Welcome to SysReqs

CPU data, benchmarks, and hardware news

Leaked Slides Reference AMD Zen 5 Microarchitecture

It is said that the next-generation "Zen 5" microarchitecture was talked about in a couple of slides that were leaked from AMD's private presentation. Inside the company, the "Zen 5" core with better performance is called "Nirvana," and the CPU core die built on "Nirvana" cores is codenamed "Eldora." The company will use these CCDs to make either Ryzen "Granite Ridge" PC processors or EPYC "Turin" server processors. Next-generation mobile CPUs from the company could also use these cores as part of heterogeneous CCXs (CPU core complexes), along with "Zen 5c" low-power cores.

Broadly speaking, AMD says that "Zen 5" will have a 10% to 15% higher IPC than "Zen 4." The core will have a 48 KB L1D cache instead of the 32 KB one it has now. When it comes to the core itself, it has an 8-wide dispatch from the micro-op queue instead of 6-wide dispatch in "Zen 4." The integer execution step will have 6 ALUs instead of 4. It is possible for the floating point unit to work with FP-512. AMD has raised the highest number of cores per CCX from 8 to 16, which is probably the most important news. At this point, we don't know if it means that the "Eldora" CCD will have 16 cores or that the "Zen 5c" cloud-specific CCD will have 16 cores in a single CCX instead of two CCXs with smaller L3 caches. When AMD makes "Eldora," they use the TSMC 4 nm EUV node. The mobile processor built on "Zen 5" might use the more advanced TSMC 3 nm EUV node.

AMD also uses an interesting way to talk about its CPU core designs on the first slide. This means that "Zen 3" and "Zen 5" are new cores and "Zen 4" and the upcoming "Zen 6" are leveraged cores. If you remember, "Zen 3" gave a huge 19% IPC boost over "Zen 2," which helped AMD take over the CPU market. Even though the "Zen 5" core is only expected to improve IPC by 15% over "Zen 4," it is still expected to have a big effect on AMD's ability to compete.

AMD is looking forward to a 10% rise in instruction per cycle (IPC) over "Zen 5," new FP16 features for the core, and a 32-core CCX (maximum core count) for the "Zen 6" microarchitecture and "Morpheus" core. This would lead to a second round of big growth in the number of CPU cores.

We can see that AMD added an even more advanced branch prediction unit to the "Zen 5" core when we look more closely. Don't forget that changes to branch predictors were the main reason for the generational IPC gain of "Zen 4." The new branch predictor can do zero bubble conditional branches, is more accurate, and has a bigger BTB (branch target buffer). We already said that the core has a bigger 48-KB L1D cache and an unknown bigger D-TLB. There is more speed in the front-end and load/store stages thanks to dual basic block fetch units, 8-wide op dispatch/rename; Op Fusion, a 50% rise in ALCs, a wider execution window, a better prefetcher, and changes to the CPU core ISA and security. The L2 cache that is set aside for each core stays at 1 MB.

Intel's 14th Gen Core is Meteor Lake-based

Leaked documents did not provide a complete description of what Intel will call its consumer CPU models. However, the newly disclosed information may shed light on the future of the Core series.

Future: 14th Generation Intel Core series is MTL/Meteor Lake in the latest hardware support chart for OpenVPL (Video Processing Library). This codename is exclusively associated with the 13th Generation Core family, and not the impending Raptor Lake refresh for which no details have been provided. This effectively disassociates the 14th Gen series from Raptor Lake Refresh, barring a minor marketing effort down the line.

Intel is expected to use a product naming scheme similar to that of its mobile series for its forthcoming consumer desktop Core series. Anyone anticipating the Raptor Lake Refresh can now look forward to a Core i9-13950K SKU or something very comparable.

No clear intentions for mobile series were included in the leaked roadmap, which only addressed Raptor Lake-S Refresh for PCs. Importantly, Intel will not be replacing the current 700-series platform, therefore the company will have two generations plus one refresh (three waves of SKU launches) for the current LGA1700 platform.

Intel's Raptor Lake Refresh, if all goes according to plan, will debut in the third quarter of 2018. Meanwhile, you shouldn't expect to see a PC version of Meteor Lake anytime soon.

