AMD in its recent interview with TechPowerUp had asserted that 3D Vertical Cache (or 3DV Cache), isn't a one-off technology and that it would be a continual part of its roadmap. In its 2022 Financial Analyst Day presentation, the company confirmed this, by announcing variants of its CPU chiplets that have 3DV Cache, extending to both the upcoming "Zen 4" microarchitecture, and the upcoming "Zen 5," which it unveiled today.

EPYC "Genoa" is codename for the upcoming line of server processors based on the "Zen 4" CPU microarchitecture, with CPU core-counts of up to 96-core/192-thread. These feature the standard "Zen 4" CCD. The company hasn't yet announced the last-level cache (L3 cache) size of the standard "Zen 4" CCD. The company will launch the EPYC "Genoa-X" processor, which much like the EPYC "Milan-X," will incorporate 3DV Cache, with a stacked L3 cache die on top of the chiplet. "Genoa-X" is slated for a 2023 debut.

AMD in its Financial Analyst Day 2022 presentation, unveiled its next-generation "Zen 5" CPU microarchitecture. The company's latest CPU microarchitecture roadmap also confirms that variants of its "Zen 4" CCDs with 3D Vertical Cache (3DV Cache) are very much in the works, and there will be variants of the EPYC "Genoa" processors with 3DV Cache, besides standard ones.

AMD stated that it completed the design goal of the current "Zen 3" architecture, by building it on both 7 nm and 6 nm nodes (the latter being the client "Rembrandt" processor). The new "Zen 4" architecture will debut on the 5 nm node (TSMC N5), and could see a similar optical shrink to the newer 4 nm node somewhere down the line, although AMD wouldn't specify whether it's on the enterprise segment, or client. The next-gen "Zen 5" architecture will debut on 4 nm, and see an optical shrink to 3 nm on some future product.

According to the latest TrendForce statistics, the top ten IC designers worldwide posted a combined revenue of US$39.43 billion in 1Q22, or 44% growth YoY. Qualcomm, NVIDIA, Broadcom ranked in the top three. After the acquisition of Xilinx, AMD surpassed MediaTek in the fourth position. In addition, according to TrendForce tracking of IC design industry trends, revenue generated by Will Semiconductor and Cirrus Logic was enough to be included in the top ten for the first time.

Benefiting from growth performance in handsets and RF front-end divisions in addition to its IoT and automotive divisions in 1Q22, Qualcomm's quarterly revenue reached US$9.55 billion, or 52% growth YoY, ranking number one in the world. The expanded application of GPUs in data centers boosted this portion of NVIDIA's revenue to 45.4%, surpassing the 45% accounted for by its gaming business, combining for a total revenue of US$7.9 billion, or 53% growth YoY. Broadcom's revenue from semiconductor solutions is substantial, including network chips, broadband communication chips, and storage and bridging chips. Its business has maintained stable sales performance, with revenue reaching US$6.11 billion, or 26% growth YoY. After the addition of Xilinx, AMD's revenue reached US$5.89 billion, or 71% growth YoY. However, even excluding Xilinx, due to strong sales in its enterprise, embedded and semi-customized divisions, AMD's own business revenue still hit an all-time high of US$5.33 billion.

The ecosystem of AMD AIBs (add-in board partners) increased this month, when Chinese consumer electronics company Weijian International launched the Soonfoals brand of AMD Radeon graphics cards. The name purportedly indicates fast-moving foals (young horses). From the looks of it, Soonfoals will focus on entry-mainstream SKUs, with its current lineup including cards based on the Radeon RX 6650 XT, RX 6600, RX 6500 XT, and RX 6400. Products include a single-slot, low-profile card based on the RX 6400, and dual-slot, dual-fan cards based on the other GPUs, including a premium-looking brand extension called "Lightning," denoting factory-overclocked RX 6650 XT and RX 6600 cards. From the looks of it, Soonfoals will look to grow locally in the Chinese market first.

