AMD Ryzen 5000 Series CPU Review Roundup
Always look at the date when you read an article. Some of the content in this article is most likely out of date, as it was written on November 5, 2020. For newer information, see our more recent articles.
Table of Contents
Introduction
On October 8th, AMD announced the new Ryzen 5000 Series processors with a range of advancements including a 19% improvement in IPC (instructions per clock) and unified 8-core complex among others. While the announcement was focused primarily on gaming performance, they did also promote performance improvements of 5% in Premiere Pro, 12% in V-Ray, 27% in SolidWorks, and 9% in GCC compile times.
To verify these performance claims and to see how they do in a range of other real-world scenarios, we have reviews looking at a number of applications across video and photo editing, game development, photogrammetry, and more. While this post includes a summary for each entry, we highly recommend reading the full article for programs you are interested in as we go into much greater detail in the individual reviews.
https://www.youtube.com/watch?v=aqNwaY13Zp4?t=18
Labs Open Office Hour: Labs Lads Discuss the New AMD Ryzen 5000-Series CPU Launch
If you want to see the full specs for the new AMD Ryzen 5000-series processors, we recommend checking out AMD’s page for these new CPUs. But at a glance, here are what we consider to be the most important specs:
| Cores/Threads | Max Boost Clock | TDP | MSRP | |
|---|---|---|---|---|
| Ryzen 5 3600XT | 6/12 | 4.5GHz | 95W | $249 |
| Ryzen 5 5600X | 6/12 | 4.5GHz | 65W | $299 |
| Ryzen 7 3800XT | 8/16 | 4.7GHz | 105W | $399 |
| Ryzen 7 5800X | 8/16 | 4.7GHz | 105W | $449 |
| Ryzen 9 3900XT | 12/24 | 4. 7GHz |
105W | $499 |
| Ryzen 9 5900X | 12/24 | 4.8GHz | 105W | $549 |
| Ryzen 9 3950X | 16/32 | 4.7GHz | 105W | $749 |
| Ryzen 9 5950X | 16/32 | 4.9GHz | 105W | $799 |
Something to keep in mind is that specs are only the tip of the iceberg in terms of actual performance. While the core counts and even max boost frequency haven’t changed much, the 19% IPC improvement AMD is advertising may automatically make each of the new CPUs up to 19% faster than the previous generation. With the changes to the cache structure and other behind-the-scenes upgrades, the difference with the new CPUs should be much more dramatic than the basic specs show.
Content Creation Workstations
Puget Systems offers a range of powerful and reliable systems that are tailor-made for your unique workflow.
Configure a System!
Labs Consultation Service
Our Labs team is available to provide in-depth hardware recommendations based on your workflow.
Adobe Photoshop
Read the full article: Adobe Photoshop: AMD Ryzen 5000 Series CPU performance
Summary: In the past, there were arguments for using either an AMD or Intel processor for Photoshop, but with the new Ryzen 5000-series CPUs, AMD takes such a large lead that it will be very difficult to justify using Intel over AMD. Because Photoshop doesn’t scale well with more cores, the higher-end Ryzen 9 5900X and 5950X might not be the best value for Photoshop since they are no faster than the Ryzen 7 5800X, but all three of these CPUs beat the Intel Core i9 10900K by a solid 10%. For those on a tighter budget, Ryzen 5 5600X outperforms the similarly-priced Intel Core i5 10600K by an even larger 22%.
Adobe Lightroom Classic
Read the full article: Adobe Lightroom Classic: AMD Ryzen 5000 Series CPU Performance
Summary: In the past, there were arguments for using an Intel processor for Lightroom Classic if you wanted to optimize for active tasks like scrolling through images, but with the new Ryzen 5000-series CPUs, AMD takes a solid lead no matter the task.
We saw some odd performance issues with the Ryzen 9 5950X, but the Ryzen 7 5800X and Ryzen 9 5900X beat the Intel Core i9 10900K by a solid 14% and 21% respectively, while the Ryzen 5 5600X outperforms the similarly-priced Intel Core i5 10600K by a bit smaller 11%.