Moore's Law Inventor Gordon Moore Dead at 94

Gordon Moore, co-founder of Intel and the Gordon and Betty Moore Foundation, died on March 24th at the age of 94, according to Intel and the Gordon and Betty Moore Foundation. According to the foundation, he died peacefully on Friday, March 24, 2023, at his home in Hawaii, surrounded by family. In July 1968, Moore and his long-time colleague Robert Noyce founded Intel. Moore was executive vice president until 1975, when he was named president. Moore was appointed chairman of the board and CEO in 1979, positions he held until 1987, when he relinquished the CEO position but remained chairman. Moore was appointed chairman emeritus in 1997 and retired in 2006.

Moore also devoted his time and energy to philanthropy, specifically environmental conservation, science, and patient care improvements. He and his 72-year-old wife founded the Gordon and Betty Moore Foundation, which has donated more than $5.1 billion to charitable causes since its inception in 2000.

Intel Xeon W9-3495X Has an Extreme Overclocking Power Draw of Up to 1,900 Watts

Overclockers may now get their hands on Intel's newest Xeon CPUs, which are built on the Sapphire River uArch architecture. The Intel Xeon W9-3495X has the highest scores ever recorded in the Cinebench R23 and R20 tests, the Y-cruncher, the 3DMark CPU test, and the Geekbench 3 benchmark, as we revealed last week. Now here we are again, with yet another attempt at severe overclocking in the hopes of breaking the world record, and yet again, we have very little information on power usage or what the new SKU is capable of. Elmor, an ASUS overclocker, has overclocked the Intel Xeon W9-3495X CPU to 5.5 GHz on all 56 cores in an attempt to set a new world record. The processor's outstanding power consumption is a more notable feat.

The central processing unit used about 1,900 watts of electricity when fueled by two Superflower Leadex 1,600 watt power supply units. Liquid nitrogen was utilized to bring the temperature of the central processing unit down to a comfortable -95 degrees Celsius, which allowed it to continue to operate normally. For this attempt, an ASUS Pro WS W790E-SAGE SE motherboard was used alongside eight GSKILL Zeta R5 DDR5 R-DIMMs. The CPU scored an astounding 132,220 on Cinebench R23, which is an incredible result. While Elmor's accomplishment is impressive, it falls short of the previous week's world record of 132,484 points, hence the prior week's record still stands. For further explanation, please watch the video.

AMD claims great performance doesn't need a power-hungry CPU

The focus of AMD's webinar is on their newly released and upcoming 3D V-Cache SKUs. Company executives discuss the significance of this introduction, the current status of these SKUs, and the benefits of using AMD's platform as opposed to Intel's. Also, there is a great deal to discuss at this time. There is a lot of manipulation, to be sure, but we saw similar results.

When it comes to gaming performance, AMD Ryzen 7000X3D processors outperform Intel's 13th Generation Core series. According to third-party benchmarks, the top-of-the-line Ryzen 9 7950X3D from AMD outperforms the Intel Core i9-13900K by 3.8% and the Core i9-13900KS by 2.7%. This is despite the fact that the base frequency of the latter Processor is 6.0 GHz. Apparently, the Ryzen 9 7950X3D is in direct competition with the Core i9-13900KS, as shown on the presentation. However, it's likely that the AMD comparisons were conducted before Intel introduced this Special Edition KS SKU, as neither the old nor the new literature in the PowerPoint deck has any direct comparisons to either SKU.

Curiously, AMD now offers three Ryzen 9 SKUs (the 7900X3D, the 7950X, and the 7900X) to compete with Intel's Core i9-13900K. Everyone can find anything suitable here. See if the forthcoming Ryzen 7 7800X3D or Ryzen 9 5900 SKUs aren't a better (and cheaper) option if you're considering an Intel Core i9-12900K or Core i7-13700K.

Another major accomplishment of the 600 platform for which AMD may take credit is its compatibility with the new PCIe standard. When using both memory and a graphics processing unit on the CPU's PCIe lanes, the X670E and B650E motherboard series simply have more PCIe lanes available and experience no bottleneck. But, Intel has a leg up because of its support for faster DDR5 memory. A problem with non-binary memory support (24/48/96GB) also exists, but AMD is said to be hard at work on an AGESA upgrade.