ASUS today announced ExpertCenter PN64 Mini PC and ExpertCenter PN52 Mini PC, a pair of mini PCs that offer amazing combinations of performance and aesthetics, including unique ribbed chassis with beveled edges for a look that blends easily into any office or commercial setting. ExpertCenter PN64 and PN52 are standout performers in the 1-liter PC segment, with a compact layout and thermal design to ensure stability for the latest Intel and AMD processors. Measuring just 120 x 130 x 58 mm, each model can support up to four displays and features comprehensive I/O connectivity. In addition, MyASUS software includes Two-Way AI Noise Cancellation to ensure crystal-clear voice communication, smart fan controls, system diagnostics and other functions.

The ExpertCenter PN64 Mini PC is powered by a 12th Gen Intel Core processor and DDR5-4800 RAM. It features one PCIe 4.0 x4 M.2 SSD slot, five USB ports—including a USB 3.2 Gen 2 Type-C port with DisplayPort 1.4 support for attaching one display—plus dual HDMI ports that support up to four displays. There's also one configurable port for efficient working scenarios. For additional flexibility, PN64 is designed to allow quick access to the storage and memory modules for easy upgrades.

ASUS today announced that a selection of its all-in-one (AIO) liquid coolers will be fully compatible with the AM5 (LGA 1718) socket designed for next-gen AMD CPUs. ROG Ryujin, Ryuo and Strix LC coolers will fit via an AM5 mounting kit, while TUF Gaming LC coolers will fit using the existing kits. A list of compatible ASUS coolers is provided in the table below.

It appears that AMD's Zen 4 based CPUs are making their ways into the hands of overclockers and so far at least one has already been delidded. Although we only get to see the IHS itself, it's clearly very thick compared to what we've seen in the past, although it appears to be fairly straightforward to remove, if it wasn't for the fact that the two CCD's and the IOD are soldered to it. Unlike current CPUs, which have a solid seal, the Zen 4 CPUs appear to only have the IHS glued to the CPU packaging in a few spots.

Judging by the looks of the area where the CCDs and the IOD attach to the IHS, this looks like a destructive delidding, although it could just be leftovers from the soldering material. The IHS has clearly been coated with some materials for a good solder interface as well, but this is nothing new, as we've seen this on delidded, soldered CPUs in the past. The person who shared this picture should most likely not have done so and as such, we won't be posting a link to the source.

Seasonic has recently updated their wattage calculator tool to include the unreleased AMD Radeon RX 7000 series however it is currently unknown if the power information comes directly from AMD. The RX 7900 XT and 7800 XT are both listed as requiring a 750 W power while for the RX 7700 XT a 650 W supply is recommended with these values all aligning with those of the corresponding Radeon RX 6000 models. The RDNA 3 architecture of these new graphics cards is expected to increase performance by up to 40% according to recent leaks with AMD confirming they are targeting a 50% performance/watt increase over RDNA 2.

AMD and ATI Technologies ULC have filed a complaint with the United States International Trade Commission (USITC) against Realtek Semiconductor and TCL Industries holdings. The complaint lists five patent infringements from both companies, mostly related to graphics technologies such as texture decompression, a unified shader approach to graphics architectures, a multi-threaded graphics processing system, as well as methods to synchronize thread wavefront data and events and a patent covering a processing unit for asynchronous dispatch.

According to AMD, both companies integrated solutions based on AMD's intellectual property without appropriate, prior licensing. The USITC has already come forward with an investigation announcement towards a number of Realtek and TCL-designed products, including graphics systems, digital televisions, and assorted components, found in some products shipped and sold in the U.S. market. The lawsuit aims for an exclusion order and cease and desist on sales of affected products.

DDR5 memory pricing has been declining faster than expected, with average pricing for modules based on the latest standard dropping by as much as 20% in a month's time. As reported by ComputerBase, pricing for 1 GB of DDR5 has fallen from around €15 by the end of 2021 down to around €5/GB at time of writing. At current pricing, an entry-level, 32 GB DDR5 kit DDR5-4800 memory (JEDEC standard) has fallen from a high of €430 down to a much more palatable €154.