Adobe After Effects
Read the full article: Adobe After Effects: AMD Ryzen 5000 Series CPU Performance
Summary: Before AMD launched the Ryzen 5000-series, either an Intel 10th Gen CPU or an AMD Ryzen 3000-series CPU would end up giving you very similar performance in After Effects. With the new Ryzen 5000-series CPUs, however, AMD takes a solid lead over Intel. Compared to the Intel Core i5 10600K, the new Ryzen 5 5600X is a solid 16% faster. And at the higher end, the Ryzen 7 5800X, Ryzen 9 5900X, and Ryzen 9 5950X range from 9% to 18% faster than the Intel Core i9 10900K.
Adobe Premiere Pro
Read the full article: Adobe Premiere Pro: AMD Ryzen 5000 Series CPU Performance
Summary: With Premiere Pro now supporting GPU-accelerated encoding/decoding of H.:strip_icc()/i.s3.glbimg.com/v1/AUTH_08fbf48bc0524877943fe86e43087e7a/internal_photos/bs/2021/o/O/LqMcS6RmSDelWORuWm2g/2014-03-24-amd-ou-intel.png)
264/HEVC media, Intel is no longer an auto-pick for Premiere Pro, allowing AMD to show what they are capable of on a level playing field. At many points of the stack, the AMD Ryzen 5000-series CPUs do extremely well, solidly beating similarly-priced Intel CPUs by 12-20%.
The exception is the AMD Ryzen 7 5800X, which still manages to pull ahead of the Intel Core i9 10900K, but only by 1% – effectively tying those two CPUs in terms of overall performance. This actually puts Intel in a very slight lead at this price point since being able to use Quick Sync rather than your GPU for H.264/HEVC encoding and decoding frees up the video card to be used for other things like processing GPU effects, debayering R3D media, etc.
DaVinci Resolve Studio
Read the full articles: DaVinci Resolve Studio: AMD Ryzen 5000 Series CPU Performance
Summary: AMD has been a strong choice for DaVinci Resolve Studio ever since the launch of the 3rd generation Ryzen and Threadripper CPUs due to their higher core counts and inclusion of PCI-E gen 4.
Interestingly, the new AMD Ryzen 5000-series processors are not much faster than the previous generation. But, it is enough for AMD to extend their lead over Intel.
At the bottom end, the Ryzen 5 5600X beats the Core i5 10600K by 14%, and the Ryzen 7 5800X sneaks by the Core i9 10900K by 5%. Further up the stack, AMD does even better with the Ryzen 5 5900X and 5950X beating the Core i9 10900X and 10940X by 29% and 22% respectively. This means that depending on your budget, an AMD Ryzen 5000-series CPU can be anywhere from 5% to a staggering 29% faster than a similarly priced Intel processor.
Unreal Engine
Read the full article: Unreal Engine: AMD Ryzen 5000 Series CPU Performance
Summary: AMD’s previous-gen Ryzen CPUs already provided a compelling option for Unreal Engine. The newly release Ryzen 5000 series CPUs have made it difficult to recommend anything else. They offer a 10-20% improvement over the previous generations and outperform any Intel CPU at similar price points or core counts.
AMD’s Threadripper is still the top choice in Unreal Engine, taking the top three spots in every test. However, if you are looking for a lower-priced system or something in a smaller form factor, AMD’s new Ryzen 5000 series is the way to go.
Cinema 4D
Read the full article: Cinema 4D: AMD Ryzen 5000 Series Rendering Performance
Summary: The fantastic single-threaded performance of these new Ryzen processors will make them excellent for working with modeling, animation, and physics in Cinema 4D – while also offering solid rendering speeds for their price point.
For pure rendering workloads, however, AMD’s Threadripper line is faster yet (and more expensive). Nothing in Intel’s Core or Core X series can compete here. It is worth noting that Cinema 4D also supports other rendering engines via plug-ins, and you can potentially get more performance with GPU-based rendering by combining multiple video cards.
Previous
Next
Previous
Next
V-Ray CPU Rendering
Read the full article: V-Ray Next: AMD Ryzen 5000 Series Rendering Performance
Summary: For the price, these new Ryzen processors offer great rendering performance – while also providing the best single-threaded speeds for modeling, animation, and other related tasks.