The AMD Ryzen 7000X3D series consumes far less power than the Intel Core i9 series, and this is likely the platform's greatest benefit. In response to Intel's claims that "You need a power-hungry processor to attain top results," AMD offers a graphic demonstrating significantly lower total system power consumption. Intel's equivalent improvement will take more time to materialize.

Up to 24% Faster in Official Ryzen 7 7800X3D Gaming Benchmarks vs. Intel's Core i9-13900K

Now that some official gaming benchmarks have been released for AMD's upcoming 8-core Ryzen 7 7800X3D processor (due out in April), it appears to outperform the Intel Core i9-13900K by as much as 24 percent. In its official lineup, AMD is pitting the Ryzen 7 7800X3D against the Intel Core i7-13700K, while reserving the 16- and 12-core Ryzen 7000X3D variants for the Core i9-13900K and Core i9-13900KS, respectively.

While the Ryzen 9 7950X3D and Ryzen 9 7900X3D are both available as of February 28th, AMD has delayed the release of its 8-core/16-thread Ryzen 7 7800X3D. Given the attractive $449 price tag, this came as quite a shock and was a major letdown. The fact that the Ryzen 7 7800X3D is simply too good to ignore could be a contributing factor, as it could put significant strain on even AMD's own SKUs, let alone Intel's lineup.

While the 12-core and 16-core SKUs are a multi-chip module with two CCDs and feature an asymmetric chiplet design, the Ryzen 7 7800X3D has a rather standard design, with a single 8-core CCD with 3D V-Cache, giving it yet another significant advantage over the rest of the Ryzen 7000X3D series.

The Ryzen 9 7950X3D and Ryzen 7900X3D both have two CCDs, but only one CCD with 3D Vertical Cache; as a result, the 3D Vertical Cache Optimizer Driver is responsible for directing gaming workloads to the CCD with 3D Vertical Cache via dynamic "preferred cores" flagging for the Windows OS scheduler.

In two new slides, AMD compares the performance of the AMD Ryzen 7 7800X3D and the Intel Core 9 13900K in four games: Rainbow Six Siege, Total War: Three Kingdoms, Red Dead Redemption 2, and Horizon Zero Dawn. When compared to the Core i9-13900K, the Ryzen 7 7800X3D is superior in all four scenarios by a range of 13–24%. On the second slide, we see how the new AMD Ryzen 7 7800X3D stacks up against the older AMD Ryzen 5800X3D in four popular games: Rainbow Six Siege, Warhammer: Dawn of War III, Counter-Strike: Global Offensive, and Dota 2 (Dota 2 is shown on the third slide).

The Ryzen 7 7800X3D, priced at $449 and able to outperform Intel's $579 priced Core i9-13900K SKU while being $130 cheaper, could be a big winner for AMD and become one of the best sellers if these benchmarks turn out to be even close to painting the realistic picture, as these are just three games handpicked by AMD. AMD is competing with the $405 Core i7-13700K with the Ryzen 7 7800X3D.

These are AMD's chosen benchmarks, so take them with a grain of salt; we'd rather see the official launch on April 6th to verify the claimed performance levels ourselves. Check out our Ryzen 7800X3D preview, which is a performance simulation with a single CCD enabled, in the meantime.

AMD Ryzen 7 7745HX Outperforms High-Priced 13th Gen Core i7 Mobile Processors in Gaming and Battery Life

Notebook makers are starting to take notice of AMD's high core-count Ryzen 7045HX series "Dragon Range" mobile processors, which offer performance and battery-life highly competitive to Intel's 13th Gen Core "Raptor Lake" processors, thanks to AMD's "Zen 4" architecture's unparalleled efficiency at lower power. The Ryzen 7 7745HX has a default TDP of 55 W and 8 cores/16 threads thanks to its single 5 nm "Zen 4" chiplet. There was a comparison between a 7745HX-powered laptop and another with a Core i7-13700HX 8P+8E in a review by the Chinese technology publication Golden Pig Upgrade. As compared to "Raptor Lake," which is hindered by aggressive power-management and an older 10 nm-class technology node, the reviewer noted that the 7745HX offered greater performance/Watt and gaming performance that either matched or surpassed it.