The price decline comes as good news for anyone aiming to upgrade their PC in wake of AMD's Socket AM5 launch for its Zen 4 architecture, which the company has already confirmed will only support the latest RAM standard. AMD itself must be riding the sea of relief, as high DDR5 pricing could significantly shape the company's next-gen platform's value compared to arch-rival Intel, which already offers DDR4 and DDR5 support with its Alder Lake chips. Expectations place the same memory support for the next-gen Raptor Lake platform. Prices for DDR4 memory seem to have hit a bottom, however, as pricing hasn't significantly moved in around six months. Like with all new technologies, expect the price difference to eventually change in favor of DDR5 memory, as manufacturers adjust their outputs towards adoption.

Sapphire formally launched its Radeon 6700 series graphics card. The AMD Radeon 6700 is an odd-ball SKU that doesn't yet feature in the company's retail product stack, but is yet being released to retail by Sapphire. So far we've not come across any other board partner with this SKU. The 6700 is unique in its branding—there's neither "RX" nor "XT" in the model name, it's called simply the "Radeon 6700."

Carved out from the same 7 nm "Navi 22" silicon as the RX 6700 XT and RX 6750 XT; the 6700 has 36 out of 40 compute units enabled, working out to 2,304 stream processors, and 144 TMUs. The card is endowed with 10 GB of 16 Gbps GDDR6 memory across a 160-bit wide memory interface. Sapphire has two cards in its lineup, one is an unnamed base model that sticks to the "reference" specs, and a factory-overclocked Pulse 6700 card.

AGON by AOC - one of the world's leading gaming monitor and IT accessories brands - adds the 27" (68.58 cm) AGON PRO AG274QS, a high-speed, 300 Hz gaming monitor to its AGON PRO portfolio. Equipped with a Fast IPS panel supporting DisplayHDR 600, 1 ms GtG response time and QHD resolution, it hits home with esports players and enthusiasts, especially fans of FPS, battle royale, and similar fast-paced titles.

AGON PRO's current models with Full HD, the 25" AG254FG (360 Hz) or the 27" AG274FZ (260 Hz) demonstrate beautifully how exceptional speed can be combined with high quality IPS panels and award-winning gamer-focussed designs. Now, with the 27" AG274QS, AGON by AOC brings sharper QHD resolution (2560x1440 pixels) up to breath-taking 300 Hz - one of the first QHD displays with such a high refresh rate on the market today.

AMD is likely preparing to launch a new mid-range graphics card SKU positioned between the Radeon RX 6650 XT and the RX 6700 XT, the new RX 6700 (non-XT). It doesn't seem like the RX 6700 is an OEM-exclusive designed to get rid of silicon. Pictures surfaced of a Sapphire branded custom-design card, complete with box-art designed to woo customers in stores. It also doesn't appear to be a China-exclusive SKU, since Sapphire tends to put Chinese-language branding on its box-art, which is missing here.

The RX 6700 is configured with 2,304 stream processors across 36 RDNA2 compute units, out of the 40 physically present on the "Navi 22" silicon. The memory is an interesting piece of specs, with the RX 6700 coming with 10 GB standard—presumably over a 160-bit wide GDDR6 memory bus. This means five 16 Gbit (2 GB) GDDR6 memory chips. The engine clocks are reportedly 2330 MHz game clock, and 2495 MHz boost; and 16 Gbps memory speed. The cards' board-design is standard Sapphire fare, with nothing that stands out from the current RX 6700 XT Pulse and base-model custom cards from the company. One interesting thing to point out, though, is a single 8-pin connector on the base-model custom card (225 W maximum power capability including the PCIe slot), which should put the typical board power around 200 W. The cards are reportedly launching in Europe on June 9, priced around 569€ including taxes.

Jon Peddie Research reports that the global PC-based graphics processor units (GPU) market reached 96 million units in Q1'22 and PC GPUs shipments decreased 6.2% due to disturbances in China, Ukraine, and the pullback from the lockdown elsewhere. However, the fundamentals of the GPU and PC market are solid over the long term, JPR predicts GPUs will have a compound annual growth rate of 6.3% during 2022-2026 and reach an installed base of 3.3 million units at the end of the forecast period. Over the next five years, the penetration of discrete GPUs (dGPU) in the PC market will grow to reach a level of 46%.

