To say that Cyberpunk 2077 has gotten off to a troubled start is putting it mildly. Bugs and glitches abound, performance on previous-gen consoles is quite terrible, and even on PC, you’ll want a fairly potent rig. We’ve looked at GPU performance, CPU scaling, and even RAM requirements. Meanwhile, the v1.05 patch arrived for PC on Dec. 20 and could potentially invalidate previous benchmarks, so we wanted to do a quick sanity check. There’s no sense in putting in dozens of hours of testing if things haven’t really changed, right?
Based on our limited testing, and depending on your hardware and settings, the performance differences between 1.04 and 1.05 range from negligible to pretty decent.
If you’re running with an AMD CPU and haven’t already used the SMT hack to enable the extra threads, v1.05 fixes that for you. How much that matters varies based on your CPU. 4-core/8-thread and 6-core/12-thread Zen 2 and Zen 3 CPUs appear to benefit the most, and 8-core/16-thread chips can also see a moderate boost. However, the 12-core/24-thread and 16-core/32-thread CPUs appear to have enough cores to go around, so SMT doesn’t really benefit the game (and may actually drop performance slightly).
Image 1 of 6
(Image credit: Tom’s Hardware)
Image 2 of 6
(Image credit: Tom’s Hardware)
Image 3 of 6
(Image credit: Tom’s Hardware)
Image 4 of 6
(Image credit: Tom’s Hardware)
Image 5 of 6
(Image credit: Tom’s Hardware)
Image 6 of 6
(Image credit: Tom’s Hardware)
Image 1 of 6
(Image credit: Tom’s Hardware)
Image 2 of 6
(Image credit: Tom’s Hardware)
Image 3 of 6
(Image credit: Tom’s Hardware)
Image 4 of 6
(Image credit: Tom’s Hardware)
Image 5 of 6
(Image credit: Tom’s Hardware)
Image 6 of 6
(Image credit: Tom’s Hardware)
What about GPUs? We originally tested the RTX 3080 and RX 6800 XT with version 1.03. On the Nvidia GPU, 1.05 performance is largely unchanged—basically within the margin of error for live frametime captures using OCAT. We did see improvements in 99.9th percentile frametimes, which should mean slightly fewer stutters/hitching, but there’s far more variability in 99.9th percentiles (i.e., a few ‘bad’ frames can quickly skew the results).
On the AMD side, our 1.03 results are higher than 1.05, which in turn is slightly to moderately higher than 1.04. However, I think I may have inadvertently had AMD’s Rage Mode overclocking enabled on the 6800 XT during the first round of 1.03 testing, which could account for the differences we’re now seeing. That or the stars were in alignment on those first tests, but that’s no longer the case. (In other words, I’ve tried to recreate the original test results on 1.03 and cannot, despite cleaning out all GPU drivers and reinstalling. To be fair, it was a crazy night of benchmarking Cyberpunk 2077 at launch.)
Otherwise, looking at current 1.04 and 1.05 results, 1.05 does appear to improve AMD’s performance at 1440p and 4K. 4K performance is 6 percent higher at medium settings and 9 percent higher at ultra settings, while 1440p medium performance is only 1 percent faster, but 1440p ultra is 5 percent faster. There’s still no support for ray tracing on AMD’s GPUs, but given the number of other bugs and glitches, we imagine that’s a lower priority than getting the game to work properly.
What about lower-tier GPUs? Some of those may see similar improvements, but we’re going to hold off on retesting any other cards. Besides, it’s not like most people would want to run with a buggier 1.03 or 1.04 build. Then again, there’s some entertainment in the various things that can go wrong in Night City—though we’d just google those. It can almost be more fun watching the best glitches than playing through the actual story.
The Intel Core i9 – 10900 and its K version have been leaked on CPU-Z and we can see the apparent performance offered by the two highest-end models that Intel will have in its 11 ª Gen for desktop.
The Intel 11 ª Gen for desktop will have two very high-end models, as expected, called Core i9 – 11900 and i9 – 11900 K . Just these two models are the ones that have just been filtered in CPU-Z, showing the score of each of them in the same program and comparing it with the previous ones to get an idea of their performance. In single core, the i9 – 11900 scores 582 while the i9 – 11900 K reaches 597 points. This, compared to its previous analogues (the i9 – 10900 and 10900 K), implies a increase in the CPI around 12% .
The Intel Core i9 – 11900 K rears its head with 8 cores and 12% more CPI than previous generation
On the other hand, in the multicore environment we see that they lose out due to the fact that have fewer cores . The above i9 – 10900 have 10 cores while i9 – 10900 have 8 . However, the multicore performance loss stays between 5 and 10%, so the IPC gain is paramount and Intel knows it. Nor are there many applications that make good use of 21 threads , so counting on 16 threads with a higher IPC it’s a nice change.
However, these processors are the ones that will have to compete face to face with the AMD Ryzen 5000 (Zen 3 ), and we already saw that the IPC of these Ryzen is really high and leaks appeared that left the Intel behind in performance.
End of Article. Tell us something in the Comments or come to our Forum!
Pablo López
With 15 years I started to overclock my PC to get every extra FPS I could in games and scratch a few milliseconds in SuperPi, while I was constantly posting about hardware on the Geeknetic forum as a user and reader. They were probably so fed up with continually reading me on the forum that I became part of the writing team, where I continue to report on the latest in technology. Astrophysics and PC games are the hobbies that, after hardware, cover most of my free time.
Available for OEM customers only, Ryzen 9 5900 and Ryzen 7 processors maintain the same specifications underlying the X proposals on the market, operating with a maximum TDP of 65 Watt
by Paolo Corsini published 21 December 2020 , at 08: 11 in the Processors channel RyZen AMD Zen
AMD would be preparing the debut of two new Ryzen processor versions of the series 5000, characterized by low consumption compared to the proposals now on the market. These would be processors Ryzen 9 5900 and Ryzen 7 5800 , therefore based on Zen 3 architecture, characterized by a TDP equal to 65 Watt : the anticipation comes from two Twitter channels at this address and this other address.
These processors will maintain the same number of cores as the corresponding proposals on the market with TDP equal to 105 Watt , the latter with the suffix X to distinguish them, as well as the various technical characteristics; the only exception would be represented by the clock frequency which will be lower in order to contain the overall consumption. The clock frequencies of these CPUs are not yet known.