A "Celestial" Xe3 iGPU will be built into an Intel "Panther Lake" processor.

"Panther Lake" is the codename for Intel's 17th Generation Core processors, which will be available in 2026-27. It is the successor to the 16th Generation "Lunar Lake" (2025-26), 15th Generation "Arrow Lake" (2024-25), and 14th Generation "Meteor Lake" (2023-24) architectures. While little is known about "Panther Lake," the first piece of information discovered on the LinkedIn profile page of one Intel Graphics engineer suggests that the processor's graphics tile will feature an iGPU based on the Xe3 "Celestial" graphics architecture, which is two generations ahead of the current Xe "Alchemist" and one generation ahead of the Xe2 "Battlemage."

In their applications, Intel's graphics architectures will remain highly scalable and modular, with variants ranging from low-power iGPUs to large client discrete GPUs and very-large HPC-AI processors. The iGPU powering "Panther Lake" will be Xe3-LPG, a highly pared-down version of the architecture for lower Xe Core counts, with just the right hardware to operate in power-constrained devices like mobile processors. According to an older company slide detailing the scalability of "Celestial," Intel will stick with the disaggregated chiplet design for its processor architectures all the way down to "Panther Lake," as an older company slide detailing the scalability of "Meteor Lake" highlighted a "next platform" processor succeeding "Meteor Lake" and its immediate successor ("Arrow Lake").

AMD talks about how the new 3D V-Cache is better than the old one

AMD began mixing nodes in 2019 when it used the 7 nm node for the Zen 2 microarchitecture's core complex die (CCD) and the 12 nm node for the IO die. AMD recently confirmed to Tom's Hardware that Zen 4 will use three nodes: 5 nm for the CCD, 6 nm for the IO die, and 7 nm for the V-Cache.

During its recent ISSCC presentation, AMD discussed some of the challenges it faced when stacking one node on top of another. To allow them to be connected, both the 7950X3D and the original 5800X3D have their V-Caches positioned over their regular L3 caches. The configuration also keeps the V-Cache away from the heat generated by the cores. While the V-Cache neatly fits over the L3 cache in the 5800X3D, it overlaps with the L2 caches on the core edges in the 7950X3D.

Part of the issue stemmed from AMD's decision to double the amount of L2 cache in each core from 0.5 MB in Zen 3 to 1 MB in Zen 4. However, it circumvented the additional space constraints by drilling holes through the L2 caches for the through-silicon vias (TSVs) that deliver power to the V-Cache. The signal TSVs are still generated by the controller in the center of the CCD, but AMD has optimized them to reduce their footprint by 50%.

AMD reduced the size of the V-Cache from 41 mm2 to 36 mm2 while keeping the same 4.7 B transistors. TSMC manufactures the cache on a new 7 nm node developed specifically for SRAM. As a result, despite the CCD being manufactured on the much smaller 5 nm node, the V-Cache has 32% more transistors per square millimeter.

All of AMD's improvements and workarounds result in a 25% increase in bandwidth to 2.5 TB/s and an unspecified increase in efficiency. Not bad for nine months of development time between the first and second generations of a supplemental chiplet. Hopefully, it will prove its worth when the Ryzen 7 7800X3D arrives in a month.

Ampere Computing Gaining Ground in the Data Center CPU Landscape

Once upon a time, the data center market was dominated by x86-64 manufacturers AMD and Intel. However, in recent years, companies have begun developing custom Arm-based processors to handle more complex workloads within smaller power envelopes and more efficiently. According to Counterpoint Research, we have the most recent data highlighting a significant new player in the data center world called Ampere Computing. The latest data center revenue share report includes revenue for Intel/AMD x86-64 and AWS/Ampere Arm CPUs. For the first time, we see a third-party company, Ampere Computing, capture up to 1.54% of the total data center market revenue in 2022. Because CPUs are available in off-the-shelf servers from OEMs, enterprises and cloud providers can easily integrate Ampere Altra processors.