Each model in the 5000 Series family is 10-20% faster in V-Ray Next than the corresponding model from the 3000 Series.
However, AMD’s Threadripper processors are still the kings of single-socket rendering. It is also worth considering whether a GPU rendering engine or a hybrid that uses both the CPU and GPUs, could fit in your workflow. Those can scale much higher in performance within a single workstation.
RealityCapture
Read the full article: RealityCapture 1.1: AMD Ryzen 5000 Series Performance
Summary: AMD’s Ryzen 9 5950X is the fastest CPU we have tested yet in RealityCapture, and the 5900X and 5800X are both close behind. These processors offer fantastic per-core processing speeds, along with an ideal number of cores for this application.
The only limitation I can see to this platform is the maximum of 128GB with regards to memory support. That is not likely to be an issue in RealityCapture, but if it is a problem for anyone then I would advise moving to AMD’s Threadripper processors as the next closest option in terms of performance.
Agisoft Metashape
Read the full article: Agisoft Metashape 1.6.4: AMD Ryzen 5000 Series Performance
Summary: AMD’s Ryzen 7 5800X is now our top-performing processor for working with maps in Metashape. The Ryzen 9 5950X and 5900X are slightly faster for dealing with complex models, rather than maps, and the only competing CPU we have tested that is in the same ballpark as these is Intel’s Core i9 10900K.
Any of those four processors would be an excellent choice for photogrammetry in Metashape, but they do all share a single weakness: 128GB of maximum memory capacity. When working with extremely large image sets there could be times that additional RAM is needed, but a trade-off on pure performance would be required to get there.
Pix4D
Read the full article: Pix4D 4.5.6: AMD Ryzen 5000 Series Performance
Summary: AMD’s Ryzen 9 5950X is undeniably the fastest CPU for Pix4D – and the 5900X is a strong contender as well.
Even the lower Ryzen 7 and 5 models offer better performance for the price than their older models or Intel’s Core series equivalents. Most of these new CPUs are ~15% faster than the previous generation in Pix4D, while the Ryzen 5 5600X is «only» 10% faster than the 3600XT.
The only reason to go for a different model would be to access higher amounts of RAM (more than the 128GB limit of the Ryzen 5000 Series) for working with larger image sets, in which case AMD’s Threadripper processors are not much slower and support up to 256GB.
How well does the AMD Ryzen 5000 Series perform overall?
Across the board, AMD’s new Ryzen 5000 series CPUs do extremely well – in most cases easily out-pacing similarly priced processors from Intel. The exact amount depends on the specific application, but the Ryzen 9 5900X and 5950X in particular do very well, beating the Intel X-series CPUs by as much as 40%!
In content creation applications like the Adobe suite, DaVinci Resolve, and Unreal Engine, the lower-end Ryzen 5 5600X and Ryzen 7 5800X beat the similarly-priced Intel Core i5 10600K and Core i9 10900K by around 10%.
However, this performance is overshadowed by the Ryzen 8 5900X and 5950X which were both nearly 40% (or more) faster than the Intel Core i9 10900X and 10940X. To be fair to Intel, many of the applications in this category (especially lightly threaded ones like Photoshop and After Effects) are actually faster with the Intel Core i9 10900K than the more expensive Intel X-series processors. So, if you were to compare AMD and Intel-based not just on the price of the CPU, but also limiting it to the best performing CPU within that budget, AMD’s lead would actually cap out at closer to 20%.
For CPU rendering applications like CineBench and V-Ray, the results are fairly interesting and not actually a straight win for AMD across the board. The Ryzen 7 5800X in particular only matches the Intel Core i9 10900K – likely because the 10900K has 10 CPU cores versus the 8 on the 5800X. Now, the Core i9 10900K is a bit more expensive than the 5800X ($488 vs $449), but that $40 difference is a smaller gap than the $75 difference between the Core i7 10700K and the 5800X ($374 vs $449).
Further up the stack, however, the AMD Ryzen 9 5900X and 5950X take a commanding lead over Intel, performing up to 40% faster than a similarly-priced CPU from Intel. It is worth noting that if you are looking for the best rendering performance, the AMD Threadripper CPUs are even faster than the new Ryzen 5000 series and well worth the cost if they are within your budget.