Proposals of this type will be offered by AMD to OEM customers only manufacturers of desktop systems, for which CPUs with low consumption despite a high number of cores represent a very interesting element for their customers.
Prices should also be slightly lower than those of the proposals with standard consumption, with a ventilated reduction in 50 dollars from the official list. Since these are processors designed for OEM customers, they should not in any case be available for purchase in the retail channel, unless special commercial policies implemented by some retailers.
Everything seems to indicate that the dry time in the world of processors is running out and the next few years will be interesting premieres. After all, we will face the debut of many interesting technologies and standards, which in turn means that we are in for an extremely interesting duel between Intel vs AMD. Several leaksters shared with us their information on the publishing plans of both giants, which show that the coming years will be extremely busy for producers. We learned the details about not only desktop processors, but also mobile units. We just have to remember that this is unofficial information and ultimately the plans of both companies may change. Well, let’s finally get to the point.
We just have to remember that this is unofficial information and ultimately the plans of both companies may change. Nevertheless, everything indicates that the dry time in the world of processors is running out and the coming years will give rise to a number of very interesting premieres.
Intel Core Processors 11. Generation is not only Rocket Lake? Core i3, Pentium and Celeron may be Comet Lake Refresh chips
We already know that Intel will release Rocket Lake-S chips in the first quarter 2021 of the year, and in the fourth quarter comes Alder Lake-S (10 nm), to consist of Golden Cove and Gracemont cores and Xe Gen graphics 12 with EU units 32. As we mentioned recently, the Blue team will additionally present the Comet Lake Refresh models, which are to represent the cheaper next year’s Core i3, Pentium and Celeron systems. as for AMD, we expect the Vermeer series (Zen 3) to continue to be Red’s core offering at least until the third quarter 2021 of the year when architecture refresh under the code name Warhol. Chips are to be released between the fourth quarter 2021 and the second quarter 2022 year. It’s not yet clear what makes Warhol different from Vermeer, but you can expect rather minor changes (maybe something like Ryzen 9 3900 X – > Ryzen 9 3900 XT). The next major evolution will take place in the third quarter 2022 of the year, when AMD will begin distributing Raphael chips based on Zen 4 (5nm) architecture. Raphael is expected to provide DDR5 support and will likely be the first generation AMD desktop units to feature integrated on-die graphics (possibly RDNA2 based).
AMD Ryzen 9 5900 HX with high Turbo clock – top APU Cezanne-H processor as powerful as Intel Core i7 – 10700 K
AMD Ryzen for laptops – we got to know the company’s detailed plans for years 2021 – 2022. First APU RDNA 2 already in 2021 on
If it is about Intel and the laptop market (AMD’s offer is described in a separate news), you can expect that Tiger Lake-H chips 45 W will debut in high-end devices in the second quarter 2021 of the year. A year later, in the second quarter 2022, the Alder Lake-H generation 55 W z 32 UE Xe Gen 12 Supports PCIe 5 and DDR5. Interestingly, the Tiger Lake-H line will also have a variant 35 W, which may appear earlier than version part 45 W with possible support for LPDDR5 RAM. In the second quarter 2022 the Alder Lake series will receive units 45 W as lower power variant with graphics 96 – EU Xe Gen 12. In terms of units from the Mobile-U series, the successors of the Tiger Lake UP3 systems 28 will be Alder systems Lake-P U 28 to offer 96 EU Xe Gen 12 and PCIe Gen5 support. At the same time, we will also see Alder Lake-P U 15 as the successor to Tiger Lake UP4 15 In the fourth quarter 2021. There is still the category of low-voltage processors – here Intel already has Tiger Lakes and UP4 (9 W – 5900 and Lakefields (5W – 7W), which will be offered until the third quarter 2021, and then they will be replaced by the Alder Lake-P U9 and Alder Lake-M models. These first chips will offer the same core and iGPU configuration as Alder Lake-P U 15, while Alder Lake-M will have 64 Xe Gen graphics EU units 12.
Yesterday, we informed about AMD’s plans for Ryzen mobile processors for the years 2021 and 2022. The company is working on several different series of APUs that are prepared for different types of laptops. Next year’s APUs include Cezanne, Lucienne and Van Gogh. The next year will be premieres of Rembrandt, Barcelo and DragonCrest. Next year, only one APU series will use completely new graphics chips based on the RDNA 2 architecture – it will be the Van Gogh family. In turn, Cezanne and Lucienne will once again be based on Vega. In 2022, APU Rembrandt and DragonCrest will already use the integrated iGPU RDNA 2. Meanwhile, more information about individual APU groups as well as the first Cezanne processors, prepared with desktops in mind.
According to new information, AMD is working on at least two Cezanne desktop processors: Ryzen 5 5600 G and Ryzen 7 5800 G. In addition, we obtained new information about the laptop APU Van Gogh, DragonCrest and Rembrandt processors.
AMD Ryzen for laptops – we got to know the company’s detailed plans for years 2021 – 2022. First APU RDNA 2 already in 2021
According to the latest reports, AMD is working on at least two APU processors Cezanne for desktops. The first is 6-core and 12 – threaded Ryzen 5 5600 G, the other is 8-core and 16 – threaded Ryzen 7 5800 G. Both chipsets are most likely prepared for the OEM market, which means that they will not be available for retail sale, as well as this year’s Renoir APU processors. In addition, the company is working on more Vermeer systems – Ryzen 7 5800 (8C / 16 T) and Ryzen 9 5900 (12 C / 24 T). Both Zen 3 processors have TDP lowered to 65 and are also prepared exclusively for the OEM market. They will therefore be available in the form of ready-made computer sets.
AMD Ryzen 9 5900 HX with high Turbo clock – top APU Cezanne-H processor as powerful as Intel Core i7 – 10700 K
In addition, we got to know new details about the mobile APU Rembrandt, Van Gogh and DragonCrest. In the case of the first of these series, we will get a slightly larger APU compared to, for example, this year’s Renoir. As a reminder, the area of the Renoir monolith chip is 156 mm². AMD Rembrandt APU, which will be produced in TSMC 6nm lithography, is expected to have a surface area of approximately ~ 209 mm² (12, 82 mm x 16, 25 mm). The larger size is due to e.g. using a more complex graphics chip, based on the RDNA 2 architecture. Interestingly, the total area of the BGA (BGA Package) will be identical and will amount to 25 x 35 mm. The stand, however, will be different – instead of FP6, the manufacturer will use FP7. However, this is more in the form of a curiosity, because these processors are soldered anyway, and cannot be replaced. In the case of APU Van Gogh and DragonCrest, another important information appeared. These are not processors made for cheap laptops. They will be prepared with the so-called premium constructions in mind. Laptops with noticeably better build quality, slim and very mobile.
