Only one laptop will get AMD's best mobile gaming processor ever, and that's a huge shame

https://ift.tt/ZJ92r6D

It seems that the rumor mill was right about AMD’s latest mobile gaming processor, as the tech giant officially announced the AMD Ryzen 9 7945HX3D. But there’s a catch.

The only caveat is that this processor will be available for a single laptop, the Asus ROG Strix Scar 17 X3D. And that's a shame as this is probably the best mobile chip ever, and the first one to use the excellent 3D V-Cache technology. It’s similar to the Ryzen 9 7945HX except that, according to PC Gamer, it has the full 64MB of L3 cache bonded on top of one of its eight-core chiplets. It also has the same TDP and the same peak 5.4GHz clock speed, if the chiplet doesn’t have the extra L3 of course.

According to AMD, the Ryzen 9 7945HX3D’s full specs are as follows: Zen 4 architecture, 16 cores and 32 threads, L2 cache of 16GB and L3 cache of 128GB, TDP between 55 and 75W, and an AMD Radeon 610M GPU.

The AMD Ryzen 9 7945HX3D processor will be launching on August 22nd, 2023, which is just after Gamescom.

AMD is wasting a great opportunity 

Considering how incredible the 3D V-Cache technology is, as it manages to get extra performance out of a chip without having to bump up the amount of energy or cooling that chip would normally need, it’s a shame that only one laptop will get to benefit from it.

Especially since this tech is ideal for portable machines, as it allows some of the best gaming laptops to get great performance without having to risk additional ventilation issues.

And this chip is even more valuable to AMD because it could give both the Intel Raptor Lake-HX and the upcoming 14th-gen CPU refresh a run for the money. It really seems that AMD’s Dragon Range series could be one of the best AMD processors and even one of the best processors in general. So it feels like Team Red is squandering a real chance to push Team Blue’s hold on the processor market.

And this isn’t the only time AMD has done this either. Previously it released the AMD Ryzen 5 5600X3D processor in limited supply only at Micro Center, which was a baffling decision to say the least. And back in 2022, the tech giant launched the AMD Ryzen 7 5800X3D but it was produced in low volume and was the only 5000-series chip to sport the extra cache until 2023.

from TechRadar: computing components news https://ift.tt/yloRMzC
via IFTTT

My Radeon RX 7900 XT just got a lot more powerful, for free – thanks, AMD!

https://ift.tt/re59ybG

Let’s be honest: GPU driver updates are rarely a source of much excitement. We’re normally talking about tiny, marginal improvements, or simply fresh support for performance-boosting software in newly-released games. But every now and then, we see an update that makes a genuine difference – and that’s what we’ve seen today.

A new graphics driver update for AMD’s Radeon Software Adrenalin Edition has shown a whopping 67% performance improvement for the Radeon RX 7900 XT. Granted, that’s just one specific GPU playing one specific game at one specific resolution (Forza Horizon 5 at 1080p), but it’s still a wildly impressive leap in framerate.

The flagship Radeon RX 7900 XTX also gets a big boost in Forza Horizon 5 and The Last of Us, though it’s not quite as pronounced. Looking at a wider slew of results as tested by YouTuber Ancient Gameplays, performance boosts are a bit less impressive at higher resolutions, with an average framerate increase of 23-24% at 4K for both cards (although, since testing was performed with a Ryzen 7 CPU, it’s possible there was a processor bottleneck at play here).

Driver updates aren’t all Adrenalin is good for

While most of the best graphics cards – certainly the most popular ones, anyway – might be made by Nvidia, it’s fair to say that AMD’s software interface has the lead. Adrenalin offers a more robust feature set than Nvidia’s GeForce Experience software (though Nvidia’s less complex offering might be preferable to users who prefer to simply hit ‘optimize settings’ and jump straight into a game).

These latest drivers really put Adrenalin in the spotlight. It’s rare to see performance leaps this significant purely from a free software update and it’s an interesting testament to the effectiveness of good driver support. Nvidia and AMD both routinely release new graphics drivers, but they sometimes feel more like an afterthought than a vital improvement.

