to look. in terms of hashes, the 2080ti lags behind the 5700xt. You don’t wrap our fur coat here)
postalman
Knowledgeable
-
-
#four
Where is it written that this is 6900xt? There is no and will not be information until the official announcement. Stop speculating.
Serzhsergio
Own person
nine0578
-
-
#five
how much money?
how many mega hashes?
how much electricity?
Jinpachi
Experienced
-
nine0586
-
#6
nek29 said:
Well, here from which side to look. in terms of hashes, the 2080ti lags behind the 5700xt. Don’t wrap our fur coat here)
Click to expand…
What hashes are behind? In my opinion, 2080ti is better on any algorithms, prada expensive infection. But this is due to positioning more…
nek29
Experienced
-
-
#7
Jinpachi said:
What hashes are behind? In my opinion, 2080ti is better on any algorithms, prada expensive infection. But this is due to positioning more…
Click to expand. ..
nine0729
Fuck your algorithms. Here the topic is for amd, respectively for kefir.
odubo
Forum friend
#8
postalman said:
Where is it written that this is 6900xt? There is no and will not be information until the official announcement. Stop speculating. nine0003
Click to expand…
Article from Overs! No one is speculating! My thoughts after the reference to overs ass!
Jinpachi
Experienced
#nine
nek29 said:
blaspheme us your algorithms. Here the topic is for amd, respectively for kefir.
Click to expand…
Don’t worry about your kefir having rigs from 2080ti. Or are you completely stupid? Do not you understand that mining does not live by kefir alone? )))
Builders80
Own person
#10
xform
Forum friend
#eleven
Another set of conjectures and fantasies. I don’t see the point in condemning fictitious «news». nine0003
In the name of blockchain, token and holy bitcoin, aisio!
Serg_VSA
Forum friend
#12
odubo said:
But these are all unproven rumors
Click to expand…
Katapu1t
Own person
#13
Jinpachi said:
Don’t worry about your kefir having rigs from 2080ti. Or are you completely stupid? Do not you understand that mining does not live by kefir alone? ))) nine0003
Click to expand…
Mining lives on kefir alone. Lest anyone say my FOUR year old women 474/478 have already brought so much that not one 2080 has ever been dreamed of (and to this day they bring no less than your new 1660 supers). I don’t even want to discuss the price difference. And whoever thinks that when the air goes to the settlement, the green ones will laugh and enjoy life, then they are deeply mistaken. Millions of red cards will simply drown all sorts of greens and crows in the trash and mining will end there. nine0003
xform
Forum friend
#fourteen
Jinpachi said:
Don’t worry about your kefir having rigs from 2080ti.
Click to expand…
How much pain in this scream…
In the name of blockchain, token and holy bitcoin, aisio!
Zephar
Experienced
#15
And why on this test is slightly higher AMD in 2080 and in 2080ti the results are almost the same?
nine0002
xform
Forum friend
#sixteen
Katapu1t said:
Millions of red cards will simply drown all sorts of greens and crows in the trash and mining will end there.
Click to expand…
I do not agree. Something else will just show up.
In the name of blockchain, token and holy bitcoin, aisio!
Katapu1t
Own person
#17
xform said:
I disagree. Something else will just show up.
Click to expand…
Not enough cards to fit. You yourself count. Divide the complexity of the ether network by 30mx. Look at how many cards there are. Calculate how much ala one 570/580 gives on green and crow. And throw at least 10% of AMD cards on a crow, you will see the complexity of the network and understand that mining is dead))
kemono
Forum friend
#eighteen
Yes lan to you =) AMD — Ethash, the rest is nVidia
Better tell me when to come to the memorial service for 4GB? At least I can stroke the suitKatapu1t
Own person
nine0578
#nineteen
kemono said:
Yes lan to you =) AMD — Ethash, the rest is nVidia
Better tell me when to come to the memorial service for 4GB? I can pet the suitClick to expand.