Up until recently, it was thought that CUDIMMs would be exclusively supported by Intel's Arrow Lake-S desktop platform. However, it looks like AMD's Ryzen 9000-series processors will also be able to work with CUDIMMs - at least according to MSI - reports VideoCardz. 

During a recent webcast, MSI confirmed that while AMD's upcoming X870 and X870E motherboards will support CUDIMMs with Ryzen 8000 and 9000 series CPUs, they are currently incompatible with the Ryzen 7000 series processor family. 

During the recent event, MSI highlighted that its X870 motherboards have been optimized for better signal quality, which will improve their ability to support ultra-fast memory modules. While Ryzen 8000 (Phoenix) and 9000 (Granite Ridge) CPUs can work with CUDIMMs, previous-generation Ryzen 7000 (Raphael) CPUs cannot yet use CUDIMM due to an unresolved issue that prevents platform booting. MSI's TOPPC mentioned that this issue might be addressed in the future, but no official confirmation has been given. 

For now, it is unclear which CUDIMM speed bins AMD's Ryzen 8000 and Ryzen 9000 processors will officially support and whether they will be on par with Intel's Arrow Lake-S platforms. 

If (or when?) support for CUDIMMs by Ryzen 7000 is introduced, motherboards may initially start at a reduced memory data transfer rate, which could then be adjusted either manually or automatically to reach optimal speeds. However, it is unclear whether AMD is interested in adding support for CUDIMMs to its previous-generation processors and whether this support brings actual benefits. 

TOPPC also commented on memory speed advancements, suggesting that achieving speeds like 10,000 MT/s is still some way off. This indicates that while overclocking memory modules with cherry-picked memory devices and heightened voltages get higher data transfer rates, their widespread adoption could take quite some time. 

When it comes to JEDEC-specified performance settings, they are very conservative in terms of timings and voltages, and any increase is considered an overclock and therefore beyond the standard-set specification. Clock unbuffered dual in-line memory modules (CUDIMMs) feature a CDK driver to 'officially' support data transfer rates of 6,400 MT/s and beyond. Leading makers of memory modules for enthusiasts have already announced their products with 9,200 MT/s and 9,600 MT/s - presumably for Intel's Arrow Lake-S platforms.

AMD has lowered prices on its Ryzen 8000G series of desktop APUs across all retailers. As spotted by VideoCardz, AMD's entire suite of current-gen processors with integrated graphics has taken price drops up to $30, indicating a planned change ahead of Ryzen 9000's release in July.

The Ryzen 8000G desktop series first released in January of this year after a long wait; the last time AMD released desktop APUs before the 8000G series was in 2021. The 8000G series is the only line of processors on the current-gen AM5 socket which offer integrated graphics or an NPU, thanks to the Phoenix iGPU cores borrowed from AMD's Hawk Point mobile APUs. Today, three models of the Phoenix chips are available for sale at the new prices: the Ryzen 7 8700G, Ryzen 5 8600G, and Ryzen 5 8500G (the Ryzen 3 8300G was also released with the line but has never been available for sale to consumers). 

Ryzen 8000G new prices and where to find them

• Ryzen 7 8700G - $299, down from $329
  • Amazon
  • Newegg
• Ryzen 5 8600G - $199, down from $229
  • Amazon
  • Newegg
• Ryzen 5 8500G - $159, down from $179
  • Amazon
  • Newegg

We reviewed the Ryzen 7 8700G and Ryzen 5 8600G when they initially released back in January, and found them both to be alright, giving ratings of 3 and 3.5 stars respectively. It was difficult to recommend the 8700G at MSRP due to its strange positioning in the market. Nudget chips and the AM5 platform still don't mix perfectly thanks to the higher prices of AM5 motherboards and DDR5 RAM compared to the older AM4 platform and its DDR4 RAM. But the Ryzen 5 8600G is currently our pick for the best CPU for entry-level gaming, thanks to it having 90% of the performance of its bigger brother the 8700G for $100 less, and this recommendation is even easier with the $30 price drop.

The Ryzen 5 8500G is unique in the lineup for its completely different architecture; while the 8700G and 8600G both use only AMD's Zen 4 processor cores, the 8500G utilizes some of the weaker Zen 4c cores with some Zen 4, an unholy split first seen on AMD's Hawk Point mobile APUs. The 8500G also uses a weaker 740M graphics engine than the Radeon 760M in the 8700G/8600G, two significant sacrifices for only $40 less than the Ryzen 8600G.

The price shift for the three APUs likely comes as a harbinger for the arrival of the Ryzen 9000 series in July. Recent leaks suggest a launch date of July 31 for the new processors, which will reportedly usher in a 16% IPC improvement over the previous-gen for the new "Granite Ridge" line of Ryzen 9, 7, and 5 chips. 

AMD is set to beat Intel in some key battles across the desktop CPU space with the new Zen 5-based 9000 series. Add to this a drop in prices in its APU line which Intel has no alternatives to, and AMD is clearly aiming to win some ground back in the never-ending skirmish between Team Red and Team Blue.

Following up on past Mini PCs like the Geekom Mini FUN11, Geekom has launched the Geekom AX8 Mini PC on JD.com with a NUC-inspired design that comes in both AMD Ryzen 8000 Series and Intel 13th Generation-based variants. 

This Mini PC comes in two colors. White for Intel-based models and blue for AMD models. There is a considerable amount of variety in the specifications on offer, particularly if you want a worthwhile iGPU, so be mindful of what if you plan on buying one of these. The most expensive model, the Ryzen 9 8945HS model, tops out at 5099 Chinese yuans, roughly $703 USD.

Geekom AX8 Models and Core Specs

Geekom AX8 I/O Ports

Both machines feature a selection of ports. All the regulars are present. Two front facing USB 3 Type-A, and additional USB 3 Type-A port at the rear along with a USB 2 Type-A port. A single USB 3 Type C port is present, along with a USB 4 Type-C port. Wired networking duties are provided by a 2.5G Ethernet port. Finally we have a 3.5mm audio jack for legacy audio connections. 

Overall, the Geekom AX8 does seem like a fairly complete mini PC offering, particularly considering the near-direct adaptation of the formerly Intel, now Asus line of NUC Mini PC designs. The blue paint job is also a nice touch for the AMD users, though some would argue blue should have been kept for the Intel models instead and AMD CPUs should have gotten the white enclosure.

We recommend that any users hoping to do light gaming or any kind of GPU-accelerated work on these mini PCs opt for the models with Iris Xe graphics or (ideally) the Radeon 780M models, as well, since Intel UHD iGPUs are notoriously weak. The Radeon 780M is also still one of the strongest iGPU solutions on the market, though a stronger Radeon 890M now exists in the Ryzen AI 300 CPUs.

AMD has officially released the Ryzen 8000 F-series processors, including the Ryzen 7 8700F and Ryzen 5 8400F. The chipmaker had silently launched the duo of Zen 4 chips in China; however, according to leaked slides from hardware detective momomo_us, they're now available worldwide.

• Ryzen 7 8700F (Amazon)
• Ryzen 7 8700F (Newegg)
• Ryzen 5 8400F (Amazon)
• Ryzen 5 8400F (Newegg)

The Ryzen 8000 F-series processors leverage AMD's existing Phoenix APU silicon, which comes directly from TSMC's 4nm oven. The chips still feature Zen 4 execution cores but lack the integrated RDNA 3 graphics engine that's present in the Ryzen 8000G counterparts. The approach is analogous to what Intel does with its F-series models, fusing off the iGPU to offer cheaper models to attack lower price points. Therefore, you must pair the Ryzen 7 8700F and Ryzen 5 8400F with a discrete graphics card. In addition to the iGPU snip, AMD disabled the Ryzen AI NPU on one of the two models. Therefore, only the Ryzen 7 8700F still has the NPU intact, which offers 16 TOPS of AI performance.

The core and cache configuration on the Ryzen 8000 F-series chips remain unaltered compared to the G-series parts. The Ryzen 7 8700F has the same octa-core, 16-thread configuration with 16MB of L3 cache as the Ryzen 7 8700G, whereas the Ryzen 5 8400F shares a hexa-core, 16-thread layout with 16MB of L3 cache like the Ryzen 5 8600G. The Ryzen 7 8700F and Ryzen 5 8400F are still 65W processors and come with AMD's Wraith Stealth CPU cooler.

AMD Ryzen 8000 F-series Processor Specifications

Besides the absence of an iGPU, the difference in clock speeds sets the F-series apart from the G-series. The Ryzen 7 8700F has a 100 MHz lower base and boost clock than the Ryzen 7 8700G. The Ryzen 5 8400F also clocks 100 MHz less than the Ryzen 5 8600G. However, there's a big gap regarding the boost clocks as the Ryzen 5 8400F exhibits a 300 MHz slower boost. 