AMD's overall market share percentage from last quarter increased 0.7%, Intel's market share decreased by -2.4%, and Nvidia's market share increased 1.69%, as indicated in the following chart.

With the release of the PICO-V2K4, AAEON has gone one step further in pushing the boundaries of innovation. By creating the smallest embedded single board computer to be powered by the AMD RYZEN V2000 embedded processor family, featuring up to 8 cores, AAEON has further reiterated why it is an industry leader in embedded solutions. The PICO-V2K4 also hosts high-performance AMD Radeon graphics and 7 nm processing technology, all on a 3.94" x 2.84" form factor, opening the door to a broader and more sophisticated range of application uses.

The PICO-V2K4 retains the characteristic versatility that AAEON's PICO-ITX board range is known for, but its unprecedented high-speed computing power, groundbreaking graphic capability and targeted I/O introduces a new level of refinement to the edge. Such elements will give users everything from unprecedented CPU performance to rugged deployment capabilities, all without sacrificing graphic sophistication. With exceptional CPU performance courtesy of the V2000 embedded processor family, the PICO-V2K4 has the power to facilitate automated industrial applications, while maintaining longevity and consistency throughout its lifecycle with a thermal design point range of 10~25 W. Tying this together is up to 64 GB NVMe onboard storage and an M.2 2280 M key port, which will enable expansion modules to suit different application types.

Sony today released the v1.0.2 patch for the PC release of "God of War," which adds support for AMD FidelityFX Super Resolution 2.0 (FSR 2.0). The latest performance enhancement by AMD lets you experience even higher image quality at a set performance level, or higher performance at a set quality level. It leverages temporal data to add detail to upscaled images, and in our testing with other games, was found to offer comparable quality and performance to NVIDIA DLSS 2.0. The latest v1.0.2 patch for "God of War" is now up on Steam.

Update 20:57 UTC: We've just posted our God of War FSR 2.0 Review.

There has been a fair bit of confusion about AMD's upcoming AM5 platform when it comes to connectivity and we're going to break things down and explain what the difference is between the B650 and X670 boards. We're also going to cover the processor connectivity, since that's an integral part of any motherboard these days. All the information in this article is based on what we've been told by various sources during Computex.

Supercomputing game has been chasing various barriers over the years. This has included MegaFLOP, GigaFLOP, TeraFLOP, PetaFLOP, and now ExaFLOP computing. Today, we are witnessing for the first time an introduction of an Exascale-level machine contained at Oak Ridge National Laboratory. Called the Frontier, this system is not really new. We have known about its upcoming features for months now. What is new is the fact that it was completed and is successfully running at ORNL's facilities. Based on the HPE Cray EX235a architecture, the system uses 3rd Gen AMD EPYC 64-core processors with a 2 GHz frequency. In total, the system has 8,730,112 cores that work in conjunction with AMD Instinct MI250X GPUs.

As of today's TOP500 supercomputers list, the system is overtaking Fugaku's spot to become the fastest supercomputer on the planet. Delivering a sustained HPL (High-Performance Linpack) score of 1.102 Exaflop/s, it features a 52.23 GigaFLOPs/watt power efficiency rating. In the HPL-AI metric, dedicated to measuring the system's AI capabilities, the Frontier machine can output 6.86 exaFLOPs at reduced precisions. This alone is, of course, not a capable metric for Exascale machines as AI works with INT8/FP16/FP32 formats, while the official results are measured in FP64 double-precision form. Fugaku, the previous number one, scores about 2 ExaFLOPs in HPL-AI while delivering "only" 442 PetaFlop/s in HPL FP64 benchmarks.

AMD's upcoming Radeon RX 7000 series of graphics cards based on the RDNA 3 architecture are supposed to feature next-generation protocols all over the board. Today, according to a patch committed to the Linux kernel, we have information about display output choices AMD will present to consumers in the upcoming products. According to a Twitter user @Kepler_L2, who discovered this patch, we know that AMD will bundle DisplayPort 2.0 technology with UHBR 20 transmission mode. The UHBR 20 standard can provide a maximum of 80 Gbps bi-directional bandwidth, representing the highest bandwidth in a display output connector currently available. With this technology, a sample RDNA 3 GPU could display 16K resolution with Display Stream Compression, 10K without compression, or two 8K HDR screens running at 120 Hz refresh rate. All of this will be handled by Display Controller Next (DCN) engine for media.