In photogrammetry applications, the AMD Ryzen 5000 series is terrific, with either the Ryzen 7 5800X or Ryzen 9 5950X being the fastest CPU we tested depending on how well threaded the particular photogrammetry application is. The performance gains over Intel are not quite as pronounced as they are in the other categories we tested, but AMD maintains a strong 27% lead over Intel with the Ryzen 9 5950X. For the other processor models, the gain over Intel is less than 10%, but the Ryzen 5 5600X and 5900X still maintain enough of a performance advantage to make them a clear choice over the Intel CPUs that are currently available.
Overall, these new AMD Ryzen 5000 series processors are a great improvement over the previous generation, and in most cases place AMD in a solid lead over Intel. The Ryzen 7 5800X is the only model that faltered in some applications, and in some cases was actually a hair slower than the similarly-priced Intel Core i9 10900K.
The Ryzen 9 5900X and 5950X, however, are particularly good versus the Intel X-series CPUs that are in the same budget range. Performance gains of nearly 40% are not uncommon, although that is somewhat artificially inflated since the lower priced Intel Core 10th Gen CPUs can actually be faster than the X-series in many applications.
If you are looking for a workstation with one of these new Ryzen CPUs, we anticipate moving a significant number of our systems to the AMD Ryzen 5000 series in the coming weeks. Be sure to visit our solutions page to view our recommended workstations for various software packages, our custom configuration page, or contact one of our technology consultants for help configuring a workstation that meets the specific needs of your unique workflow.
AMD Ryzen Workstations
Puget Systems offers a range of powerful and reliable systems that are tailor-made for your unique workflow.
Configure a System!
Labs Consultation Service
Our Labs team is available to provide in-depth hardware recommendations based on your workflow.
Related Content
Latest Content
View All
Tags: 5600X, 5800X, 5900X, 5950X, After Effects, AMD, Cinebench, DaVinci Resolve, Lightroom CLassic, Metashape, Photoshop, Pix4D, Premiere Pro, RealityCapture, Unreal Engine, V-Ray
How Does AMD’s Ryzen 5 1600 Compare to Mainstream $200 CPUs?
In 2017, AMD’s Ryzen 5 1600 was a popular choice among PC enthusiasts because of its incredible price-to-performance ratio across gaming and productivity workloads. Not only did AMD’s offering perform admirably against Intel’s competing Kaby Lake lineup, but it also redefined how value-oriented CPUs should be in the long run.
With significant improvements in microprocessor architecture over the past five years, the Ryzen 5 1600 is now more of an entry-level option. Regardless, how does it stack up compared with mainstream CPUs launched by AMD and Intel in 2022? Let’s find out!
AMD Ryzen 5 1600: Specifications and Overview
In terms of processing resources, AMD’s Ryzen 5 1600 is a 6-core, 12-thread CPU based on the Zen microarchitecture and built using GlobalFoundries’ 14nm manufacturing node. Featuring a base clock of 3.2GHz and a maximum boost clock of 3.6GHz, the Ryzen 5 1600, at $219, was meant to compete against Intel’s Core i5-7500 and i5-7600K.
Unlike its «X» SKU, the Ryzen 5 1600 comes at a TDP (Thermal Power Design) of just 65W and is bundled with AMD’s 95W Wraith Spire cooler capable of pushing this CPU well beyond its stock frequency. Nevertheless, both variants benefit from 16MB of L3 cache and support for XFR (eXtended Frequency Range), a feature that utilizes AMD’s SenseMI Technology to ramp up core clocks based on available thermal headroom.
Additionally, one of the major advantages that the Ryzen 5 1600 had over its direct competitors was the inclusion of an unlocked ratio multiplier. Paired with an inexpensive yet capable B350 motherboard and high-speed DDR4 memory, an overclocked Ryzen 5 1600 matched Intel in terms of per-core performance and offered stiff competition against its own lineup.
Despite glaring issues with memory compatibility and power management modes at launch, the Ryzen 5 1600 provided exceptional value for budget-conscious gamers and content creators. Following its success, AMD continued to expand its Ryzen 5 lineup by releasing mainstream, affordable CPUs with gen-on-gen upgrades every subsequent year.