It's a little surprising that the F-series chips have lower clock speeds. Considering that the iGPU is disabled in these SKUs, there should be enough headroom for equivalent or higher clocks. The Ryzen 7 8700F and Ryzen 5 8400F are likely employing subpar silicon that doesn't comply with the requirements for the higher-tier levels. Silicon recycling is something every chipmaker does to maximize their yields. Fortunately, manual overclocking is enabled on the F-series processors, so there's always a chance that you can get them up to speed with their G-series peers.

AMD provided charts comparing the performance of the Ryzen 7 8700F and Ryzen 5 8400F to the competition. As with any launch, we recommend cautiously approaching vendor-provided benchmarks.

According to AMD, the Ryzen 7 8700F delivers 12% to 24% higher gaming performance than the Core i5-14400F. Meanwhile, the Ryzen 7 8700F reportedly has a performance advantage in applications that varies between 10% and 22%, depending on the workload. On the other hand, the Ryzen 5 8400F appears to outperform the Core i5-13400F with gaming margins between 1% and 14%.

The leaker didn't share the AMD slide with the pricing of the Ryzen 7 8700F and Ryzen 5 8400F. Nonetheless, we've already seen an early Amazon listing for the Ryzen 7 8700F at $299, and momomo_us believes that the Ryzen 5 8400F could retail for $189. 

According to AMD's numbers, the Ryzen 7 8700F offers up to 24% faster gaming performance and up to 22% higher application performance than the Core i5-14400F. However, the Ryzen 7 8700F is 50% more expensive than the Core i5-14400F, which retails for $199.98. On the other hand, the Ryzen 5 8400F could prove to be a very valuable proposition, assuming AMD's benchmarks are accurate. The Ryzen 5 8400F seemingly delivers up to 14% better gaming performance than the Core i5-13400F while costing less than 2% more than the latter. For reference, the Core i5-13400F has a $185.99 price tag.

While having more options and competition on the market is always great, it may pay to wait a few before considering an upgrade. AMD's next-generation Zen 5 processors should be on the market before 2024 concludes, as will Intel's rivaling Core Ultra 200 series (Arrow Lake) processors.

AMD announced its Ryzen Pro portfolio today as it extended the 'Hawk Point' 8040-series to commercial laptop and workstation users while simultaneously offering its Ryzen 8000 'Phoenix' APU models for commercial desktop PCs. As we've seen in the past, the Pro series is based on AMD's existing consumer-oriented processor models but comes with additional features that tailor them for the commercial market. With the consumer versions of these processors, AMD was the first x86 company to bring its AI-processing neural processing unit (NPU) to the mobile and desktop PC market. Those same AI acceleration features are now headed to commercial users, allowing AMD to lay claim to being the first with professional CPUs armed with NPUs for laptops and workstations.

AMD's in-built XNDA engine powers the NPU, and the company touts that its mobile processors have an advantage over Intel's competing Core Ultra processors with 16 TOPS of NPU performance, outperforming the Meteor Lake NPU's 11 TOPS. AMD also maintains a slimmer lead in overall system TOPS, which includes both the CPU and GPU AI processing power, taking the lead with 39 TOPS over Intel's 34 TOPS. AMD's desktop Ryzen 8000 APUs also have an in-built NPU engine that delivers 16 TOPS, whereas Intel has yet to release a processor for desktop PCs with an integrated AI engine.

Notably, neither AMD nor Intel's chips meet Microsoft's next-gen AI PC requirement for 45 TOPS of performance from the NPU, though both companies say their next-gen chips, Strix Point and Lunar Lake, respectively, will meet that bar. The 45 TOPS requirement is meant to enable Microsoft to run AI elements of Copilot locally, and it isn't clear how a future Windows update to enable that functionality will pan out with the current generations of AMD and Intel processors. Qualcomm's Snapdragon X Elite Arm chips will debut with 45 TOPS of performance from its NPU in the middle of the year. Apple's M3 processors provide 18 TOPS of NPU performance but obviously aren't impacted by Microsoft's requirements. 

The ability to run AI workloads locally helps alleviate critical privacy concerns in the commercial space. It also delivers latency, performance, and battery life advantages for AI applications (power efficiency is dramatically better with the NPU than running the same AI workload on the GPU). However, unleashing that performance requires tightly coupled software solutions, and in many respects, the early war for AI dominance will come down to developer partnerships.

The AI ecosystem is still raw as developers work to leverage NPU acceleration in their products, but AMD, Microsoft, and others are working to build the foundations that will enable a broad swath of AI-accelerated features in software and Windows. AMD cites that it will have 150+ ISV (independent software vendor) partnerships to develop AI-driven solutions in 2024, in contrast to Intel's broadening AI Developer Program, which has 100 ISV and 100+ IHV (integrated hardware vendor) partnerships thus far.

Naturally, unleashing the power of locally-run AI starts with the hardware. Let's take a closer look at the silicon.

AMD Ryzen Pro 8040-series processors

As noted, the Ryzen Pro 8040-series processors are based on the consumer-oriented 'Hawk Point' processors — you can learn the details of this Zen 4-powered lineup here. The U-series processors slot in for the 15-28W TDP range, while the HS variants are carved into 20-28W and 35-54W swimlanes spanning the Ryzen 9, 7, and 5 product lines, with the 45W models geared for workstations.

Core counts range from six cores and 12 threads with Ryzen 5 up to eight cores and 16 threads with Ryzen 7 and 9. Peak clock rates remain the same as the consumer variants. Notably, the Ryzen 5 Pro 8540U comes without an integrated NPU, just like the consumer variant. All other key criteria, like clock speeds, iGPU, and cache, remain the same.

AMD shared benchmarks comparing its silicon to Intel's competing chips, but as with all vendor-provided benchmarks, we should take them with a grain of salt. AMD claims an average 30% advantage with the Ryzen 7 Pro 8840U at 15W over the 15W Intel Core Ultra 7 165U in a range of workloads spanning multitasking, rendering, productivity, graphics, and content creation workloads. The company also claims an 18% average advantage over the 28W Core Ultra 7 165H.

AMD also says its 45W Ryzen 9 Pro 8945HS for workstations delivers a 50%+ advantage in the Topaz Labs AI-powered video application over the 45W Intel Core Ultra 9 185H and up to 77% faster time to the first token with the Llama 2 LLM. AMD also shared benchmarks with a wider range of AI models, like facial and object recognition, among others. Claimed performance advantages range from 5% to 23% across a spate of more general CPU-centric applications. 

The company also claims a 46% battery life advantage over the Intel Core Ultra 7 165H during a Teams workflow and a slight advantage over the Apple M3 Pro CPU. However, the Apple processor is missing from any of AMD's direct performance comparisons.  

AMD Ryzen Pro 8000-series APUs

The Ryzen Pro 8000-series for desktop PCs is built upon the foundation of the powerful 8000G APU series for the consumer market, which you can read more about here and here. All of these models have an in-built NPU AI acceleration engine.

Once again, all relevant specs remain identical to the consumer models, with options spanning from four cores and eight threads up to eight cores and 16 threads across the Ryzen Pro 3, 5, and 7 families. The standard 8000G models have a configurable 45-65W TDP range, whereas the GE models slot in for a lower-power 35W TDP.

Overall, AMD claims a 19% average uplift over Intel's competing Core i7-14700, with a peak of a 3X advantage in the Time Spy graphics benchmark. That isn't surprising, given the powerful combination of the Zen 4 CPU with the RDNA 3 graphics engine. AMD also claims from 33% to 76% less power consumption across a span of different configurations.   

AMD Ryzen Pro Technologies

The integrated AMD Pro Technologies suite is the key difference between AMD's Pro models and its standard consumer chips. This hardware and software suite includes features like AMD Pro Security, which includes multiple layers of security that leverage proprietary OEM solutions and in-built Windows 11 features, AMD memory guard, Microsoft Pluton, and the AMD Secure processor. The 8000-series family marks the first desktop PCs to integrate Microsoft Pluton security functionality and cloud-based remote manageability. 

The AMD Pro Manageability features simplify provisioning, system imaging, and deployment tasks, while the AMD Pro Business Ready suite ensures stability and includes a quality and reliability guarantee. AMD points out that it offers the full suite of Pro Technologies on all of its processor models, whereas Intel bifurcates support for segmentation purposes.

AMD's OEM platforms have a range of refreshed products built on the underpinnings of the 8040- and 8000-series processors. The Ryzen Pro series is available to AMD's OEM partners now.

Sometimes new drivers point to products that aren't yet released, but for whom launch is imminent. This may be the case with AMD's Ryzen 9000-series products, as the company's chipset driver V6.03.19.217, released by Asus this week, included mentions of the Ryzen 9000 platform management framework (PMF), according to X user @9550pro.

Release notes for AMD's chipset driver V6.03.19.217 for its AMD X670E-based motherboards this week indicate that, in addition to AMD's Ryzen 6000-, 7040-, 8000-, and 7736-series platforms, the software package also supports AMD's (as of yet) unreleased Ryzen 9000-series platforms. This doesn't necessarily mean AMD's Ryzen 9000-series processors are just around the corner, but it at least indicates the company is prepping for launch. 