We already know the approximate specs for the AMD Cezanne-H APUs, which will debut next month. AMD is preparing for the conference at the CES virtual fair 2021, during which the following units will be shown : AMD Ryzen 5 5600 H, Ryzen 7 5800 H, Ryzen 9 5900 HS and Ryzen 9 5900 HX. The latter is a new, flagship model designed for the most efficient laptops. New performance tests AMD Ryzen 9 5900 HX in GeekBench 5 have appeared on the network. The data reported by the benchmark shows that that the Cezanne-H system worked with a Turbo clock at 4.7 GHz, so also 100 MHz higher than the samples from a few weeks ago. The processor based on Zen 3 cores almost defeated one of the most efficient Intel desktop processors from the Comet Lake-S line in the new test.
The first test of the AMD Ryzen 9 processor 5900 HX from the Cezanne-H series APU laptop. The 4.7 GHz processor achieved a result comparable to that of the Intel Core i7 in the GeekBench test – 10700 K.
AMD Ryzen for laptops – we got to know the company’s detailed plans for years 2021 – 2022. First APU RDNA 2 already in 2021
The base clock of the AMD Ryzen 9 processor 5900 HX is 3.3 GHz, the same as in the predecessor in the form of Ryzen 9 4900 H. Here, however, newer and more refined Zen 3 cores were used. Interestingly, according to the reported tests, the Turbo clock was 4.7 GHz. It is therefore the highest clocked Ryzen ever to hit laptops. In addition, it will be the only Cezanne-H APU with an unlocked multiplier, and thus the possibility of additional OC. We hope that you will also be able to go crazy with your memories. Regular DDR4 Usage 3200 MHz CL 22 to such a unit is almost a crime.
In the single thread test, the results of the top APU Cezanne-H look quite promising. CPU obtained 1534 / 1547 points – this is a bit less than the Zen 3 desktop flagships from the Vermeer family, but also noticeably more than the Renoir systems. Ryzen 9 5900 HX did very well in the multi-thread test, scoring just over 9000 points. This is a noticeably higher result than both Ryzen 7 5800 H and Ryzen 9 4900 H. Interestingly, at the same time the top APU Cezanne almost equals the desktop Intel Core i7 – 10700 K. Here, the results depend heavily on specific platforms – sometimes the Comet Lake CPU performs lower, sometimes slightly higher. Anyway, this level of performance with incomparably lower energy consumption is even more impressive than what we have already seen on the occasion of APU Renoir. AMD Ryzen 9 5900 HX with a score of approx. 9000 points also beats the current Intel flagship in laptops in the form of Core i9 – 10980 HK – this chip usually does not exceed 8000 points in the multithreaded test.
AMD Ryzen 9 5900 HX vs AMD Ryzen 7 5800 H – single thread test
AMD Ryzen 9 5900 HX vs AMD Ryzen 7 5800 H – multi-thread test
AMD Ryzen 9 4900 H vs AMD Ryzen 9 5900 HX – comparison of the top Renoir and Cezanne.
Several newest bad tests of the current Intel flagship in laptops – Core i9 – 10980 HK.
Cyberpunk 2077 is the most hotly-anticipated game in recent history, but it has proven to be punishing for even the most well-equipped PCs. We set out to test the new game with a wide range of the best CPUs on the market to determine which processors provide the best performance in a wide range of resolutions. We’ve tested AMD’s Ryzen 9, 7, 5, and 3 processors across several generations, along with Intel’s Core i9, i7, i5, and i3 processors as well.
This article focuses entirely on CPU perforamnce: If you’re looking for a closer look at how the game performs with a range of different GPUs, be sure to check out our in-depth Cyberpunk 2077 performance analysis. Thus far, our CPU testing has been with the current 1.04 patch, which brought minor performance enhancements over the original game code. However, CD Projekt Red has announced that it plans to release a new 1.05 patch for PCs soon (consoles already have the patch) with some performance enhancements, but the impact of those changes remains to be seen.
We’ll update our testing after the new patch, but except for select AMD models’ performance, we don’t expect huge changes to the CPU benchmark hierarchy we’ve created below. We could see slight increases in each processor’s overall performance but we expect that the performance deltas between the chips, and the rankings, will remain similar.
Cyberpunk 2077 has certainly earned a reputation for buggy performance, and one of the most surprising errata revolves around how the game code identifies AMD’s Ryzen processors. Resourceful redditors have discovered that the game code significantly harms performance on some AMD processors, because it apparently checks to see if your CPU uses Team Red’s 2011-era Bulldozer platform and, if not, disables multi-threading.
The enthusiast community discovered an easy-to-apply patch, which you can learn about here, that corrects the issue on select AMD models. CD Projekt Red says that it will also include similar code in the upcoming 1.05 patch. However, the game developer says the patch will only address Ryzen models with six or fewer cores, and that eight, 12, and 16 core models are unaffected. Our testing says otherwise — we continue to see performance gains with eight-core Ryzen 5000 series processors with the fix applied, so we’ll have to see how the upcoming patch addresses the eight-core models. The remainder of the pending changes listed for the 1.05 patch don’t appear to have massive implications for the game engine.
In either case, we’ve included testing both with and without the patch applied for the Ryzen 3000 (Zen 2) and 5000 (Zen 3) series processors with eight or fewer cores, but we found that the patch does little to improve performance with the Ryzen 1000 (Zen) and Ryzen 2000 (Zen+) models. For the older Ryzen models we’ve found that the change introduces buggy behavior or performance regressions, so we’ve excluded those extra tests.
With the preamble out of the way, let’s see how the Cyberpunk 2077 CPU performance hierarchy stacks up.
Cyberpunk 2077 CPU Benchmarks Setup
We tested with the highest number of NPC’s available and followed the same benchmark methodology used for our Cyberpunk 2077 PC Benchmarks, Settings, and Performance Analysis article.