Part of why these sorts of driver updates are important is how they impact older GPUs. AMD’s last-gen RX 6000 graphics cards also benefited from a 10-25% performance boost in the relevant games with this new update, and some of those GPUs are close to three years old now. There’s a constant push from PC hardware manufacturers to upgrade and buy new components, but performance boosts this significant purely from free, instantly downloadable software tweaks indicate that maybe, just maybe, nobody actually needs to buy a new GPU every two years.

Think about it. AMD’s recently-released RX 7600 outperformed its predecessor the RX 6600 by about 19% in our testing process. If that sort of performance boost in the latest games could be secured for the 6600 without a hardware upgrade, why bother with the 7600 at all? It’s a question some GPU makers probably don’t want us asking – especially with two hotly-anticipated AMD GPUs now very close to release.

from TechRadar: computing components news https://ift.tt/M0uIeym
via IFTTT

Possible upcoming AMD Ryzen 9 7950X3D CPU could give Intel a run for its money

https://ift.tt/re59ybG

According to new rumors along the tech grapevine, the AMD Ryzen 9 7945HX3D CPU has been spotted in the wild along with some possible specs.

The Ryzen 9 7945HX3D might be based on the Zen 4 core architecture with 16 cores, 32 threads, and support for PBO overclocking and tuning capabilities. The leaker's Weibo tech account (via Twitter user HXL) also said that it might have a 3D V-Cache offering 128 MB capacities, with “dual-CCD configuration with 64 MB of stacked cache on one of the two CCDs and the rest of the 64 MB cache coming from the two Zen 4 dies.”

Another Twitter user, Everest, seems to have unearthed more specs through an Asus ROG Strix Scar 17 laptop spec sheet. From there we find out that clocks speeds reach up to 5.4 GHz and has a configurable TDP between 55W and 75W.

Judging from these leaks, the Ryzen 9 7945HX3D will likely be for high-end laptops. The ROG Strix Scar 17 listed in the specs sheet has an Nvidia RTX 4090, 32GB of DDR5 memory, and a 240Hz IPS display. It’s also listed through an Australian retailer for AU$5,599 (about $3,780 / £2,945), which, while pricey, makes sense considering the components.

AMD could bring the heat to Intel 

What makes this rumor so interesting is that, going by the possible specs, the Ryzen 9 7945HX3D could be giving the Intel Raptor Lake-HX and the upcoming 14th-gen CPU refresh a run for its money. And considering what an excellent CPU the AMD Ryzen 9 7950X3D turned out to be, this gives the 7945HX3D even more credibility.

Not to mention that the AMD Ryzen 9 7945HX was already very close to the Intel 13th Gen Core i9-HX series,, according to preliminary benchmark scores. AMD’s Dragon Range series of CPUs could be shaping up to be not only one of the best AMD processors but also one of the best processors in general.

Pricing might be the biggest obstacle for the 7945HX3D. The 7950X3D is more expensive, at $699, than the Intel Core i9-13900K, which has an MSRP of $589 / £699 / AU$929. No matter how solid the performance, it could be a hard sell to get cash-strapped PC gamers to purchase these CPUs, since the upgrade would only make sense for those with an AMD Zen 3 or 11th-gen Intel processor or earlier. Budget-minded buyers would be more likely to look for a cheaper option. 

from TechRadar: computing components news https://ift.tt/jBAUtCa
via IFTTT

Ultra Ethernet Consortium Formed, Plans to Adapt Ethernet for AI and HPC Needs

https://ift.tt/ieanJPW

This week the Linux Foundation has announced that the group will be overseeing the formation of a new Ethernet consortium, with a focus on adapting and refining the technology for high performance computing workloads. Backed by founding members AMD, Arista, Broadcom, Cisco, Eviden, HPE, Intel, Meta and Microsoft, the new Ultra Ethernet Consortium will be working to improve Ethernet to meet the low latency and scalability requirements that HPC and AI systems need – and which the group says current Ethernet technology isn't quite up to the task for.

The top priority of the new group will be to define and develop what they are calling the Ultra Ethernet Transport (UET) protocol, a new transport-layer protocol for Ethernet that will better address needs of AI and then HPC workloads.

Ethernet is certainly one of the most ubiquitous technologies around, but demands of AI and HPC clusters are growing so fast that the technology will run out of steam in the future. The size of large AI models is increasing rapidly. GPT-3 was trained with 175 billion of parameters back in 2020. Today GPT-4 is said to be accommodating already a trillion of parameters. Models with the larger number of parameters require larger clusters and then these clusters send larger messages over the network. As a result, the higher bandwidth and the shorter latency these network feature, the more efficient the cluster can operate.