The availability of DisplayPort 2.0 capable monitors is a story of its own. VESA noted that they should come at the end of 2021; however, they got delayed due to the lack of devices supporting this output. Having AMD's RDNA 3 cards as the newest product to support these monitors, we would likely see the market adapt to demand and few available products as the transition to the latest standard is in the process.

The mention of "170 W" in one of the slides of AMD's Computex 2022 reveal of the upcoming Ryzen 7000 "Zen 4" desktop processors, caused quite some confusion as to what that figure meant. AMD issued a structured clarification on the matter, laying to rest the terminology associated with it. Apparently, there will be certain SKUs of Socket AM5 processors with TDP of 170 W. This would be the same classical definition of TDP that AMD has been consistently using. The package-power tracking (PPT), a figure that translates as power limit for the socket, is 230 W.

This does not necessarily mean that there will be a Ryzen 7000-series SKU with 170 W TDP. AMD plans to give AM5 a similar life-cycle to AM4, which is now spanning five generations of Ryzen processors, and the 170 W TDP and 230 W PPT figures only denote design goals for the socket. AMD, in a statement, explained why it needed to make AM5 capable of delivering much higher power than AM4 could—to enable higher CPU core-counts in the future, more on-package hardware, and for new capabilities like power-hungry instruction-sets (think AVX-512). AMD has been calculating PPT as 1.35 times TDP, since the very first generation of Ryzen chips. For a 105 W TDP processor, this means 140 W PPT, and the same formula continues with Ryzen 7000 series (230 W is 1.35x 170 W).The AMD statement follows.

Biostar is seemingly trying to become more competitive in the consumer motherboard market and although the company has some catching up to do with the tier one motherboard brands, the company has put out some more interesting products in the past couple of years. Its VX670E Valkyrie motherboard seems to sit near the middle of the X670E models that have been announced so far, although it's a little bit hard to tell, as the company only provided partial specs. As this is an AMD X670E based board, the PCIe x16 slot is using PCIe 5.0, although it's multiplexed with the second x16 slot, which means if both slots are used, the bandwidth drops down to eight lanes per slot. The board layout doesn't suggest any PCIe 5.0 M.2 slots, but as this is mandatory on X670E board, we have to presume that the M.2 slot right below the x16 PCIe slot, is the PCIe 5.0 one. The board has a further three M.2 slots, as well as what appears to be an empty M.2 E-keyed slot for a WiFi/Bluetooth module.

Other features listed by Biostar includes 2.5 Gbps Ethernet via a Realtek chip, HDMI and DisplayPort outputs, six SATA 3 ports, one rear and one front header for USB Type-C Gen 3.2 2x2 (20 Gbps) ports, as well as Realtek based audio and a pair of ARGB headers. The board also has a debug LED and a few buttons and switches for resetting the CMOS, and powering the board on. Interestingly, Biostar also provided figures for the memory clock speed, as the company listed support for up to four sticks of DDR5 memory at 5600 MHz plus. AMD has already demoed higher memory clocks of 6000 and even 6400 MHz during its keynote, as supported by the footnotes that went alongside it, suggesting that this might just be a placeholder. Intel officially only supports 4800 MHz DDR5 memory, but speeds in excess of 6000 MHz doesn't appear to be an issue with the right motherboard and CPU combination. We'll have to wait and see what the official figures will be from AMD.

To usher in the new era of DDR5 and satisfy the demands for precise, high-performance, real-time computing, and computer multi-tasking in servers, data centers, and industrial computers, the world-leading memory brand TEAMGROUP is officially announcing today that TEAMGROUP's DDR5 Industrial Server Memory is now in mass production. The DDR5 Industrial Server Memory targets applications in high-end servers and large databases and is integrated with TEAMGROUP's patented technology, welcoming a total upgrade for next-generational servers such as Intel Eagle Stream and AMD Zen 4 Genoa.