We begin our test sequence at the entrance to V’s apartment building and walk a path down the stairs, across the street to the left, hang a right at the vending machine, then head across the street towards the two policemen on the other side. From there, we proceed to the right across the pavilion, down another set of stairs, and to the building on the far side.
Be aware that performance can, and will, vary in other scenes and locations (up to a 20% performance range), but this scene gives us a good-enough sense of the type of performance you’ll see with the various CPUs.
Our test systems easily outweigh the recommended minimum system requirements for Cyberpunk 2077. We outfitted our test systems with the Gigabyte GeForce RTX 3090 Eagle to reduce the GPU bottleneck as much as possible. Be aware that you’ll see a smaller performance delta with lesser graphics cards. If you’re already shackled by a GPU bottleneck, you might not see appreciable gains with the same processors we’ve tested below.
We strove to remove memory as a potential bottleneck, too, so we equipped our tests system with 32GB of DDR4 memory spread across two dual-rank DIMMs. We assigned the memory to operate at each respective processor’s stock specifications, but be aware that additional memory tuning/overclocking could yield higher performance results. Likewise, single-rank memory kits could also result in less performance. As we noted in our recent look at how Cyberpunk 2077 runs with varying memory kits, you can definitely run the game with 8GB of memory, and 16GB should be sufficient, but using two dual-rank DIMMs (or four single-rank DIMMs) can improve performance. If you’re not running a top-end GPU, memory capacity probably won’t matter much.
You can see a further breakdown of our test system at the end of the article.
Cyberpunk 2077 CPU Scaling RT Ultra Settings Benchmarks
We included the highest-end models of the 1000, 2000, 3000, and 5000 series Ryzen processors. We’ve also included the Core i9, i7, i5, and i3 models from Intel’s 9th- and 10th-generation processors. We’ll add testing with HEDT models after we assess performance impacts from the upcoming 1.05 patch.
Image 1 of 3
(Image credit: Tom’s Hardware)
Image 2 of 3
(Image credit: Tom’s Hardware)
Image 3 of 3
(Image credit: Tom’s Hardware)
Processors marked in the charts with ‘Modified’ have the ‘Ryzen fix’ applied. We tested with three resolutions at various settings, but our first album of results focuses on the RT Ultra preset. This enables DLSS Auto (which uses the Quality mode at 1080p, Balanced mode at 1440p, and Performance mode at 4K), and also engages ray tracing, making it the most demanding preset. After we applied the patch, it also yielded big advances for a few of the AMD Ryzen processors.
AMD’s Ryzen 9 5950X takes the lead at 1080p RT Ultra, trailed closely by the Ryzen 7 5800X, 5900X, and Core i9-10900K. You’ll notice that these top-end processors are closely matched — most users would be hard-pressed to notice a difference between these CPUs.
The eight-core 16-thread Ryzen 7 5800X only makes its way into the upper tier of the chart after we apply the fix. The performance improvement is also apparent with the eight-core Ryzen 7 3800XT, too, casting some doubt on CD Projekt Red’s claims that the ‘Ryzen bug’ doesn’t impact AMD’s eight-core processors.
Stepping down the chart, we see the Core i7-10700K trailing the chart-topping Ryzen 9 5950X by 9%. We also see a substantial gap between the 9th and 10th gen Intel processors, with the fastest 9th gen processor, the Core i9-9900K, trailing the Core i9-10900K by 17% and the Core i7-10700K by 12%.
A quick look at the bottom of the chart shows how far AMD has come since it’s Zen 1 and Zen+ processors. The Ryzen 7 1800X falls behind Intel’s current i3-10100 despite having twice as many cores and threads, and the Ryzen 7 2700X only matches the Core i5-9600K. The SMT bug still affects those Ryzen CPUs, but using the modified code made the game unstable so it’s not currently a solution.
Flipping through the album to the QHD results shows that the deltas between the chips shrink tremendously as we increase resolution. Bear in mind that we’re testing with the ridiculously-priced RTX 3090, so the difference between these chips will shrink dramatically with lesser GPUs. The 4K resolution, and the resulting GPU bottleneck, stands as the great equalizer, though, dragging all of the chips (except the Ryzen 3 1300X) into closer competition.
Cyberpunk 2077 CPU Scaling Medium Preset Benchmarks
Image 1 of 3
(Image credit: Tom’s Hardware)
Image 2 of 3
(Image credit: Tom’s Hardware)
Image 3 of 3
(Image credit: Tom’s Hardware)
Dropping all the way down to the medium preset, maximum performance is much higher and CPU bottlenecks continue to play a significant role. This preset features lower fidelity settings than the Ultra preset below, so we see more substantial deltas between the processors.
These standard rasterization tests stress the processor differently than the RT Ultra settings above, and at 1080p, we see the ‘patched’ Ryzen 7 5800X take a slight lead over the rest of the test pool. However, we see several of the same AMD contenders appear in the top five, with the Ryzen 9 5900X and Ryzen 9 5950X landing within a few percent of the leader.
AMD’s core-heavy Ryzen 3000 models are impressive in this series of tests, but the Ryzen 5 5600X, our go-to recommendation for gamers, holds its own with 106.5 fps, basically matching the more expensive Core i7-10700K. Given the mostly-reasonable pricing for both the 5600X and the 10700K, it’s a tough sell to step up to significantly more powerful chips from either camp, at least if you’re looking for the best bang for your buck.
On the opposite end of the spectrum, the first-gen Ryzen processors struggle to keep pace in the 1080p tests, and hitches and stuttering were outwardly visible during testing. If you’re playing at this low of a resolution, a second-gen or newer Ryzen processor is a better fit.
Again, it’s readily apparent that the magnitude of the performance gains relies heavily upon how close the system is to a GPU bottleneck. Performance deltas steadily decline as we increase the resolution; however, unlike with the RT Ultra settings in the previous series of 4K tests, we see a bit more differentiation between the processors at the highest resolution. Anything below a Ryzen 7 2700X will introduce visible stuttering and hitching, and the Ryzen 3 1300X is a very rough experience. (We’d call it unplayable, but some people have lower standards.)
Cyberpunk 2077 CPU Scaling Ultra Preset Benchmarks
Image 1 of 3
(Image credit: Tom’s Hardware)
Image 2 of 3
(Image credit: Tom’s Hardware)
Image 3 of 3
(Image credit: Tom’s Hardware)
Finally, we turn the dial up to the Ultra preset, but without ray tracing active. As expected, the gains again shrink as we increase resolution and push closer to a graphics bottleneck that makes CPU performance less of a factor.