While similarly priced 1600 alternatives like the Ryzen 5 2600, 3600, and 5600 brought major performance and efficiency gains over their predecessors, the same cannot be said about AMD’s Zen 4 CPUs, specifically the Ryzen 5 7600X. Architectural improvements aside, when you factor in the cost of upgrading to a brand-new platform, the 7600X’s value proposition seems somewhat underwhelming with regard to expensive AM5 motherboards and DDR5 memory kits.
On the flip side, mainstream CPUs from Intel, such as the Core i5-12400 and i5-12600K, show how much of a difference competitive pricing can mean for the average consumer. Furthermore, when comparing Intel’s 12th Gen Alder Lake CPUs against AMD’s Ryzen 5000 Series, it is evident that Ryzen has the upper hand in terms of platform compatibility, whereas Alder Lake’s dominance lies in its support for both DDR4 and DDR5 memory configuration.
However, when we pitch the Ryzen 5 1600 against current-gen, $200 products like the Ryzen 5 5600 and Core i5-12400, it is fascinating to see just how far CPU architecture has evolved in the last few years. Take a look at their specifications below to better understand the design philosophy behind each of these mid-range chips.
|
Specifications |
Ryzen 5 1600 |
Ryzen 5 5600 |
Core i5-12400 |
|---|---|---|---|
|
Cores/Threads |
6/12 |
6/12 |
6/12 |
|
Base/Boost Clock |
3. |
3.5GHz/4.4GHz |
2.5GHz/4.4GHz |
|
L3 Cache Size |
16MB |
32MB |
18MB |
|
Default TDP |
65W |
65W |
65W |
|
Lithography |
14nm |
7nm |
10nm |
|
Socket Type |
AM4 |
AM4 |
LGA 1700 |
|
System Memory Type |
DDR4 |
DDR4 |
DDR4/DDR5 |
|
Maximum Memory Speed |
Up to DDR4-2667 |
Up to DDR4-3200 |
Up to DDR4-3200/DDR5-4800 |
|
Memory Channels |
2 |
2 |
2 |
|
Integrated Graphics |
None |
None |
Intel UHD Graphics 730 |
|
PCI Express Version |
PCIe 3. |
PCIe 4.0 |
PCIe 4.0/5.0 |
|
Overclocking Support |
Yes |
Yes |
No |
|
Supplied Cooler |
AMD Wraith Spire |
AMD Wraith Stealth |
Intel Laminar RM1 |
|
Launch Date |
Q2’17 |
Q2’22 |
Q1’22 |
|
MSRP |
$219 |
$199 |
$199 |
2GHz/3.6GHz
0Ryzen 5 1600 vs. Ryzen 5 5600 vs. Core i5-12400: Synthetic Benchmarks
To determine the relative performance of AMD’s Ryzen 5 1600 against the Ryzen 5 5600 and Core i5-12400, we’ve decided to put them through a handful of free benchmark programs for Windows.
Regarding test setup, both CPUs from AMD have been paired with MSI’s B450 TOMAHAWK motherboard, 16GB (2×8) of G.Skill Flare X DDR4-3200 CL14 RAM, and an ASUS TUF Gaming GeForce RTX 3060 OC Edition GPU.
Meanwhile, the Intel test bench used MSI’s MAG B660 TOMAHAWK motherboard and the same GPU and memory configuration.
Since the Ryzen 5 1600 isn’t natively supported on the latest version of Windows 11, we ran all the benchmarks by installing a fresh copy of Windows 10 Pro (Version 21h3) on a WD_Black SN850 NVMe PCIe Gen4 SSD. So without further ado, let’s get straight to the test results!