"Many HPC and AI users are finding it difficult to obtain the full performance from their systems due to weaknesses in the system interconnect capabilities," said Dr. Earl Joseph, CEO of Hyperion Research.

At a high level, the new Ultra Ethernet Consortium is looking to refine Ethernet in a surgical manner, improving and altering only those bits and pieces necessary to achieve their goals. At its onset, the consortium is looking at improving both the software and physical layers of Ethernet technology — but without altering its basic structure to ensure cost efficiency and interoperability.

Technical goals of the consortium include developing specifications, APIs, and source code to define protocols, interfaces, and data structures for Ultra Ethernet communications. In addition, the consortium aims to update existing link and transport protocols and create new telemetry, signaling, security, and congestion mechanisms to better address needs of large AI and HPC clusters. Meanwhile, since AI and HPC workloads have a number of differences, UET will have separate profiles for appropriate deployments.

"Generative AI workloads will require us to architect our networks for supercomputing scale and performance," said Justin Hotard, executive vice president and general manager, HPC & AI, at Hewlett Packard Enterprise. "The importance of the Ultra Ethernet Consortium is to develop an open, scalable, and cost-effective ethernet-based communication stack that can support these high-performance workloads to run efficiently. The ubiquity and interoperability of ethernet will provide customers with choice, and the performance to handle a variety of data intensive workloads, including simulations, and the training and tuning of AI models." 

The Ultra Ethernet Consortium is hosted by the Linux Foundation, though the real work will be undertaken by its members. Between AMD, Cisco, Intel, and other founders, these companies all either design high-performance CPUs, compute GPUs, and network infrastructure for AI and HPC workloads or build supercomputers or clusters for AI and HPC applications, thus have plenty of experience with appropriate technologies. The work of UEC is set to be conducted by four working groups that will work on Physical Layer, Link Layer, Transport Layer, and Software Layer.

And while the group is not explicitly talking about Ultra Ethernet in relation to any competing technologies, the members of the founding board – or rather, who's not a founding member – is telling. The performance goals and HPC focus of Ultra Ethernet would have it coming into direct competition with InfiniBand, which has for over a decade been the networking technology of choice for low-latency, HPC-style networks. While developed by its own trade association, NVIDIA is said to have an outsized influence on the group vis-a-vie their Mellanox acquisition a few years ago, and they are noticeably the odd man out of the new group. The company makes significant use of both Ethernet and InfiniBand internally, using both for their scalable DGX SuperPod systems.

As for the proposed Ultra Ethernet standards, UEC members are already plotting plans how to integrate the upcoming UET technology into their products.

"We are particularly encouraged by the improved transport layer of UEC and believe our portfolio is primed to take advantage of it," said Mark Papermaster, CTO of AMD in a blog post. "UEC allows for packet-spraying delivery across multiple paths without causing congestion or head-of-line blocking, which will enable our processors to successfully share data across clusters with minimal incast issues or the need for centralized load-balancing. Lastly, UEC accommodates built-in security for AI and HPC workloads that in turn help AMD capitalize on our robust security and encryption capabilities."

Meanwhile, for now UEC does not say when it expects to finalize the UET specification. It's expected that the group will seek certification from the IEEE, who maintains the various Ethernet standards, so there is an additional set of hoops to jump through there.

Finally, the UEC has noted that it is looking for additional members to round out the group, and will begin accepting new member applications from Q4 2023. Along with NVIDIA, there are several other tech giants involved in AI or HPC work that are not part of the group, so that would be their next best chance to join the consortium.

Source: The Linux Foundation, The Register

­­­­

from AnandTech https://ift.tt/CWUYjt0
via IFTTT

AMD Ryzen 8000-series CPUs built on Zen 5 might beat Intel's Arrow Lake to market

https://ift.tt/epaJC5Q

AMD Ryzen 8000-series CPUs have hit a major milestone on the way to launch, as major industry benchmarking utility AIDA64 is now able to recognize AMD's new Zen 5 microarchitecture.

The new AMD architecture popped up in FireWire's most recent update release notes to its popular AIDA64 tool, which many professionals and PC enthusiasts use to help measure the health and performance of their system. FireWire couldn't have done this on its own, and it would have needed details from AMD to introduce preliminary support for the architecture.