Again, the Ryzen 9 5950X, 5900X, Ryzen 7 5800X, and 10900K all push the RTX 3090 to the limits during the 1080p tests, meaning we’ll have to wait for faster GPUs to see any meaningful difference between these chips at these resolution and fidelity settings. We don’t see any massive changes to the performance hierarchy with these settings — the chips largely stack up the same as we saw with the 1080p medium preset, albeit with some differences in average framerates.
It’s interesting to look at some of the other differences that are present thanks to the memory configurations we’re using for testing. For example, Jarred tested the Core i9-9900K at DDR4-3600 with 16-18-18 timings, and performance was around 14% higher than the i9-9900K running at official Intel settings (DDR4-2666), even with tighter 14-14-14 timings. AMD’s Zen 2 and Zen 3 chips have an advantage in that they officially support up to DDR4-3200, without overclocking.
Many enthusiasts at a bare minimum enable the XMP profile in the BIOS (which might be called A-XMP, AMP, DOHC, or some other name on AMD platforms). With better memory, this can definitely improve performance, but it technically voids the warranty.
Cyberpunk 2077 CPU Scaling Thoughts
Cyberpunk 2077 is one of the most demanding games on the market, and it’s pretty much impossible to devise one series of tests that can give a definitive single answer for how your CPU will perform in the massive number of possible hardware configurations. As such, you should view these tests as basic indicators of how the processors stack up when paired with the highest-end gaming GPU that money can buy, and then adjust your expectations based on your GPU.
Naturally, the differences between our test subjects will diminish rapidly with lesser graphics cards. Still, if you’re determined enough, you could even scrape by with integrated graphics with below-bare-minimum settings. Not that we’d recommend that to anyone, but the overriding point is that with some time and patience you can adjust the fidelity settings to make do with a slower GPU.
On the flip side of the coin, playing the game with a capable GPU can be a punishing experience for your processor. Today’s newest high-end chips, namely the Ryzen 9 and 7 5000 series and 10th-gen Intel i9 and i7 chips, are obviously the best fit if money is no object and you’re trying to squeeze out every last frame. However, if you’re looking for the best blend of price and performance, mid-range modern CPUs, like Intel’s Core i5 and AMD’s Ryzen 5, still offer the best blend of performance and price.
Even the previous-gen Core i5-9600K and Ryzen 5 3600XT held up relatively well during our testing, but those should be considered a bare minimum if you’re looking to game with higher-end GPUs.
Be sure to hit our article covering the recommended system specifications for different price points, and our expansive Cyberpunk 2077 testing with a wide spate of GPUs to see where your system could slot in based on your hardware configuration.
In the meantime, we’re eagerly awaiting the Cyberpunk 2077 1.05 patch to see how much it impacts overall game performance. We’re also working to add a few HEDT models to the tests above, so stay tuned.
In a few weeks, AMD will present a new generation of APU processors for laptops. This year, the manufacturer introduced a third series called Renoir, which is equipped with up to eight Zen 2 cores. Next year, the company will launch its assault with Cezanne-U and Cezanne-H chips, which will also use up to 8 cores. This time, however, they will be greatly improved Zen 3 cores with much higher single-threaded performance. The Cezanne series is just the beginning, because we have learned more details about the APU laptop premiere schedule for the years 2021 – 2022. We learned that the manufacturer plans at least several different series of processors depending on the destination and energy budget. We have also obtained confirmation that the first mobile APU with the RDNA 2 graphics chip will go to laptops next year.
We got to know the detailed publishing plan of AMD Ryzen APUs for laptops for years 2021 – 2022. The company is working on the series Cezanne, Lucienne, Van Gogh, Rembrandt, Barcelo, DragonCrest and Pollock.
In the next two years, AMD intends to introduce APU processors with the following TDP factors (for laptops): 45 W, 35 W (special variants of normal 45 W), 15 W, 9 W and 4.5 W. A total of 4 APU series for notebooks are planned for next year: Cezanne-H (45 W, 35 W), Cezanne-U (15, Lucienne-U (15 W), Van Gogh (9 W) and Pollock ( 4.5 W). Only Cezanne processors next year will use the new Zen 3 cores, while only Van Gogh will offer an integrated graphics chip based on the RDNA 2 architecture. In all other cases it will be the well-known and polished Vega.
As you can see, Van Gogh APU in addition to RDNA graphics chips 2 will also introduce support for LPDDR5 memory, most likely 5400 MHz. Scheduled APUs for the 2022 year will be even more varied. The main family of Rembrandt-H and Rembrandt-U will use the improved Zen 3+ cores in TSMC 6nm lithography. Here, too, the Vega architecture will be abandoned in favor of RDNA 2. There will also be support for DDR5 and LPDDR5 memory. In 2022, the Barcelo-U series will also be introduced alongside Rembrandt-U, replacing Lucienne-U. Barcelo will be based on Zen 3 cores and Vega graphics chips – thus the differences between Barcelo and Rembrandt will be even greater than between Cezanne and Lucienne. Barcelo will also use 7 nm lithography, not 6 nm. The Van Gogh family in 2022 will in turn be replaced by the DragonCrest APU, which will basically only be a minor improvement over Van Gogh . These processors will have a lower TDP of 9 W. For the next 2 years, the 4.5 W chips will be dominated only by Pollock in lithography 14 nm.
Ampere Computing is moving purposefully in the direction of a CPU alternative for data center use. The Ampere Altra is based on the Neoverse-N1 design from ARM and offers 80 ARM v8.2 + cores in one socket. Up to 80 cores in a power budget of 250 W should be offered to established manufacturers such as Intel and AMD compete. As Altra Max 2021 even variants with 80 cores should follow.
So far, however, there have only been many announcements and few tangible ones, even if Ampere naturally presented its own benchmarks and wanted to emphasize, among other things, the advantages of a stable clock over dynamic boost clock rates. Now Anandtech presents a first detailed and independent test of two Altra Q 80 – 33 (80 cores with 3.3 GHz) in a dual socket system (Mount Jade). The processor in the test system still includes 256 GB of RAM per socket – in eight DIMMs per socket in order to fully utilize the 8-channel memory interface.