|
Applications |
Ryzen 5 1600 |
Ryzen 5 5600 |
Core i5-12400 |
|---|---|---|---|
|
Cinebench R23 (Single-Core) |
892 |
1476 |
1664 |
|
Cinebench R23 (Multi-Core) |
6468 |
11126 |
12342 |
|
Geekbench 5 (Single-Core) |
902 |
1552 |
1636 |
|
Geekbench 5 (Multi-Core) |
5235 |
7895 |
8054 |
|
CPU-Z (Single Thread) |
398 |
600 |
659 |
|
CPU-Z (Multi Thread) |
3157 |
4738 |
4838 |
|
Blender Open Data Benchmark (Median Score) |
106. |
166.69 |
175.92 |
|
7-Zip (Compression Rating) |
36.690 GIPS |
66.708 GIPS |
55.562 GIPS |
|
7-Zip (Decompression Rating) |
52.608 GIPS |
54.694 GIPS |
40.522 GIPS |
14In CPU-intensive benchmarks like Cinbench R23 and Geekbench 5, the Ryzen 5 5600 and Core i5-12400 outperformed AMD’s Summit Ridge CPU by a significant margin. Although the Ryzen 5 1600’s score in Blender seems adequate for daily workloads, creative professionals might want to get their hands on a brand-new Zen 3 or Alder Lake chip, especially if they’re willing to cut down on render times.
A similar pattern can be observed in 7-Zip’s File Compression benchmark, where the Ryzen 5 5600 and Core i5-12400 enjoyed a massive performance boost of around 50-80% over the older Zen microarchitecture. While the i5-12400’s Decompression Rating is lower than either of the Ryzen CPUs, the former makes up for it purely in terms of Compression and Decompression speeds.
Ryzen 5 1600 vs. Ryzen 5 5600 vs. Core i5-12400: Gaming Benchmarks
Now, one of the major downsides of AMD’s 1st-Gen Ryzen CPUs was their lackluster performance in single-threaded applications. For games heavily reliant on single-core performance, even an overclocked Ryzen 5 1600 failed to match or beat Intel’s Core i5-7500 and 7600K in terms of average frame rates.
To ascertain whether the Ryzen 5 1600 is a capable performer against the 5600 and Core i5-12400 in modern AAA games, we’ve opted for an identical test rig, pairing all three CPUs with an RTX 3060 GPU from Nvidia. Furthermore, the testing methodology involved running five of the most popular yet demanding AAA titles at 1080p resolution and Ultra settings to produce a more realistic workload on both the CPU and GPU.
|
Games |
Ryzen 5 1600 |
Ryzen 5 5600 |
Core i5 12400 |
|---|---|---|---|
|
Assassin’s Creed Valhalla (Ultra High, DX12) |
45 FPS |
56 FPS |
59 FPS |
|
Cyberpunk 2077 (Ultra, DX12) |
43 FPS |
56 FPS |
57 FPS |
|
Red Dead Redemption 2 (Favor Quality, Vulkan) |
47 FPS |
59 FPS |
61 FPS |
|
Shadow of the Tomb Raider (Highest, DX12) |
86 FPS |
109 FPS |
119 FPS |
|
Counter-Strike: Global Offensive (High/Very High, DX9) |
292 FPS |
390 FPS |
393 FPS |
Across all of our gaming benchmarks, it is evident that the Ryzen 5 1600 struggles to maintain a steady frame rate compared to either of the current-gen CPUs.
For example, in Assassin’s Creed Valhalla, both the Ryzen 5 5600 and Core i5-12400 were at least 20-30% faster than the Ryzen 5 1600 with a similar performance uplift in Cyberpunk 2077 and Red Dead Redemption 2.
Although Shadow of the Tomb Raider’s in-game benchmark shows favorable results for the Ryzen 5 1600, which managed an average of 86 FPS, it is nowhere near the performance of the newer Zen 3 or Alder Lake chip. Meanwhile, CS:GO puts both the Ryzen 5 5600 and Core i5-12400 in an entirely different territory as they delivered around 100 frames more than the Ryzen 5 1600’s average FPS.
For gamers with a high refresh rate display, upgrading to a Ryzen 5 5600 or Core i5-12400 seems like the better option, as both CPUs will help maintain a smooth gameplay experience across the board. However, if you own a Ryzen 5 1600 and are willing to make the most out of its stock gaming performance, consider investing in a FreeSync monitor that can run Nvidia’s Adaptive-Sync technology at a wide VRR range.
Is the Ryzen 5 1600 Still Relevant in 2023?
Given that the Ryzen 5 5600 and Core i5-12400 are versatile, mainstream chips launched in 2022, AMD’s Summit Ridge CPU can no longer be considered a viable option for 2023. For early Zen adopters, switching to the Ryzen 5 5600 seems like a no-brainer, especially when it is a drop-in upgrade on existing B350 and B450 motherboards.