TEAMGROUP's DDR5 ECC DIMM and DDR5 R-DIMM Industrial Server Memory modules can reach speeds of up to 6400MT/s and storage capacities of up to 128 GB, with total energy consumption reduced to 1.1 V. TEAMGROUP has also introduced Row Hammer Protection Technology to support the DFE (Decision Feedback Equalization) function to offer robust support for data storage. The memory is equipped with TEAMGROUP's patented "TRUST technology", which stands for Temperature, Robust, Unique, S.M.A.R.T. technology, and Trusty. With the TRUST technology, the industrial server memory can ensure reliable performance under harsh environments of high temperatures, high humidity, sulfurization, vibration, and shock as well as improving the overall reliability and durability of the product to guarantee that the memory can satisfy the diversified demands of next-gen servers. In addition, the DDR5 Industrial Server Memory has adopted a brand new framework. The greatest difference between the DDR5 and DDR4 server memories is that the power management is now transferred onto the DIMM, in other words, TEAMGROUP has optimized redundant circuit designs on the server, providing its DDR5 products a greater system stability, lower power consumption, better performances, larger capacities, and more stable signals.

GIGABYTE TECHNOLOGY Co. Ltd, a leading manufacturer of motherboards, graphics cards, and hardware solutions today announced all AIO liquid coolers lineup and tower fans for multi-platform are attached with brackets for Socket AM5 to support the new coming AMD motherboards, which provides users the optimal CPU heat dissipation and hands-down system upgrade without changing coolers.

Considering the inconvenience of upgrade CPU fans along with platforms, GIGABYTE always takes into account of the platform variation in the initial design stage of Cooler and tower fan. After confirmed the continuity of AM4 brackets on the Socket AM5 platform, GIGABTYE proved that all AIO liquid coolers lineup and tower fans can provide the optimal thermal performance to the new platform through rigorous tests. For users who has bought GIGABYTE coolers or tower fans designed for multi platforms, no change or adjustment will be needed to suffice the new platform.

AMD today announced new Confidential virtual machines (VMs) on the existing the N2D and C2D VMs on Google Cloud, all powered by AMD EPYC processors. These VMs extend the AMD EPYC processor portfolio of Confidential Computing on Google Cloud with the performance of 3rd Gen EPYC processors in compute-optimized VMs.

A key Confidential Computing component provided by AMD EPYC processors is AMD Secure Encrypted Virtualization (SEV), part of AMD Infinity Guard. This advanced hardware-based security feature encrypts full system memory and individual virtual machine memory as well as isolating the VM memory from the hypervisor, without dramatically impacting performance. With the expansion of Confidential Computing in N2D and C2D VMs, Google Cloud customers now have access to advanced hardware enabled security features powered by 3rd Gen AMD EPYC processors that will help protect sensitive, wide-variety workloads.

AMD is launching two distinct classes of next-generation enterprise processors, the 4th Generation EPYC "Genoa" with CPU core-counts up to 96-core/192-thread; and the new EPYC "Bergamo" with a massive 128-core/256-thread compute density. Pictures of the "Genoa" MCM are already out in the wild, revealing twelve "Zen 4" CCDs built on 5 nm, and a new-generation sIOD (I/O die) that's very likely built on 6 nm. The fiberglass substrate of "Genoa" already looks crowded with twelve chiplets, making us wonder if AMD needed a larger package for "Bergamo." Turns out, it doesn't.

In its latest Corporate presentation, AMD reiterated that "Bergamo" will be based on the same SP5 (LGA-6096) package as "Genoa." This would mean that the company either made room for more CCDs, or the CCDs themselves are larger in size. AMD states that "Bergamo" CCDs are based on the "Zen 4c" microarchitecture. Details about "Zen 4c" are scarce, but from what we gather, it is a cloud-optimized variant of "Zen 4" probably with the entire ISA of "Zen 4," and power characteristics suited for high-density cloud environments. These chiplets are built on the same TSMC N5 (5 nm EUV) process as the regular "Zen 4" CCDs.