AMD's current-generation Zen 4-based products for high-performance desktops belong to the Ryzen 7000-series range, while Zen 4-powered processors with integrated Radeon graphics for mainstream desktops belong to the Ryzen 8000-series family. Therefore, AMD's next-generation Zen 5-based CPUs are poised to find themselves in a brand new product family — which will likely be the Ryzen 9000-series lineup.

AMD has publicly stated it will continue to support the AM5 socket until at least through 2025, so its Zen 5-based processors set to launch this year will use the existing socket and existing motherboards. So it's not surprising that AMD and its motherboard partners are getting ready to support Zen 5-based CPUs (most likely, Ryzen 9000-series) on existing platforms, such as the AMD X670E. 

Note that AMD's chipset drivers tend to start adding support for new processors months before they're released — we saw this with the the Ryzen 8000-series accelerated processing units in late 2023. AMD is expected to release Zen 5-based Ryzen 9000-series offerings this fall, so it's about time for the company to start prepping for launch. Computex 2024 is just weeks away, and we wonder if AMD and its partners are getting ready to showcase Ryzen 9000-based machines at the Taipei-based tradeshow. 

AMD's Zen 5-based processors are expected to increase performance compared their predecessors in various workloads, so high-end desktop enthusiasts are eagerly awaiting them.

AMD is reportedly working on a revamped version of its AM5 socket, the socket AM5+. The MC Extractor tool, a program that parses microcode binaries of AMD and Intel CPUs, recently added detection of AM5+ and AMD processors with the 00B40Fxx CPUID, as hardware detective HXL noticed.

The tool has 'Adjusted AMD microcode pattern for AM5+' and adds 'AMD microcode CPUID 00B40Fxx detection.' Speaking of 00B40Fxx CPUs, the tool can detect 00B40F40 and 00B40F00 processors. HXL believes that 00B40Fxx CPUs could be AMD's codenamed Granite Ridge or Fire Range processors, but this is speculation.

While simultaneously mentioning the unreleased CPU and socket may mean that AMD's processors with the said identification will use the AM5+ socket, this may not be the case at all, which is unrelated.

An avid reader will probably ask what Socket AM5+ is in this case. The most logical answer is that it will replace the original AM5 sometime in 2026 and beyond. In this case, AM5+ will add certain new features, such as different voltages, different power supply circuitry, different PCIe bifurcation, new interface capabilities (e.g., USB4 version 2, DisplayPort with UHBR20), or something else.

It should be noted that while we traditionally call CPU sockets by the number of pins they have (in the case of AMD's AM5, which will be 1718 pins), modern CPU sockets tend to have more pins than they use. Therefore, the socket AM5+ could be just the socket AM5 with some extra pins enabled. Meanwhile, AM5+ CPUs will probably have a mechanical key to ensure they are not installed into an AM5 motherboard.

The critical aspect of AM5+ is that it is poised to reuse as many AM5 mechanical features as possible to lower costs for AMD's partners. Also, AMD might try to ensure backward compatibility of AM5+ motherboards with AM5 processors, which helps to quickly launch the platform across different market segments.

Meanwhile, AMD has publicly stated that it will support the AM5 socket through 2025 and beyond, so it is safe to say that Zen 5-based processors launching this year will use the established socket. That said, it is reasonable to expect the emergence of AM5+ sometime in 2026.

AMD reports that its older Ryzen mobile 7040 Phoenix and Ryzen mobile 8040 series processors outperform Intel’s Core Ultra Meteor Lake CPUs by up to 79% in various large language models (LLMs). The CPU manufacturer unveiled a plethora of benchmarks against Intel’s Core Ultra 7 155H CPU compared to the Ryzen 7 8740U. Both chips sport hardware-based Neural Processing Units (NPUs).

AMD put together several slides featuring performance results in Mistral 7b, Llama v2 and Mistral Instruct 7B with the two CPUs. In Llama v2 Chat using a Q4 bit size, the Ryzen chip achieved 14% faster tokens per second than the Core Ultra 7 155H. With the same bit size in Mistral Instruct, the Ryzen chips achieved 17% faster tokens per second. In the same LLMs, but looking at Time to First Token for Sample Prompt, AMD’s competitor was 79% faster than the Core Ultra 7 in Llama v2 and 41% faster in Mistral Instruct. 

AMD showed another chart of Llama 2 7B Chat using a plethora of different bit sizes, block sizes, and quality levels. On average, the Ryzen 7 7840U was 55% quicker than the Intel counterpart and up to 70% faster in the Q8 results. Despite Q8 being the fastest, AMD recommends a 4-bit K M quantization for running LLMs for real-world use and setting a 5-bit K M for tasks requiring extreme accuracy, like coding.  

We are not surprised that AMD is currently winning the AI performance war with Intel. Despite its Ryzen 7040 series architecture having the same level of performance (in TOPS) as Meteor Lake, we discovered late last year that AMD often outperforms Meteor Lake in AI-based workloads. This appears to be a problem with LLM optimization rather than a hardware or driver issue. We noticed AMD notably wins in AI workloads that don’t take advantage of Intel’s OpenVINO framework, which is optimized for Intel products only. OpenVINO appears to be vital to significantly boosting Intel AI performance. Intel’s A770, for instance, gets a tremendous 54% performance improvement purely from OpenVINO optimizations.

Don’t expect this performance behavior to last long. We are only at the beginning of NPU development, after all. If more apps don’t embrace OpenVINO, we expect Intel to switch gears and try a better optimization route—one that will be adopted by more developers. Intel is also getting ready to unleash its next-generation Lunar Lake mobile CPU architecture later this year, which will reportedly feature 3x the AI performance of Meteor Lake (on top of huge IPC improvements for the CPU cores). 

For now, AMD’s slides demonstrate that it currently has the edge in NPU performance, especially with its Ryzen 8040 series CPUs, which have even more NPU performance than the Ryzen 7 7840U. But by the end of this year, the tables could turn depending on how successful Intel is with Lunar Lake and its AI optimization plans.

The headlining purpose of AMD's Beijing AI PC Innovation Summit yesterday was to launch the Ryzen 8040 series and 8000G desktop solutions for the Chinese market. We reported on the fresh Hawk Point info from the event, but there was a slide we didn't spot during the two hours of footage. Twitter's HXL shared the image below, showing Spencer Pan, AMD SVP and President of Greater China, revealing that the Ryzen 7 8700F and Ryzen 5 8400F are being prepared for launch.

Based on prior AMD (and Intel) SKUs, the F suffix indicates that these processors lack onboard graphics. Perhaps the iGPU failed some tests in the chip plant, and instead of these parts going to landfill or being converted into shoes, we have some F suffix chip options. The F suffix chips usually offer a modest discount compared to their fully-fledged brethren. PC DIYers who insist on discrete graphics typically aren't that bothered about missing an iGPU.

The above slide shows that the Ryzen 7 8700F and Ryzen 5 8400F are part of AMD's plan of "attacking every price point." It looks like a wide range of Ryzens will be maintained for this purpose, ranging from the Athlon 3000G through Ryzen 5000 series parts, plus these new Ryzen 8000 desktop APUs (sans graphics) that will provide more options for owners or buyers of AM5 motherboards.

We don't have official specs for the newly revealed Ryzen 7 8700F and Ryzen 5 8400F desktop processors, but it would be reasonable to assume they are similar to the AMD Ryzen 8000G Series 65W Phoenix APUs (without onboard graphics).

For example, the AMD Ryzen 7 8700G (review link) has an 8C/16T CPU configuration using the Zen 4 architecture, with boost speeds up to 5.1 GHz and a TDP of 65W. It also has a powerful Radeon 780M (12 CU) GPU using the RDNA 3 architecture. We wouldn't be surprised if all the CPU-related specs remain the same, but no onboard graphics option exists.

While there are AMD Ryzen 5 8600G, Ryzen 5 8500G, and Ryzen 3 8300G desktop APUs, there isn't a corresponding Ryzen 5 8400G to guestimate Ryzen 5 8400F specs and features. However, if pushed, we would say it is likely that the new Ryzen 5 part will have a 6C/12T (or possibly 6C/6T) Zen 4 architecture CPU configuration sans graphics.

Last but not least, the new Ryzen 7 8700F and Ryzen 5 8400F chips may be China-only products, so we don't have to ponder over what kind of a discount they will be sold for vs. G-suffixed parts. HXL has a hunch that these are only being prepared for sale in China, but it isn't certain. Remember, AMD's China-only parts can sometimes escape beyond the Great Wall - we now have Radeon 'Golden Rabbit Edition' GPUs available at retailers in the West.