However, if you’re building a PC from scratch, we recommend going for Intel’s Core i5-12400, as it happens to be the superior product on most occasions.
CES-2023. Intel and AMD introduced new productive processors for laptops
Fresh number
RG-week
Rodina
thematic applications
Union
Fresh number
Category:
Technologies 11.01.2023 13:59
Vladimir Tikhonov
Intel and AMD announced many high-performance mobile processors at CES 2023. What interesting new products await users in the coming year?
screenshot from video/ intel.
ru
In the performance segment, AMD introduced the Ryzen 7045HX series, designed for heavy gaming and creative tasks. The chips are based on a 5nm process and nominally generate 45-55W of heat. It is noteworthy that the new line of AMD processors uses a multi-chip layout (one I / O die and two crystals with computing cores), which was previously only characteristic of the company’s desktop processors. The flagship of the series, the Ryzen 9 7945HX, offers users 16 cores (32 threads) based on the Zen 4 architecture, clock speeds up to 5.4 GHz, and 80 MB of L3 and L2 cache. AMD also introduced Ryzen 97845HX (12c/24t*), Ryzen 7 7745HX (8c/16t) and Ryzen 5 7645HX (6c/12t).
Intel is pursuing a similar strategy, integrating the best features of desktop processors into the high-performance mobile segment with changes in heat and clock speeds. The flagship of the line, the Core i9-13980HX, has 8 productive and 16 energy-efficient cores (32 threads in total, like the desktop i9-13900K), a clock frequency of 5.
6 GHz, and a heat dissipation of 55 watts. Intel also released i9-13950HX and i9-13900HX, which differ from their older counterpart in lower clock speeds, i7-13850HX (20c / 28t), i7-13700HX (16c / 24t), i7-13650HX (14c / 20t), i5-13600HX and 15-13500HX (13c / 20t, differ in frequencies), as well as i5-13450HX (10c / 16t).
In the mid-range segment, AMD released 35-45W Ryzen 7040HS series processors. The line consists of three models: Ryzen 9 7940HS (8c/16t), Ryzen 7 7840HS (8c/16t, but with reduced frequencies), and Ryzen 5 7640HS (6c/12t). The manufacturer significantly increased the amount of cache memory in 7040HS chips compared to the previous generation: from 16 MB to 32 MB of L3 cache.
Intel also updated the mid-segment with the announcement of 11 new Raptor Lake H family processors. The 45W chips have up to 14 compute cores (maximum 6 performance and 8 power efficient cores) and up to 24MB L3 cache.
HX series processors from both companies will be in demand in high-performance laptops with discrete graphics, while AMD HS / Intel H chips, due to a more powerful iGPU, will look more interesting in devices without a discrete graphics card.
The first devices based on new chips from Intel and AMD are expected in the first half of 2023.
#Technologies
#processors
#CES 2023
Main
-
The head of Chechnya Kadyrov reported the capture of the Armed Forces of the Armed Forces
-
Minous Minor Proposed to increase social pensions from April 1 by another 3.3 percent
-
Putin signed the law on the protection of the Russian language
-
Ministry of Defense of the Russian Federation: Air Defense Forces conducted training with the involvement of fighters in the western zone
-
SK opened a case after the crash of the Mi-8 helicopter in the Murmansk region
Intel and AMD created a headache for all data centers. The latest processors turned out to be very voracious and too hot
Technique
|
Share
The new 4th generation AMD Epyc 4 Genoa and Intel Xeon Scalable server chips are not suitable for use in modern data centers.
They are too energy-intensive and too hot — many data centers are not able to provide servers based on them with the necessary amount of electricity and high-quality cooling. A massive transition to these CPUs will require a radical review of the infrastructure of data centers, both existing ones and those being prepared for construction.
Data centers have problems
New server processors from Intel and AMD can create a lot of trouble for data center owners if they want to switch to using them. As The Register writes with reference to the Uptime Institute, the infrastructure of many modern data centers is simply not designed for such processors, and the point is not at all that these centers are outdated, but that AMD and Intel have created real champions in terms of energy consumption and allocated heat.