The opponents of the two Altra Q 80 – 33 were two AMD EPYC 7742 each with 64 cores with a TDP of 225 W and two Intel Xeon Platinum 8280 each with 28 Cores at 205 W.
The processors and the system leave a different picture in the benchmarks. In the single-threaded test using Spec 2017 integer, the Altra Q intersects 80 – 33 very good – is on a level with the Xeon Platinum 8280 and ahead of the EPYC 7742. In the floating point area, however, the established processors are stronger. It looks much better if the Altra Q 80 – 33 can use all of its 80 cores.
The ARM processor is really good in the LLVM compiling and the NAMD performance. This is where the old Q 80 – 33 in 1S and 2S configuration on par with one or two EPYC 7742. In the chosen benchmark in 1S configuration, these two processors are even faster than two Xeon Platinum 8280.
Across all tests it can be said that the Altra Q 80 – 33 not only quickly, but also efficiently are. Although the TDP provides for a power budget of 250 W, the system does not use this for the 3.3 GHz clock rate. This must also be taken into account when assessing performance. Price also plays a role. Ampere calls for the Altra Q 80 – 33 a list price of 4. 050 US dollars, an EPYC 7742 costs around 5 in this country 200 Euro and a Xeon Platinum 8280 at least 8. 600 Euro, albeit one should note here that these prices in the free trade have very little to do with what server providers are offered at reduced prices.
Whether and where amps can be compared with the Altra processors in the data center Segment will be able to establish itself remains to be seen. Hyperscalers like Amazon have already developed their own processors. AMD is nibbling on Intel’s market share, but the chip giant is supposed to strike back 2021 / 22 . First, however, AMD will launch the 3rd generation EPYC processors based on the Zen 3 architecture.
Side 1: ASUS ROG Strix B 510 – XE Gaming WiFi in the short test: hardly better than its predecessor
AMD’s new Ryzen – 5000 – Processors deliver a very good performance and can convince in all disciplines for the first time. So it’s no wonder that the Zen 3 CPUs are in great demand. In order to be well equipped for overclocking, ASUS has the ROG Strix B 510 – E Gaming (Hardwareluxx test) according to replaced with a new refresh model in a short time. A significantly better figure should therefore be the new ROG Strix B 510 – Make XE Gaming WiFi. In this short test, we want to determine whether this is indeed the case and where the differences lie.
It is not the first time that ASUS replaced a mainboard with a revised model after a relatively short time. In the recent past, this also affected the ROG Zenith II Extreme (Hardwareluxx test), which is known to have been replaced by the ROG Zenith II Extreme Alpha (Hardwareluxx short test). The same fate now overtakes the ROG Strix B 510 – E Gaming as previous B 550 – ASUS flagship. The heir to the throne is the ROG Strix B 550 – XE Gaming WiFi.
However, it should be known in advance that the differences are in the details and must be looked for with a magnifying glass. In any case, ASUS did not make any changes to the connections, but rather revised the CPU power supply. It will be interesting to see which voltage converters ASUS are now using for the ROG Strix B 550 – XE Gaming WiFi has provided.
Visually there are no major differences, but rather details -Improvements. On the one hand, ASUS designed the I / O panel cover differently. Furthermore, the manufacturer no longer relies on two separate VRM coolers, but has connected the now slightly larger heat sinks with a heat pipe.
These are the specifications of the ASUS ROG Strix B 550 – XE Gaming WiFi:
The data of the ASUS ROG Strix B 550 – XE Gaming WiFi in the overview
Mainboard format
ATX
Manufacturer and Designation
ASUS ROG Strix B 510 – XE Gaming WiFi
CPU socket
AM4 (for Matisse)
Power connections
1x 24 – ATX pin 1x 8-Pin EPS 12 V 1x 4-pin + 12 V
phases /Do the washing up
17 (16 x CPU, 1x RAM)
Price
292 Euro
Website
ASUS ROG Strix B 550 – XE Gaming WiFi
Southbridge / CPU features
Chipset
AMD B 550 Chipset
Memory banks and type
4x DDR4 (dual-channel), max. 5. 100 MHz
Memory expansion
Max. 100 GB RAM UDIMM, ECC support
SLI / CrossFire
SLI (2-Way), CrossFireX (3-Way)
Onboard features
PCI-Express
2x PCIe 4.0 x 16 (electrical with x 16 / x8) via CPU 1x PCIe 3.0 x 16 (electrical with x4) via AMD B 550 2x PCIe 3.0 x1 via AMD B 550
PCI
–
Storage
6x SATA 6 GBit / s about AMD B 550 1x M.2 with PCIe 4.0 x4 via CPU (M-Key, PCIe 4.0 x4) 1x M.2 with PCIe 3.0 x4 via AMD B 550 (M-Key)
USB
CPU: 3x USB 3.2 Gen2 (10 Gbit / s, 3x external) via CPU (Ryzen 3000)
Chipset: 1x USB 3.2 Gen2 (internal) via AMD B 550 2x USB 3.2 Gen1 (5 GBit / s , 2x internal) via AMD B 550 9x USB 2.0 (5x external, 4x internal, 1x Type-C audio) via AMD B 550
Graphic interfaces
1x DisplayPort 1.2 1x HDMI 2.1
WLAN / Bluetooth
WiFi 802. 11 a / b / g / n / ac / ax over Intel Wi-Fi 6 AX 200, dual band, max . 2.4 GBit / s, Bluetooth 5.1
1x 4-pin AIO pump header 3x 4-pin chassis FAN header
LED lighting
I / O panel cover, chipset cooler 2x 4 -Pin RGB header 2x 3-pin ARGB header
Onboard comfort
Status LEDs, Debug LED, CMOS Clear Jumper, BIOS Flashback Button
The Supplied accessories
the following We could find accessories in the packaging:
Mainboard manual including driver and software DVD
four SATA cables
2T2R-WLAN-Ant enne
seven cable ties
ROG sticker
RGB extension cable
ARGB extension cable
five M. 2 screws and thread including three M.2 rubber spacers
USB Type-C audio cable
ROG key fob
ROG thank you card
Hyper-M.2-x 16 – Gen4 additional card
The accessories were also from ROG Strix B 550 – E Gaming adopted almost 1: 1. The only exception is the expansion card, on which four additional M.2 M-Key SSDs can be installed and which can each be connected with up to PCIe 4.0 x4. ASUS also supplies additional M.2 screws for this. The Hyper-M.2-x 16 – Gen4 card is, as the name suggests, complete with 14 Gen4 lanes are connected and thus provides the necessary bandwidth to run four PCIe To fire 4.0 x4 M.2 SSDs. A small radial fan ensures the necessary ventilation.