AMD has added support for its upcoming processors based on the next-generation Zen 5 microarchitecture into the GCC compiler. The 'znver5' is now a target in the GCC Git repository in time for the GCC 14.1 stable release, reports Phoronix. This move ensures that developers will have the necessary compiler support for Zen 5-based Ryzen and EPYC CPUs ahead of their launch later this year. 

In February, AMD announced the addition of the Zen 5 target to the GCC compiler, introducing new instructions such as AVXVNNI, MOVDIRI, MOVDIR64B, AVX512VP2INTERSECT, and PREFETCHI. These enhancements build on the capabilities of the previous Zen 4-powered processors, offering improved performance and efficiency for AMD's upcoming processors for client and datacenter PCs. 

The inclusion of Zen 5 support in the GCC compiler ahead of any product launches is a natural move by AMD as the company wants software makers to prep for its next-generation products. The addition of znver5 target to GCC allows software makers to optimize, fine-tune, and fully leverage capabilities of the new processors in their programs. As a result, software will be able to fully use the potential of Zen 5-based systems from day one. 

Currently, the Zen 5 support in GCC uses the Zen 4 cost table, but AMD is expected to provide additional optimizations and enhancements in future patches. This ongoing development will further refine the compiler's ability to take full advantage of Zen 5's capabilities.

AMD is expected to release its Zen 5-based Ryzen and EPYC processors for client and datacenter computers later this year. Some of these CPUs will be made on TSMC's 3nm-class fabrication process, whereas others will rely on a 4nm-class manufacturing technology. It is projected that Zen 5-based processors will noticeably increase performance both in single-threaded and multi-threaded workloads. 

What remains to be seen is when exactly AMD will release its Zen 5-based offerings. Some believe that this will happen at the end of the year, so it is reasonable to expect Zen 5-based systems on the shelves by the holiday season. Meanwhile, the ramp of next-generation EPYC will likely happen in 2025.

Right now at Newegg, you can find quite a few AMD processors for some notably low prices. This includes a wide range of stock starting with AM4 Ryzen 5000 series processors all the way up to some of the latest AM5 Ryzen 8000G series. Discounts vary across the board but we'll highlight a few notable deals that you won't want to miss out on.

Your best budget option right now is the 6-core AMD Ryzen 5 5500, an AM4 processor that can reach up to 4.2 GHz. When using promo code GUDDQA69, you can snag this CPU for $92. A step up from the Ryzen 5 5500 is the AMD Ryzen 7 5700X3D, an 8-core processor that can reach 4.1 GHz, which is $241 when you use promo code GUDDQA77. For those looking for something strong on the performance end, the AMD Ryzen 9 5950X is listed for just $372 with promo code GUDDQA65.

The AMD Ryzen 7000 series is also part of the site-wide price drop. The 5.0 GHz AMD Ryzen 7 7800X3D with 3D V-Cache is listed for just $399. For a little bit less, you can find the AMD Ryzen 9 7900, a 12-core AM5 processor, priced at $354 when you use promo code GUDDQA74 at checkout.

The 8000 series has the smallest discounts available, but they are discounts, nonetheless. The Ryzen 5 8600G is a 6-core CPU capable of reaching 5.0 GHz with Max Turbo enabled. Promo code GUDDQA78 will save you an extra $5 off its asking price, putting it at $224. The Ryzen 7 8700G runs a little higher at $324 when using promo code GUDDQA76 but it also comes with beefier specs to justify the price tag like 8 cores and the ability to reach speeds as high as 5.1 GHz.

There are more AMD Ryzen discounts than these across the Newegg website but hopefully, this cherry-picked list will jumpstart your research. Many of these discounts are scheduled to end by March 3rd and we aren't sure when we'll see them again.

The Tianbao Gem12 Mini PC, with AMD Ryzen Hawk Point APU option, launched on JD.com on March 1, reports ITHome. Among other features, this Mini PC has a nifty little built-in screen with a fingerprint reader, plus configuration options including various mobile Ryzen 7 and Ryzen 9 CPUs. 

The most attractively powerful of the available CPU options is the newest Ryzen 7 8845HS, which is basically a higher-power version of the Ryzen 7 8840U and 7840U that we've seen become commonplace in other Mini PCs and handheld gaming PCs alike thanks to the built-in Radeon 780M— one of the best 720p/1080p iGPUs on the market, though Intel's competition is definitely heating up. 

Like the 8840U and 7840U, the 8845HS is specced for 8 cores at 16 threads with a maximum boost clock of up to 5.1 GHz...but has a 17 Watt higher default TDP (45W vs 28W), which should help squeeze more performance out of those components. The advertising for this Mini PC states that it can actually cool up to 75 Watts, though we'll have to wait and see whether or not we can actually push TDP that high without causing crashes. With the information available to us, though, this should be one of the most performant Ryzen 7 Hawk Point PCs on the market, period.

Another point in its favor is that it seemingly supports full-power DDR5 RAM at up to 5600 MT/s rather than lower-bandwidth LPDDR5, though you would need to mount "naked" RAM sticks if you were bringing your own, as there definitely isn't room for RAM heatsinks.

Below, we'll be providing some data points for this Mini PC from the original JD.com listing before wrapping things up.

• All versions come in barebones (no RAM/NVMe), 16GB RAM w/ 512GB NVMe, or 32GB RAM w/ 1TB NVMe configurations.
• All Secondary Screen versions include the fingerprint sensor.
• The three APU options are Ryzen 9 6900HX, Ryzen 7 7840HS, and Ryzen 7 8845HS (recommended). The cheapest barebones 8845HS version starts at CNY2699, or roughly $379 USD.

Besides the core specs and unique form factor with the built-in mini screen, this Mini PC boasts most of the usual I/O we've come to expect: Dual Ethernet, an undefined USB Type-C port, HDMI 2.1 and DisplayPort 1.4 video ports, four USB 3.2 Gen2 ports, a headphone jack, a USB4 Type-C port, and most interestingly, an OCuLink port. The OCuLink port should allow for the unit to maximize the potential of an attached eGPU, especially compared to the slower Thunderbolt and USB4 standards.

For now, this little mini PC is a China-exclusive that requires some shipping costs to bring elsewhere in the world, but Tianbao may very well end up bringing it stateside via the Aoostar branding, depending on its reception.

AMD launched its Ryzen 8000G-series accelerated processing units (APUs) with Zen 4 and Zen 4c cores for desktops just a little over a month ago and apparently the company is prepping to release 35W versions of these processors shortly. At least, some of AMD's partners are already listing some of the company's Ryzen 8000GE-series products as supported by some of their motherboards on their website (as noticed by @harukaze5719). This is not official information from AMD, so take it with a grain of salt.

For now, Asus lists six AMD Ryzen 8000GE processors with built-in Radeon RDNA 3 graphics and a 35W thermal design power with four of them being Ryzen Pro 8000GE-badged offerings and only two are regular Ryzen 8000GE APUs. At least, as far as the Ryzen Pro 8000GE family is concerned, it looks complete with the eight-core Ryzen 7 Pro 8700 GE, six-core Ryzen 5 Pro 8600 GE, six-core Ryzen 5 Pro 8500 GE, and quad-core Ryzen 3 Pro 8300 GE. By contrast, the non-Pro family members listed by Asus includes the six-core Ryzen 5 8500 GE, and quad-core Ryzen 3 8300 GE.

To reduce power consumption of its Ryzen 8000GE products to 35W, AMD lowered their clocks, but it looks like they preserve core count and cache sizes of their regular counterparts that feature a thermal design power of 35W.

It is possible that AMD plans to address business desktops with Ryzen Pro 8000GE-series processors first as these systems tend to be small and require low-power CPUs and only then release the full stack of its Ryzen 8000GE-series non-Pro APUs. Traditionally, AMD's Ryzen Pro APUs support ECC memory, memory encryption, secure boot, DASH remote management capabilities, and Microsoft Pluton. 

Just like in case of regular Ryzen 8000G-series APUs, it looks like the Ryzen 5 8600G and Ryzen 7 8700G are expected to employ the premium Phoenix silicon, boasting up to eight Zen 4 cores and Radeon 780M graphics, whereas Ryzen 3 8300G and Ryzen 5 8500G models rely on AMD's Phoenix 2 silicon, featuring up to 6 cores and Radeon 740M graphics.AMD's Phoenix will naturally provide superior performance for both general-purpose computing and graphics tasks as well as incorporates the Ryzen AI accelerator to handle machine learning workloads.

When AMD formally introduced its Ryzen 8000G-series accelerated processing units for desktops in early January, the company mentioned that they supported ECC memory capability. Since then, the company has quietly removed mention of the technology from its website, as noted by Reddit users. 

We asked AMD to clarify the situation and were told that the company has indeed removed mentions of ECC technology from the specifications of its Ryzen 3 8300G, Ryzen 5 8500G, Ryzen 5 8600G, and Ryzen 5 8700G. The technology also cannot be enabled on motherboards, so it looks like these processors indeed do not support ECC technology at all.