In particular, the new Epyc 4 Genoa, which AMD showed at the end of 2022, as well as even more modern fourth-generation Intel Xeon Scalable, which was presented in January 2023, are mentioned.
These are the most productive, but at the same time, energy-intensive chips of the famous American duo as of January 2022. Processors consume 400 and 350 watts, respectively.
One gets the impression that neither Intel nor AMD cares at all how the buyers of their server processors or ready-made solutions based on them will solve the problem with their cooling and power supply. In the desktop and mobile (laptop) processor segments, the two companies are conversely aiming to release the least hot and energy-hungry processors. It turns out, however, so far only with AMD, which has recently introduced 5-nanometer processors (the smaller the technical process, the colder and more economical the CPU without losing performance). Intel is stuck at 10nm.
Tiny data center-wide microchip will become a source of unnecessary difficulties
Uptime Institute is a consortium of companies formed in 1993 to maximize data center efficiency. The organization is known, in particular, for the creation of a worldwide multi-level classification of data centers — Tier I — Tier IV.
Heating is no longer needed
The increased power consumption and heat dissipation in this case is due to the increase in the number of cores and their clock frequency in the new AMD and Intel server processors compared to their predecessors. According to The Register, the achievements of companies in the server CPU segment now allow you to assemble a compact system on two processors with 192 (in the case of new AMD chips) or 120 cores (Intel) in total in the case of the most common rack server in 2U format. A few years ago, something like this was unattainable.
This is the problem with many of today’s data centers. In fact, they get a very compact solution, inside which there can be two not the coldest chips. Placing several such servers in one rack with a minimum distance between them, as well as several fully occupied racks next to each other, will inevitably lead to difficulties in cooling them.
At the same time, each such server, taking into account all its internal equipment, can easily consume significantly more than 1 kW of electricity.
The task of «feeding» a horde of such servers will also be very, very non-trivial.
Too many consequences
The Uptime Institute report says that the current trend to rapidly move to hotter, more power hungry servers is reversing decades of data center capacity planning, says Uptime, adding: destabilize the existing foundations about the design of facilities,” the report says.
Consortium analysts noted that one typical 42U rack in current data centers has a design power of less than 10kW. But with modern systems striving for higher computational density and, accordingly, «power density», this is no longer enough. The Uptime Institute estimates that energy consumption will quadruple within a few years, to approximately 42kW.
At this rate, AMD and Intel will quickly give way to the server CPU market to much cooler and more economical solutions based on ARM that are not inferior in performance
An increase in the total energy consumption of the racks will be accompanied by an increase in the heat generated by them.
However, many data centers are now very limited in their ability to provide the necessary airflow to cool racks with densely spaced servers, the heat dissipation of which is constantly growing.
Nikolai Agrinsky, Infosecurity: The issue of not fragmentary import substitution, but the achievement of full digital sovereignty remains open
Safety
Air-cooling in data centers is the standard, and its capabilities are becoming insufficient. Currently, liquid cooling systems for data centers are gradually developing. They have long become the norm in home system units, but so far they are almost never used in data centers. In addition, all these systems have not yet been brought to a single standard, which hinders the growth in the pace of their implementation.
Attempts to resolve this issue are certainly being made. In particular, Intel, the largest supplier of x86 server processors in the world, is working on the standardization of liquid cooling.
Also, Intel and other organizations are working on the creation of standards for the so-called «immersion» cooling (immerse — immerse, English), in which the cooled object is literally immersed inside the coolant. To do this, at the beginning of 2022, Intel announced the creation of a “megalaboratory” worth $700 million, which will oversee the development of standards for both immersion and liquid cooling. Meanwhile, members of the open community Open Compute Project have been working in this direction since 2018.
As part of the Open Compute Project, participants share developments in the field of software, hardware and physical design of modern data processing centers and equipment for them. After five years, they still have not achieved the adoption of cooling standards for data centers other than air.
Ecology and economics
As data centers migrate to new Intel and AMD processors, their electricity bills will also rise. Moreover, this growth can be rapid not only because of the increase in the energy intensity of the chips themselves, but also against the background of the rise in the cost of electricity itself in Europe and North America.