For this, however, the 16 CPU lanes are completely sacrificed, so that no more lanes for a potent graphics card would be left. Only over the B 550 – chipset could be in the lowest x 16 – slot for a graphics card with PCIe 3.0 x4 be connected. This is easily sufficient for pure image output, but less for games. Another option would be to have the M.2 expansion card in the top or middle x 16 – to leave the slot and the graphics card in the other CPU-PCIe-x 16 – slot to be left. In this case, however, the 16 CPU lanes divided in x8 / x8 mode, which is precisely why the M.2 expansion card is severely reduced in performance.
As previously indicated , ASUS has significantly revised the VRM cooling, and not without reason. Instead of using two individual coolers, as in the non-XE model, ASUS not only enlarged the two coolers in the XE version, but also connected them with a heat pipe. This should distribute the heat better and the cooling should generally be better. But ASUS also left a little bit alone 30 – mm axial fan that is supposed to improve the cooling.
Unfortunately, the fan cannot be manually operated in the BIOS control, but only switch it on or off, which we think is a shame. The good news, however, is that the fan never started working. If you want to avoid the risk in general, simply deactivate the fan in the BIOS.
Now we come to the biggest difference between the E and XE version. The number is with 16 CPU performance levels are identical, but ASUS has the voltage converters against 90 – A models replaced while the ROG Strix B 550 -E Gaming 70 A converter to had to offer. On the voltage converters, which come from Texas Instruments, the model designation “X 95410 “to be recognized. In general, however, it is with the effective 12 + 2 configuration remained, but through the 90 – A voltage converter is offered by the ASUS ROG Strix B 550 -XE Gaming WiFi on the theoretical side more power for extended CPU overclocking.
When it comes to RAM overclocking, ASUS has the ROG Strix B 550 – E Gaming with up to effectively 4. 600 MHz excellent, for the new ROG Strix B 550 – XE Gaming WiFi is ASUS even up to effectively 5th 100 MH z up. However, it depends very much on the choice of DIMM whether you can penetrate into such areas.
<> ASUS ROG Strix B 550 – XE Gaming WiFi in the short test: hardly better than its predecessor Overclocking
Already 12 January NVIDIA during The online conference will showcase the new generation of GeForce RTX graphics cards 3000 for laptops. No wonder that manufacturers have been working on refreshed or completely new designs that will use the potential of Ampere architecture for some time. Information about one of the novelties of the Taiwanese company MSI, which will show at least one novelty in January, has leaked to the network. I am talking about the GS laptop 76 Stealth, which will be the successor of the GS model 75 Stealth, the device introduced on the occasion of the premiere of GeForce RTX cards 2000. The new notebook will undergo a major visual makeover, as will last year’s GS 66 Stealth. What can we expect from a notebook?
MSI at CES 2021 will present a new gaming GS laptop 76 Stealth. It will be the successor to the GS model 75, which will be equipped with NVIDIA GeForce RTX cards 3000 and 360 Hz IPS screen.
We got to know the first details about the MSI GS laptop 76 Stealth, which will be presented at CES 2021. The laptop will receive a similar design to this year’s GS model 66 Stealth. Strong, black colors and better quality. Looking at the photos, we can see a full-size keyboard, of which the numeric block will be more distinct from the rest of the keyboard. New MSI GS Laptop 76 Stealth comes with 17, 3 “Full HD IPS screen with refresh rate 360 Hz. We can confirm that more such devices will be presented in January.
New MSI GS Laptop 76 Stealth with GeForce RTX cards 3000 and screen 360 Hz.
In terms of performance, we expect a big jump compared to the current generation. The hardware will be powered by NVIDIA GeForce RTX 3060, GeForce RTX 3070 and GeForce RTX 3080. The approach to the processor seems disappointing – the Taiwanese manufacturer will most likely decide on the systems 10 Intel Comet Lake generation. In the case of this manufacturer I would not expect a sudden turn towards no AMD Cezanne-H units with Zen 3 cores. Nevertheless, the laptop as a whole seems to be very interesting anyway. For sure, however, we expect very high prices, which the company has already got used to when it comes to slim notebooks from the Stealth line.
The first performance data of the AMD Ryzen Zen 3 for laptops have already been leaked, specifically the Ryzen 7 model 5800 H, showing more than one 20% increase in IPC compared to Zen 2 laptops.
While AMD Ryzen 4000 are a great option for laptops given their high performance and low consumption, AMD already has the following prepared, based on the Zen 3 architecture and with a large increase in the CPI. We have already seen this architecture in desktop processors, for example with the Ryzen 9 5950 X, and there is no doubt that It is a more than important leap on the part of AMD.
Through a leak of TUM_APISAK on Twitter, we have been able to see the performance of an AMD Ryzen 7 5800 H in the Geekbench 5 benchmark. The result obtained by this processor for laptops is 1475 points for single core and of 7630 points for multi core. Of note is the fact that this processor is expected to have 8 cores and 16 threads, as previously filtered.
The leak points to more than 20% increase in CPI compared to Ryzen 4000
Comparing it with those of the previous generation, for example with the Ryzen 7 4800 H (scoring 1194 in single core), means a increase of the 23% in CPI . Of course, this is still a leak and we are not sure if it is true, but if it is, we could see highly capable notebooks soon.
End of Article. Tell us something in the Comments or come to our Forum!
Pablo López
With 15 years I started doing I overclocked my PC to get every extra FPS I could in gaming and scratch a few milliseconds in SuperPi, all the while relentlessly posting about hardware on the Geeknetic forum as a user and reader. They were probably so fed up with continually reading me on the forum that I became part of the writing team, where I continue to report on the latest in technology. Astrophysics and PC games are the hobbies that, after hardware, cover most of my free time.
In less than a month, AMD will officially present new laptop processors. I am talking about the Ryzen series 5000, which will include low-voltage Cezanne-U and Lucienne-U units, as well as full-voltage Cezanne-H systems. Processors from the Cezanne family will use Zen 3 cores, which have been significantly improved mainly in terms of single-threaded performance. We already know exactly what processors the manufacturer is preparing, both for the Lucienne and Cezanne series. We expect not only the continuation of the idea for the Cezanne-H processor line, which will have a reduced TDP to 35 W (so-called HS) . A new addition will be the top Ryzen 9 5900 HX, characterized by, among others unlocked multiplier. Meanwhile, Acer is quietly working on a new version of the Nitro 5 laptop, which will use an 8-core and 16 – a threaded AMD chip Ryzen 7 5800 H.