While it would be nice to have ECC support on AMD's latest consumer Ryzen 8000G APUs, this is a technology typically reserved for AMD's Ryzen Pro processors. So it would be an odd move for the company to add it to regular products. That said, expect AMD's yet-to-be-announced Ryzen Pro 8000G APUs to support ECC, along with other Pro-grade capabilities, such as DASH remote management, AMD Memory Guard DRAM encryption technology, and other AMD Pro advantages.

AMD's Ryzen 8000G-series lineup includes four models compatible with AM5 platforms. The inexpensive Ryzen 3 8300G and Ryzen 5 8500G use the company's Phoenix 2 silicon, featuring up to 6 cores (two high-performance Zen 4 and four high-density Zen 4c cores) and equipped with the Radeon 740M graphics processing unit. The more expensive Ryzen 5 8600G and Ryzen 7 8700G employ the more advanced Phoenix silicon, which includes up to eight high-performance Zen 4 cores and come with the superior Radeon 780M GPU. Phoenix-based APUs are designed to deliver greater performance in both general computing and graphical tasks. In addition, the Phoenix platform incorporates a Ryzen AI accelerator to enhance machine learning capabilities. 

While ECC support is not coming to regular AMD Ryzen 8000G-series processors (as it requires validation and appropriate testing), it will certainly be supported by AMD's Ryzen Pro 8000G-series accelerated processing units, whenever they arrive. For now, it'sd unclear when that will happen, but it is reasonable to expect them to arrive in the coming months. 

AMD has added support for its next-generation accelerated processing units codenamed Strix Point to the ROCm software stack used to program GPUs, as noticed by @Kepler_L2. The addition confirms that AMD's upcoming processors based on the Zen 5 microarchitecture will feature GPUs based on the RDNA 3.5 architecture.

AMD's Strix Point Halo APU will feature AMD's GFX1151 GPU IP with RDNA 3.5 clusters, whereas regular Strix Point processors will feature another GPU identified as the GFX1150 and feature fewer RDNA 3.5 GPU clusters. The exact difference between AMD's RDNA 3 and RDNA 3.5 GPU architectures is unknown, though we can certainly expect performance and power efficiency improvements.

AMD's Strix Point products will belong to the company's Ryzen 8050-series processors and are due to be released in 2024 or 2025. Meanwhile, Strix Point Halo is projected to be based on the Zen 5 microarchitecture and higher-performance RDNA 3.5 GPU. In contrast, the regular Strix Point is rumored to feature a combination of Zen 5 and Zen 5c cores (for a total of 12 cores) and a smaller, lower-performance RDNA 3.5 GPU. Also, all Ryzen 8050-series APUs are expected to feature an XDNA 2-based AI engine.

There is also speculation circulating among hardware enthusiasts that AMD's Strix Point Halo will feature a different design than the regular version: the vanilla Strix Point is expected to feature a monolithic design, whereas the Halo is expected to feature a chiplet structure. However, there is no evidence available so far to either substantiate or refute these claims.

One of the particularly interesting things about AMD's Strix Point is that their mentions were found in ROCm code. ROCm is AMD's software stack for artificial intelligence (AI) and high-performance computing (HPC), so the addition of AMD's GFX1150 and GFX1151 GPUs may indicate that the APU will be able to run AI or HPC applications on its GPU cores.

While initial Strix Point APUs will primarily address laptops and small form-factor desktops, AMD is expected to eventually release desktop versions of these processors that will fit into AM5 motherboards.

AMD is gearing up to release its Ryzen 8000G-series accelerated processing units for desktops with built-in RDNA 3-based GPU, and the latter tend to be hungry for memory bandwidth. PCWorld decided to ask AMD whether users buying advanced versions of AMD's next-generation desktop APUs should also buy expensive memory kits for their new rigs to avoid any performance bottlenecks for built-in Radeon GPUs. The company explained that its Ryzen 8000G will be fine with a dual-channel DDR5-6000 memory subsystem. 

"So, first thing is dual-channel RAM, absolute must, that is a huge bandwidth advantage," Donny Woligroski, Technical Marketing Manager at AMD, told PCWorld. "Do not skimp, you gotta have dual-channel. DDR5-6000 is pretty cheap nowadays. If you can do dual-channel DDR5-6000, you are gonna hit those great frame rates and really playable performance and that is definitely that we will steer people."

A dual-channel DDR5-6000 memory subsystem offers a peak memory bandwidth of 96 GB/s, which is shared between Zen 4 CPU cores, Radeon 7000-series iGPU, and an NPU. AMD previously said that DDR5-6000 memory is the sweet spot for AMD's Ryzen 7000-series CPUs featuring from eight to 16 Zen 4 cores, so it looks like the company is just re-iterating its claims. 

Meanwhile, 96 GB/s does not seem to be a lot for a decent iGPU. AMD's Ryzen 7 8700G is expected to feature the company's range-topping RDNA 3-based Radeon 780M integrated GPU with 768 stream processors. The mobile version of the Radeon 780M can operate at clocks between 800 MHz and 2700 MHz, thus offering FP32 performance of up to 8.29 TFLOPS, making it a powerful graphics processor. To add some context, AMD's Radeon RX 6600 XT features a peak FP32 performance of 8.928 TFLOPS and boasts a 256 GB/s peak memory bandwidth and 32 MB of Infinity Cache. 

Meanwhile, we should remember that the performance of virtually all high-performance integrated GPUs is constrained by memory bandwidth, so AMD and Intel are in the same boat here. Therefore, AMD's Ryzen 8000G promises to be very competitive regarding graphics performance. A Ryzen 7 8700G-based rig should also not be too expensive as one can grab a 32 GB DDR5-6000 dual-channel kit for less than $100.

Another advantage of AMD's Ryzen 8000G APUs is that they are part of the AM5 platform, which the company promised to support for the next few years.

As reported on the Asus ROG and HardwareLuxx forums, Asus has officially launched a series of new beta BIOS updates supporting AMD's latest AGESA ComboAM5 microcode update 1.1.0.1. While nothing has been formally confirmed about this new microcode update, it's evident that it helps AMD's upcoming Ryzen 8000G series desktop APUs that are expected to arrive next year.

The new beta BIOS updates apply to several AM5 ROG, TUF Gaming, and ProArt X670E, B650E, and B650 motherboards, including the ROG Crosshair X670E Hero, Gene, Extreme, ROG Strix X670E-A Gaming WiFi, -E Gaming WiFi, -F Gaming WiFi, and -I Gaming WiFi. ROG Strix B650E-E Gaming WiFi, -F Gaming WiFi, -I Gaming WiFi, and B650-A Gaming WiFi, TUF Gaming X670E Plus, Plus WiFi, B650-Plus, Plus WiFi, B650M-Plus, and Plus WiFi. ProArt X670E Creator WiFi, and B650-Creator. The updates can be downloaded from Asus via Google Drive links, but be warned: these new BIOS updates are not entirely stable and might encounter bugs or instability issues.

Ryzen 8000 is the terminology for AMD's upcoming Zen 4-based APUs aimed squarely at desktop systems. Leaks and rumors report that these new chips will be equipped with Phoenix and Phoenix 2 silicon from AMD's mobile CPUs, which not only includes a high-performance RDNA 3 GPU but also a dedicated XDNA/XDNA2 Neural Processing Unit (NPU) that is designed to accelerate artificial intelligence applications at a hardware level.

AMD Ryzen 8000G Specifications*

*Specifications are unconfirmed.

We know four SKUs that will be part of AMD's new APU lineup, including the Ryzen 7 8700G, Ryzen 5 8600G, Ryzen 5 8500G, and Ryzen 3 8300G. The Ryzen 7 8700G reportedly has eight Zen 4 cores and a Radeon 780M iGPU sporting 12 CUs. The Ryzen 5 8600G will reportedly come with six Zen 4 cores and 760M with 8 CUs.

The latter two chips will seemingly come with AMD's more complex Phoenix 2 silicon, which incorporates a hybrid core design consisting of Zen 4 and Zen 4c cores. The Ryzen 5 8500G will purportedly come with two Zen 4 cores and four Zen 4c cores, while the Ryzen 3 8300G will have just a single Zen 4 core and three Zen 4c cores. The iGPU on both models includes the Raedeon 740M with 4 CUs.

These new chips are expected to be unveiled at CES 2024, which debuts on January 9th and ends on January 12th.

Renowned hardware leaker @harukaze5179 has found a list of AMD's upcoming Ryzen 8040-series accelerated processing units with built-in graphics. These upcoming Ryzen 8040-series processors are codenamed Hawk Point, though these are essentially rebadged Ryzen 7040-series CPUs known as Phoenix and Phoenix 2.