Intel, the market's largest player, saw a 70.77% share of total revenue; however, this is a decrease from 2021 data, which stated an 80.71% revenue share in the data center market. This represents a 16% decrease year on year. This decrease is not due to a lack of demand for server processors, as the global data center CPU market revenue fell only 4.4% year on year in 2022, but rather to a high demand for AMD EPYC solutions, with team red capturing 19.84% of the revenue from 2022. This represents a 62% increase over last year's revenue share of 11.74%. AMD is slowly but steadily eating Intel's lunch. Another source of revenue is Amazon Web Services (AWS), which the company has filled with its Graviton CPU offerings based on Arm ISA. AWS Graviton CPUs accounted for 3.16% of market revenue in 2021, a 74% increase from 1.82% in 2021.

Intel Releases Windows Workaround and Patch for Ethernet Stuttering and Disconnects

We reported on issues with Intel's i226-V 2.5 Gbps Ethernet controllers, which are used on a wide range of motherboards with both Intel and AMD CPUs, at the end of January, where users were experiencing stuttering and connection drops. Intel has now issued a workaround and patch for the issue, which affects not only the i226, but also the i225 and the Killer E3100 2.5 Gbps network controller. The workaround turns off the Energy-Efficient Ethernet mode, or EEE as it's also known. EEE is only supposed to activate when an Ethernet connection is idle, and it is claimed to cut power consumption by up to 50%.

However, it appears that EEE kicks in even when the Ethernet connection is active, causing stuttering and connection dropouts. The patch disables EEE for all speeds above 100 Mbps, but EEE can also be disabled manually in Windows device settings. MSI is the first motherboard manufacturer to release an updated driver on its website, but because this isn't a permanent solution to the problem, it's unlikely that we'll hear anything else about it. Intel appears to be working on determining the root cause of the issue, but it's unclear whether it will be possible to patch it in software or if a hardware revision will be required in the end.

Better Performance in Video Games on the AMD Ryzen 9 7900X3D


AMD Ryzen 7000X3D processors with 3D V-Cache optimize performance automatically based on workload. The latest chipset driver will be able to dynamically allocate which core the Windows operating system will use. In other words, if the workload requires cache, the operating system will be altered to prioritize cores having 3D V-Cache. If frequency is more critical, higher frequency cores will be picked instead.

According to the business, this function is controlled by the driver, however it can be adjusted manually in the BIOS. Users will be able to choose between 'cache' and 'frequency' as they see fit.

AMD also says that the same drivers will include the PPM Provisioning File Driver. This feature is especially intended for gaming workloads because it dynamically allocates CCX with 3D V-Cache and parks unused CCX without additional cache. If more threads are required, the CCX without 3D V-Cache will also be enabled.

According to AMD, these features will boost performance depending on the workload, but the primary goal of both technologies is to boost gaming performance.

It's worth noting that these are aimed at Ryzen 9 7000X3D CPUs (7950X/7900X), as only these two of of three disclosed SKUs have an asymmetric chiplet design, with one CCX having 3D V-Cache and the other without. This issue does not occur on the single chiplet 8-core Ryzen 7 7800X3D, which will be available in April.

ASUS Releases AMD Ryzen 7000 Series BIOS Updates with 3D V-Cache

ASUS today announced the availability of a BIOS update for its X670 and X670E motherboards, allowing complete compatibility with the latest AMD Ryzen 7000 Series X3D CPUs with AMD 3D V-Cache technology.

The AMD Ryzen 7 5800X3D CPU's proven gaming prowess has made it a long-standing favorite among PC builders. ASUS motherboards provided those builders with high-performance, feature-rich options for putting together a powerful gaming machine based on one of these CPUs. AMD is now stepping up with two new Ryzen 7000 Series CPUs that feature AMD 3D V-Cache technology: the Ryzen 9 7950X3D and the Ryzen 9 7900X3D. These chips, which combine a 3D vertical cache and all of the benefits of the cutting-edge X670 platform, are ready to take the gaming performance crown.