A new version of the Acer Nitro 5 laptop with AMD Ryzen 7 processor has been tested in the GeekBench database 5800 H. The Cezanne-H APU is 35% better than its predecessor in the single thread test.
AMD Ryzen 7 5800 H will be one of the most powerful Cezanne-H APUs, ready for 2021 year. It has 8 cores and 16 threads. According to GeekBench, the base clock is 3.2 GHz. For comparison, AMD Ryzen 7 4800 H had a base clock of 2.9 GHz, Ryzen 9 4900 HS was clocked at 3 GHz, while the top Ryzen 9 4900 H clocked at 3.3 GHz. According to the same program, the turbo clock speed is expected to reach 4.4 GHz. Ryzen 9 5900 HS (4.5 GHz) and Ryzen 9 5900 HX (4 , 6 GHz + OC possibility). The new APU will also receive twice as much L3 cache – instead of 8 MB we will get 16 MB. One of the vendors currently working on laptops with AMD Cezanne-H is Acer. GeekBench reveals the presence of a new version of Acer Nitro 5.
In the test single-threaded, AMD Ryzen 7 5800 H scored 1475 points, which is on average 20 – 35% better result (it all depends on specific laptops) than the predecessor AMD Ryzen 7 4800 H. Anyway, the increase in single-threaded performance should be visible to the naked eye during specific tests and everyday use. It is no accident that the new Cezanne-H APUs will also be combined with top NVIDIA GeForce RTX cards 3000. In the multithreaded test the result is 7630 points, which is on average 10 – 15% higher result compared to its predecessor. Therefore, another powerful processor from AMD for laptops is being prepared.
Memory module manufacturer Teamgroup has announced that from the third quarter 2021 first DDR5 Wanting to deliver kits to end customers. The period is agreed with the two chip manufacturers AMD and Intel. Means: Platforms that use DDR5 modules should reach the market at this point.
According to their own statements, Teamgroup is currently testing the modules with the mainboard manufacturers ASRock, Asus, Gigabyte and MSI – a common procedure, especially before the introduction of a new memory generation. DDR5 4800 modules that start with 2022 MHz and an operating voltage of 1.1 volts. For DDR4-SDRAM, 2400 MHz (DDR4 – 4800 ) only possible with strong overclocking and significantly increased voltage beyond the JEDEC specification of 1.2 volts. Memory manufacturer SK Hynix had previously only announced DDR5 bars for servers.
End customer DDR5 platform It is questionable which processors should be able to handle DDR5-SDRAM in summer 2021. At Intel, Alder Lake-S – presumably as Core i – 12000 – for the CPU version LGA 1700 appear with DDR5. However, the previous generation Rocket Lake-S (Core i – 11000) comes in the first quarter 2021 and a replacement half a year later seems unlikely.
AMD usually takes a little over a year for a new generation of processors, with which Zen 4 stands for end 1700 or beginning 2022 would have to be applied. The website Videocardz had in August 2021 an alleged AMD roadmap is shown, according to which 1700 the Zen 3 refresh “Warhol” should appear. A transitional generation with the previous Zen 3 chip sets, but with a new I / O die including DDR5 controller and AM5 platform, would be quite conceivable. AMD could concentrate on the new CPU version and DDR5-SDRAM without having to worry about Zen 4.
AMD Ryzen 5000 undervolten: This is how the CPU becomes quieter and cooler The CPU knows its limit Less performance, high consumption Undervolting with offset Undervolten with the Curve Optimzier Easier, economical and quiet: The eco mode Conclusion table Of 20 on 90 degrees in a few seconds despite good cooling: under full load, some Ryzen 5000 as a veritable heating plate. Even if AMD regards these high temperatures as safe and the processors meticulously adhere to this maximum limit, in the end it means at least louder fans and, in the worst case, loss of performance due to throttling.
Ours, with still moderate 105 watts TDP specified, Ryzen 7 5800 X does not even get a water cooling confidently captured as soon as maximum performance is required from all eight cores. This is surprising insofar as the Ryzen processors based on the Zen 3 architecture are quite efficient and should run as close to their optimum as possible without user intervention. But the repeatedly unmistakable fan noise raises the question: Can AMD’s tuning be improved?
The Ryzen processors from AMD are inherently fast and efficient, but with a little experience you can squeeze a little more performance out of them.
In order to get a cooler processor without loss of performance, we concentrate particularly on the so-called undervolting: The CPU is operated with less voltage, which results in lower power requirements and thus should lead to less waste heat. Our test system consists of the Ryzen 7 5800 X and the mainboard Gigabyte X 570 Aorus Master. Its beta BIOS F 31 l already offers the “Curve Optimizer” promised by AMD, which, among other things, should make it easier to adjust the voltage.
Access to all contents of heise + exclusive tests, advice & background: independent , critically founded c’t, iX, Technology Review, Mac & i, Make, c’t read photography directly in the browser register once – read on all devices – can be canceled monthly first month free, then monthly 9 , 95 € Weekly newsletter with personal reading recommendations from the editor-in-chief Start FREE month Now Start FREE month heise + already subscribed?
Sign in and read Register now and read the article immediately More information about heise +
We use cookies on our website to give you the most relevant experience. By clicking “Accept”, you consent to the use of ALL the cookies.
This website uses cookies to improve your experience while you navigate through the website. Out of these, the cookies that are categorized as necessary are stored on your browser as they are essential for the working of basic functionalities of the website. We also use third-party cookies that help us analyze and understand how you use this website. These cookies will be stored in your browser only with your consent. You also have the option to opt-out of these cookies. But opting out of some of these cookies may affect your browsing experience.
Necessary cookies are absolutely essential for the website to function properly. This category only includes cookies that ensures basic functionalities and security features of the website. These cookies do not store any personal information.
Any cookies that may not be particularly necessary for the website to function and is used specifically to collect user personal data via analytics, ads, other embedded contents are termed as non-necessary cookies. It is mandatory to procure user consent prior to running these cookies on your website.