The product names indicate that AMD will offer Ryzen 9, Ryzen 7, and Ryzen 5 products within the Ryzen 8040-series lineup, and those CPUs will feature a TDP of 28W or 45W, depending on their actual specifications. While the range-topping Ryzen 9 8940H and Ryzen 7 8840H/HS models will carry eight multi-threaded Zen 4 cores and an RDNA 3-based GPU with up to 768 stream processors, some Ryzen 5-branded parts could either pack six Zen 4 cores or use a hybrid design featuring two high-performance Zen 4 cores and four smaller Zen 4c cores.

AMD's Ryzen 8040-series processors will be available in three different packaging options, just like their Ryzen 7040 predecessors. The FP7 (LPDDR5X) and FP7r2 (DDR5) packages are tailored for slimmer designs that require a high-performance APU, while the larger FP8 package is intended for the highest functionality and performance. The FP8 is significantly bigger than the FP7 variants and supports more advanced interfaces like AMD's MIPI CSI, a fast interface for linking things like cameras to host APUs.

It is likely that the new parts will be drop-in compatible with existing motherboard designs aimed at Ryzen 7040-series products. This will greatly simplify the lives of PC makers as they will not have to redesign anything and still get some kind of upgrade. 

What remains to be seen is whether Ryzen 8040-series parts will be faster than their direct predecessors and if so, how much faster these are going to be. Keeping in mind that TSMC's N4 process technology (4nm) is getting better, AMD could probably increase the clocks of its parts without risking yields, but the company has not formally announced any specifications of its Ryzen 8040-series CPUs.

A Chiphell contributor has published a flyer advertising an alleged Alienware desktop with AMD's yet-to-be-announced Ryzen 9000-series processor. This could be a typo on Alienware's side, a forged picture, or AMD could be close to introducing its next-generation Ryzen 9000-series CPUs for desktop PCs. Whatever the case, we need to spread a healthy dose of salt on this revelation.

Given the fact that AMD is set to announce Zen 5-based Ryzen processors for desktops and Zen 5-powered EPYC CPUs for servers in 2024, it is possible that the company will reserve Ryzen 9000 branding for Zen 5 parts. What is a bit surprising is that AMD's partner Alienware is already talking about these CPUs, which leads further credence that this is just a typo on behalf of the PC maker. Meanwhile, if AMD announces its Zen 5-based Ryzen 9000-series at CES, it is about time for Alienware to start teasing its systems based on these CPUs.

Yesterday, an unofficial source claimed AMD is prepping Ryzen 8000-series desktop accelerated processing units (APUs) based on Zen 4 and Zen 4c cores with built-in RDNA 3 graphics. The company is also expected to release Ryzen 8000-series offerings for laptops with Zen 4 and Zen 4c-based designs, so it looks like AMD's Ryzen 8000 lineup will still be based on Zen 4 or older technology. That said, inserting high-end Zen 5-based CPUs into AMD's Ryzen 8000 lineup is illogical.

AMD has a somewhat convoluted strategy when naming its Ryzen series. The company has used a mix of microarchitectures within one generation of Ryzen several times. For example, the Ryzen 7000 family includes Zen 4, Zen 3, and Zen 2-based products. 

Meanwhile, to ensure enthusiasts receive the correct messaging about all-new hardware, introducing the Ryzen 9000 for Zen 5 seems like the best course of action for AMD.

The first details about AMD's upcoming Ryzen 8000G-series accelerated processing units (APUs) for desktops have leaked. They indicate that the company looks set to use both high-performance Phoenix processors based on Zen 4 cores and compact Phoenix 2 silicon featuring Zen 4 and Zen 4c cores. Based on leaked information from HKEPC, AMD is set to offer at least four Ryzen 8000G APUs for midrange and entry level desktops.

AMD's next-generation desktop APU will include four models, Ryzen 3 8300G, Ryzen 5 8500G, Ryzen 5 8600G, and Ryzen 7 8700G, according to data from AGESA Combo AM5 PI 1.1.0.0 firmware. Keep in mind that the information comes from unofficial source, so take it with a pinch of salt.

The Ryzen 3 8300G and Ryzen 5 8500G models claim to be based on AMD's Phoenix 2 silicon with up to 6 cores and Radeon 740M graphics whereas the Ryzen 5 8600G and Ryzen 7 8700G are expected to use the more expensive Phoenix with up to eight Zen 4 cores and Radeon 780M graphics. The latter will naturally offers higher performance both for general-purpose computing and for graphics. In addition, Phoenix integrates Ryzen AI accelerator for machine learning workloads.

As for the launch timeframe of AMD's Ryzen 8000G, insiders reportedly told HKEPC that AMD has yet to finalize its release schedule and communicate it to its partners. At present, it is suggested that the company could release its next-generation desktop APUs either late this year or potentially early in 2024.

Considering the fact that AMD tends to unveil new products at CES, it is likely that the new Ryzen 8000G will be formally introduce in early 2024. This scenario looks logical as if AMD planned to release all-new AM5 APUs this year, its partners would be plenty of Ryzen 8000G samples.

When AMD introduced its Ryzen Z1 and Ryzen Z1 Extreme accelerated processing units (APUs) for handheld game consoles earlier this year it made it clear than the two processors are very different. But a new CPUID dump discovered by @InstLatX64 shows that the vanilla Ryzen Z1 is actually based on Phoenix 2 silicon with Zen 4 and Zen 4c cores.

AMD's Ryzen 7040-series (Phoenix) APU boasts eight high-performance Zen 4 cores and a GPU with 12 compute units (768 stream processors), while the Phoenix 2 has a mix of six cores: two high-performance Zen 4 cores and four energy-efficient Zen 4c cores that are notably more compact. Meanwhile, the Phoenix 2 also integrates a Radeon GPU with only four compute units (256 stream processors) that is about three times less powerful than the one used in the Z1 Extreme. 

Configuration of Phoenix 2 suggests that AMD is targeting this silicon for more budget-friendly laptops rather than for gaming solutions and yet Ryzen Z1 is officially positioned as an APU for gaming consoles.

Right now, the only device that uses AMD's Ryzen Z1 is the cheap Asus ROG Ally version that is priced at $599. Yet, the full-fat Asus ROG Ally based on Ryzen Z1 Extreme costs $699 and it is unclear whether a $100 lower price justifies three times lower GPU performance and a lower-end CPU configuration.

While we can only wonder how significantly Zen 4c cores are slower than Zen 4 cores in case of Ryzen Z1 versus Ryzen Z1 Extreme, the clear performance bottleneck of this chip in games is evidently its GPU rather than CPU.

An alleged die shot of AMD’s upcoming hybrid accelerated processing unit with Zen 4 and Zen 4c general-purpose cores was published by hardware leaker HXL. The APU is expected to be the younger brother of AMD’s Phoenix and will sit below these parts in the Ryzen lineup.

The supposed Phoenix 2 processor die shot clearly shows a large L3 cache (in the middle of the left side, green structures), two high-performance Zen 4 cores (below the L3 cache), four smaller Zen 4c cores (three above L3, one to the right of Zen 2 cores), and a large built-in GPU (on the right side). In addition, there are several DDR5/LPDDR5 PHY interfaces on top of the die, as well as PCIe, USB, and other physical interfaces on the APU's bottom and left and right sides of the APU.

While AMD’s Ryzen 7040-series (Phoenix) APU carries eight high-performance Zen 4 cores, Phoenix 2 only features six cores, two high-performance Zen 4 cores, and four energy-efficient Zen 4c cores that are considerably smaller than their counterparts, which points to the fact that AMD positions Phoenix 2 as a lower-cost APU for cheaper laptops. What remains to be seen is what else AMD cut down from this chip to make it smaller and more affordable, but it looks like we will learn this when AMD officially launches these parts.

AMD's first hybrid CPU design will bolster its competitiveness against Intel's Alder Lake and Raptor Lake offerings in the laptop processor segment. By incorporating a mix of high-performance and high-efficiency cores, AMD can achieve performance metrics at power levels previously unattainable. Meanwhile, given the specifics of Zen 4c and Phoenix 2 design, it will be considerably cheaper than the original Phoenix, opening some new doors for Zen 4.

AMD's codenamed Phoenix 2 processors were expected to hit the market this year, but it is September, and we do not see this APU anywhere, so perhaps AMD decided to make it a part of its Ryzen 8000-series lineup set to arrive in 2024. Meanwhile, if AMD's Phoenix 2 comes to market in 2024, it will have to rival Intel's entry-level Meteor Lake, which is projected to be considerably more competitive than the current offerings.

AMD's next-gen Ryzen 8000-series 'Strix Point' accelerated processing unit may pack quite a punch with twelve Zen 5-class cores and a reworked RDNA 3.5-based integrated GPU if a leak from Golden Pig Upgrade Pack is to be believed (via HXL). While Golden Pig Upgrade Pack has a good reputation, the information is unofficial and should be taken with discretion.

According to the leak, AMD's Ryzen 8000 'Strix Point' APU will pack four fully-fledged Zen 5 cores with 16 MB L3 cache, eight Zen 5c cores with 8 MB L3 cache, and an integrated GPU with 8 WGPs (1024 stream processors) based on the RDNA 3.5 architecture into a big monolithic die. The CPU cores will purportedly be distributed over two core complexes (CCX).

The Strix Point APU will be AMD's first hybrid processor featuring 'big' and 'little' cores, so it will be interesting to see how the processor will stack up against AMD’s existing offerings as well as Intel’s Meteor Lake offerings, which are said to come with up to 14 cores that will process up to 20 threads simultaneously. Meanwhile, since both Zen 5 and Zen 5c cores allegedly support threading, Strix Point will be able to process up to 24 threads at once.

The information about 12 Zen 5 cores in the upcoming APU has been reported before, so some might consider the new information as yet another sign that AMD is working on such an APU. Meanwhile, it is entirely possible that the company will have Strix Point APUs with different configurations, though we'll refrain from making assumptions about possible configs.

Arguably the most intriguing part of the Strix Point leak is that the APU will feature a graphics processor with eight RDNA 3.5-class WGPs, which points to 1024 ALUs. We can only guess about the clock rate of the iGPU as well as its intended performance, but the fact that the company decided to increase the number of stream processors in its built-in graphics processor indicates that there is a plan to offer higher performance with this unit.

AMD's Ryzen 8000-series Strix Point processors for laptops are expected to hit the market in 2024, so it is highly likely that some of AMD's partners are already test-driving them. Therefore, it is not surprising that there are leaks of their specifications, and chances that the APUs will indeed feature 12 cores and a potent iGPU are fairly high.

Nonetheless, at this point, take the information with a grain of salt since it comes from an unofficial source and can be inaccurate.

AMD is keeping the pedal to the metal, as evidenced by a host of new processors name-checked in a DigiTimes report today. Probably of greatest interest to PC DIY enthusiasts will be the new Ryzen 8000 Granite Ridge processors, purportedly delivering Zen 5 processor cores in late 2024. However, AMD isn’t expected to migrate desktop and mobile chips to TSMC processes more advanced than 4nm until 2026, according to industry insider information.

Embedded above, you can see an AMD processor roadmap put together by Twitter’s @Harukaze5719 using data from DigiTimes, and some supplemental seasoning from YouTuber MLID.

Working from the top, AMD is expected to refresh its server offerings with a new series of Epyc processors in 2025. Insiders indicate this could be the first AMD chip fabbed at 3nm. There are no other 3nm chips on this roadmap.

HEDT users will also enjoy a refresh in 2025 when the Shimada Peak Threadripper processors debut if this roadmap is accurate. It isn’t certain whether these powerful high core count chips will use TSMC’s 4nm or 6nm processes, however, they will likely be based on the Zen 5 architecture. Meanwhile, we are still waiting for the Storm Peak Threadripper CPUs, based on the Zen 4 architecture, which are expected in Q3 this year.

Luckily for mainstream desktop PC users, there isn’t such a big gap until a significant refresh arrives on their platform of choice. According to this roadmap, distilled from the utterings of DigiTimes sources, Ryzen 8000 Granite Ridge processors burst onto the scene in 2024. Specifically, DigiTimes says Granite Ridge will be released “at the end of 2024.” This will likely give AMD desktop users the first taste of the Zen 5 architecture, but it is again uncertain what TSMC process node we will be looking at (4nm and/or 6nm).

The roadmap gets very busy within the mobile space, reflecting the importance of providing options appealing to as wide a range of the populace as possible. Even desktop users might admit that the currently emerging Ryzen 7000 mobile APUs are quite exciting, with some combining the attractions of Zen 4 CPU cores and RDNA 3 GPU cores in a portable-friendly TDP. We expect the Ryzen 8000 mobile chips to keep AMD on your mind if you are considering a new laptop.

In 2024, the Dragon Range and Phoenix Point Ryzen 7000 mobile chips will be replaced by an even bigger family of targeted Ryzen 8000 APUs, with most moving to the newest Zen 5 CPU cores. Then, by 2025, all the AMD mobile APU options will be Zen 5-based, except perhaps an entry-level chip codenamed Escher. Sadly, the roadmap does not mention what GPU architectures will be used, as this is an essential consideration with mobile platforms with limited upgrade options.

Lastly, with reports like this based on industry leaks and rumors, please add a pinch of salt.

When Gigabyte's enterprise unit revealed last week that AMD would introduce a successor to its Ryzen 7000-series CPUs this year, it made quite a splash among enthusiasts. But the company has removed the claim about the launch timeframe of AMD's Ryzen 8000-series products and changed the wording of its press release since then, as noted @9550pro and ComputerBase. 

"[We] do not know when the Ryzen 7000 successor will be released," Liam Quinn, a spokesman for Gigabyte, told TechRadar, confirming that the phrasing was a mistake. "[We will] make a comment on the press release to clarify but keep our original wording." 

AMD's official roadmap slides released in mid-2022 indicate that the company planned to release CPUs based on its Zen 5 and Zen 5c microarchitectures by 2024 (the year in the bottom right corner of AMD's roadmap slides typically indicates 'by the beginning of' rather than 'by the end of,' unless AMD changed something in its way of presenting data), which means that it was reasonable to expect Zen5-powered Ryzen 8000-series processors this year. 

In fact, AMD's public client roadmaps clearly indicate that its Zen 5-based Strix Point CPU for notebooks and Granite Ridge CPU for desktops will be released by 2024. This is perhaps why Gigabyte stated this in its press release.

However, a Zen 5/Zen 5c launch in late 2023 would not exactly fit AMD's cadence of releasing new microarchitectures for client PCs roughly every two years. Furthermore, given the slow demand for PCs, releasing an all-new CPU family this year might be counterproductive. To that end, even if AMD previously had plans to launch Zen 5-based products in calendar 2023, such plans might have changed by now.  

"The next generation of AMD Ryzen desktop processors that will come out later this year will also be supported on this AM5 platform, so customers who purchase these servers today have the opportunity to upgrade to the Ryzen 7000 series successor," the original press release read. 

"The next generation of AMD Ryzen desktop processors will also be supported on this AM5 platform, so customers who purchase these servers today have the opportunity to upgrade to the Ryzen 7000 series successor," the altered statement reads.

AMD's upcoming codenamed Phoenix accelerated processing units (APUs) are set to feature a so-called hybrid design containing both high-performance Zen 4 cores and  energy-efficient Zen 4c cores, according to an AMD's processor programming guide discovered by InstLatX64.

AMD's processor programming guide clearly points to Performance Cores marked at 01h and Efficiency Cores descripted as 1h, an evident indicator that AMD's upcoming processor will feature two types of general-purpose cores. Given the timing, we believe that we are dealing with AMD's codenamed Phoenix processor featuring Zen 4 and Zen 4 cores, but keep in mind that AMD calls them otherwise.

The document sheds some light on how AMD's hybrid Big.Little-like design will work. Just as earlier AMD's documents pointed out, the company's high-performance and energy-efficient cores feature a different feature set. Therefore, software makers are advised to design their programs accordingly. 

"Read-only. Reset: Fixed,Xh. Defines per-core architectural feature differentiation (microarchitectural resources, etc.) that may lead to a different performance, core clock boost, and power characteristic," a statement in the AMD PPR reads.

Of course, a hybrid CPU design will make AMD's offerings more competitive with Intel's Alder Lake and Raptor Lake products in the laptop CPU market. Usage of a hybrid design containing both high-performance and high-efficiency cores will enable AMD to hit performance levels at wattages not attainable today. Meanwhile efficient usage of low-power cores depends on software in general and operating system in particular. 

When we reported about AMD's alleged Phoenix 2 processor with a hybrid design a week ago, it left us wondering why would AMD use the same — Phoenix APU — codename for completely distinctive processors. Apparently, these CPUs are not that different as, according to AMD's programing guide, they both feature a Big.Little-like design.

Anyhow, AMD's codenamed Phoenix processors are expected to hit the market this year, so expect independent tests of these units at Tom's Hardware.

An unannounced AMD processor identified as Family 25 Model 120 Stepping 0 recently showed up in the MilkyWay@Home database. The CPU can process 12 threads simultaneously and the CPU expert @InstLatX64 believes that this is AMD's codenamed Phoenix 2 processor, packing two high-performance Zen 4 cores and four energy-efficient Zen 4c cores. 

AMD Eng Sample processor marked 100-000000931-21_N [Family 25 Model 120 Stepping 0] features 12 logical cores (i.e., six physical cores with simultaneous multithreading) and reports about 1MB of cache, which indicates that the MilkyWay@Home client cannot correctly determine the amount of cache featured by the chip. The listing itself does not prove that we are dealing with AMD's hybrid Phoenix 2 processor with Big.Little-like core configuration, but six physical/12 logical cores featured by an unknown CPU gives us a hint that this may match the rumors.

AMD's Phoenix 2 processor (which does not have a lot in common with the company's Phoenix APU) is rumored to feature two 'big' Zen 4 cores with 2MB L2 and 4MB L3 cache as well as four 'small' Zen 4c cores equipped with 4MB L2 and 4MB L3 cache, which is a rather surprising cache configuration. The APU is also said to pack an RDNA 3-based integrated GPU with 512 stream processors and has a DDR5/LPDDR5X-supporting memory subsystem, according to 3DCenter. @InstLatX64 claims that AMD's Phoenix 2 APU has an A70F8x CPUID, whereas CoelacanthDream asserts that the CPUID of the processor is 0x00a70f80. 

For now, any information about Phoenix 2 in general and the 100-000000931-21_N [Family 25 Model 120 Stepping 0]processor in particular should be taken with a grain of salt since AMD has loads of products in the pipeline. 

The alleged Phoenix 2 processor with two Zen 4 cores and four Zen 4c cores has been running MilkyWay@Home client since early March, which indicates that someone within AMD or even outside of the company is test driving the chip. This may be a sign that the CPU will be released in the foreseeable future, though it is unclear when exactly. Meanwhile, based on unofficial information, AMD is set to release its Phoenix 2 APUs in the second half of 2023.

Yesterday's AMD RDNA 3 leak not only exposed three discrete Navi 30-series graphics processors based on the next-generation architecture but also revealed one integrated GPU. The integrated GPU will be used for an accelerated processing unit codenamed Phoenix, and this chip appears to pack quite a punch if unofficial information about it is correct.

AMD's codenamed Phoenix APU appears to be based on the company's upcoming Zen 4 microarchitecture, but the number of cores is unknown (we would speculate about eight Zen 4 cores, though we would keep in mind that AMD also has cut-down Zen 4c cores). The graphics part of the APU seems even more interesting as it uses the upcoming RDNA 3 architecture and may pack as many as 16 GPU compute units. In the case of RDNA 3, that translates into 1536 stream processors (today's APUs feature 512 SPs), according to ComputerBase, citing various unofficial sources. Of course, AMD won't talk about its Phoenix APU any time soon.

Phoenix could have a GPU offering performance akin to a current-generation discrete Radeon RX 6500 XT graphics chip, or perhaps even closer to the RX 6600, depending on what RDNA 3 brings to the table. Meanwhile, to provide enough memory bandwidth for such a monstrous integrated GPU, AMD intends to support DDR5 and LPDDR5. The APU is also said to support a PCIe 5.0 interface. Unfortunately for desktop gamers, AMD seemingly only plans to offer Phoenix for notebooks and perhaps compact or all-in-one desktops.

In recent years, AMD has packed its latest Zen CPU cores into its APUs to make them competitive in general-purpose workloads. However, the GPU part of those APUs was based on the Radeon RX Vega design, which uses a highly modified GCN (graphics core next) architecture from 2011. As a result, while these APUs are good enough for everyday workloads, they are barely adequate for gaming and media (because of outdated video decoding/encoding units and video processing engines).

With its upcoming Ryzen 6000-series 'Rembrandt,' AMD finally weds its up-to-date Zen 3+ cores with a contemporary RDNA 2-based GPU. However, the combination of Zen 4 and RDNA 3 featured by the future Phoenix APU seems even more appealing. The only question is when exactly this chip will become available.

Another question is which process technology (or technologies?) will be used to make AMD's Phoenix. AMD's Zen 4 cores are designed for TSMC's N5 node, whereas the company's RDNA 3 GPUs are rumored to use N5 and/or N6. Given AMD's focus on Zen, we are pretty sure that the Zen 4-based Phoenix will use N5, just like Raphael. Meanwhile, keeping in mind that AMD tends to re-use its GPU IP in APUs without many changes, we are less sure how things will play out there. Throwing an N5 GPU IP into an APU may not be feasible from a manufacturing costs perspective, whereas redesigning an N6 IP may be too expensive. Perhaps AMD will switch to a tiled design for its APUs, but that's pure speculation.

Twitter user Kepler has discovered a new AMD patent application that details moving tasks (threads) between different types of cores in a heterogeneous processor. In other words, the patent lays the groundwork for a microarchitecture that resembles Arm's big.LITTLE design that uses clusters of 'big' high-performance cores paired with 'little' efficiency cores. 

AMD files over a plethora of patent applications every year, so there's no guarantee that all of them will manifest as real products. That said, there are plenty of reasons to think that we could see a hybrid AMD design come to market. Intel has already embraced a hybrid design on its desktop parts with the upcoming 12th-Gen Alder Lake family, and it's reasonable to assume that AMD will eventually hop on the hybrid bandwagon at some point in time.

Although the Method of Task Transition Between Heterogeneous Processors patent application was just published a couple of days ago, AMD filed it back in 2019. This patent may be an extension of a similar application that AMD also filed in the same year about implementing instruction set architecture (ISA) in a heterogeneous processor.

There's an ongoing rumor that AMD's Ryzen 8000 (reportedly codename Strix Point) APUs could arrive with a hybrid setup. The chips allegedly feature high-performance Zen 5 cores and low-powered Zen '4D' cores. Unless AMD has been diligently working behind the scenes, it's unlikely that Strix Point will make it to the market in time to compete with Intel's Alder Lake chips that may launch in late 2021 or early 2022. However, the APUs will probably go head-to-head with Raptor Lake, the alleged successor to Alder Lake.

The patent explains that the process to relocate a task or tasks from the first processor to the second processor will be based around performance metrics based on certain thresholds or some other trigger. AMD didn't specify which cores are which, but for the sake of conversation, we can assume that the first processor refers to the big cores and the second processor refers to the power-efficient cores.

Obviously, the whole point behind a hybrid configuration is to optimize performance-per-watt while also improving performance. To achieve this goal, tasks must be moved quickly and efficiently between the big and small cores. AMD's method consists of comparing one or multiple metrics to a threshold on a checklist to determine whether or not to pass the task from one processor to another. Once the assessment is complete, the first processor essentially pauses operations while the information is transferred over to the second processor.

AMD mentions numerous examples of the type of metrics that the chipmaker could leverage for the task relocation process. The chipmaker mentions task execution time, core utilization, memory usage, idle state of a single core, or duration of a single-core execution - just to mention a few scenarios.

In one example, AMD measures the period of time that the small cores are running at the maximal clock speed and compares it to a threshold. If the duration is greater than the established time threshold, the task shifts over to the bigger cores. In another example, AMD takes into account an external factor: memory. If the memory utilization is less than the threshold established on the small cores, the task will remain on said cores.

Hybrid processors won't succeed unless there is proper software support. Recent rumors point to a new, more efficient scheduler in Windows 11 that's optimized for hybrid setups. That new update is rumored to land later this year at the same time as Alder Lake, which should pave the way for better support for hybrid processors. 

AMD's Zen 4 processors haven't even launched yet, and rumors are already emerging about the chipmaker's future Zen 5 chips. Given the time frame that we expect for the Zen 5 chips to debut, the information from the leaks makes sense. Nevertheless, we still recommend you approach all rumors with caution.

The partial AMD roadmap, which originated from China, claims that AMD will market its Zen 5 processors under the Ryzen 8000 branding, with both mainstream Ryzen processors and APUs reportedly uniting under the Ryzen 8000 branding umbrella. The rumored codenames for Ryzen 8000 chips and APUs are Granite Ridge and Strix Point, respectively.

ExecutableFix, a hardware leaker with a solid record, believes that Zen 5 is based on the 3nm process node. Given AMD's long-lasting relationship with TSMC, it's very likely that Zen 5 will tap into the foundry's 3nm (N3) manufacturing process. Then again, Samsung also has its own 3nm process node, so it'll be interesting to see whether the South Korean giant can woo AMD away from TSMC.

Granite Ridge is the successor to Raphael (Zen 4), so it should be compatible with the AM5 socket, which is allegedly making a transition from a Pin Grid Array (PGA) design over to a Land Grid Array (LGA) design. The general speculation for Zen 4's debut is sometime next year, meaning we shouldn't see Zen 5 until 2023.

Strix Point is rumored to feature a hybrid microarchitecture, similar to what Intel will do with its impending 12th Generation Alder Lake processors. In AMD's case, Ryzen 8000 APUs could leverage the Zen 5 and Zen 4 cores, implying that AMD's APUs may finally switch to a multi-chip module (MCM) approach. 

Apparently, the little cores inside Strix Point, if you want to put it that way, are called Zen 4D. The only problem we have with this leak is the discrepancy between manufacturing processes. Zen 4 is expected to feature the 5nm process node, while Zen 5 is presumably based on the 3nm process node. However, we don't discard the possibility that AMD could refresh the Zen 4 cores to put them on the 3nm process node.

On the server side, the speculation says the EPYC 7005 (purportedly codenamed Turin) lineup will replace the EPYC 7004 (Genoa) family in 2023. Logically, Turing will also be on Zen 5 cores and the 3nm process node. Threadripper's fate, on the other hand, continues to be a mystery for all.