This doesn't mean we have any details on when the Zen 5 architecture will hit the store shelves in the form of AMD Ryzen 8000-series desktop processors, as Tom's Hardware notes. AIDA64 introduced preliminary support for Intel's Meteor Lake processors all the way back in 2021, and we aren't expecting these chips until later this year.

AMD's roadmap for releases puts the Ryzen 8000-series launch sometime in 2024, though, so there's every reason to think that AMD is on track with its release schedule.

Improved support for Meteor Lake and Lunar Lake, but AMD still looks to gain on Intel

The release notes also indicate that there is improved support for Intel's future processor generations: Meteor Lake (purported to be mobile-only), Arrow Lake (Desktop and mobile), and Lunar Lake.

Lunar Lake is set for a late-2024/early-2025 release, and will introduce a whole new microarchitecture for Team Blue's chips. The upcoming 14th-gen Raptor Lake refresh and Meteor Lake chips – with both falling under the 14th-gen label and expected to land later this year (with the latter being reserved for mobile devices) – and Arrow Lake, expected in 2024, will be sticking with the hybrid performance-efficiency core design of Intel's current Raptor Lake chips.

Arrow Lake, which is set to fab on TSMC's 3nm process node, has been rumored to be dealing with delays, something not entirely unheard of when dealing with TSMC, which fabs the processor wafers for AMD, Nvidia, Apple, and many others. With 3nm set to be an extremely popular node, production schedules are sure to be tight.

Zen 5, meanwhile, will also be using TSMC's 3nm node, which could be both a blessing and a curse for AMD. Intel locked down a lot of 3nm production capacity back in 2021, ostensibly for an Arc GPU and several server chips. 

If Intel is swapping in Arrow Lake chips instead, this means Intel is likely to have more stock of its chips when Arrow Lake launches than AMD has of its own 3nm Zen 5 chips. On the bright side, it looks like AMD got its orders into TSMC for Zen 5 first, so its likely that Zen 5 chips will be rolling off the line well before Intel's Arrow Lake chips do. 

Being first to market has its advantages, but AMD Zen 4 also beat Intel to market back in September 2022, only to be quickly outshined by Intel Raptor Lake chips within a month. Whether history repeats itself remains to be seen.

from TechRadar: computing components news https://ift.tt/VeLsyGg
via IFTTT

Intel Set to Exit NUC PC Business - Pushes Partners to Develop More SFF PCs

https://ift.tt/SXYe0r8

Intel has disclosed today that it will halt further development of its small form factor Next Unit of Compute (NUC) PCs. The tech giant expects its partners to take over and keep serving markets served by its NUC systems as it focuses on much more profitable chips businesses. 

"We have decided to stop direct investment in the Next Unit of Compute (NUC) Business and pivot our strategy to enable our ecosystem partners to continue NUC innovation and growth," a statement by Intel reads. "This decision will not impact the remainder of Intel’s Client Computing Group (CCG) or Network and Edge Computing (NEX) businesses. Furthermore, we are working with our partners and customers to ensure a smooth transition and fulfillment of all our current commitments – including ongoing support for NUC products currently in market."

Intel entered PC business with its ultra-compact NUC desktops in 2013, around the time it exited motherboards market. Initially, the company only targeted SOHO market with its NUC barebones and PCs, but eventually it greatly expanded its NUC range with systems aimed at corporate users that need things like remote management and appropriate support, and even gaming machines. 

Intel's NUC systems have garnered considerable popularity over the years, going toe-to-toe with similar offerings from established PC brands. While small form factor existed before the NUC (and will exist after), Intel's efforts to invigorate the space with its NUC designs were by and large successful, and a lot of the public experimentation we've seen done in the space over the last several years has come from Intel.

Nevertheless, Intel's OEM partners/rivals haven't been standing still during this time, and Intel has fallen behind in some areas. None of Intel's NUCs ever moved to being powered via USB Type-C (something ASUS already does), and Intel's more recent enthusiast NUCs have struggled to differentiate themselves from the competition. Overall, maintaining a wide variety of desktops and laptop platforms has been somewhat taxing (if not distracting) for Intel, whose primary focus lies in the semiconductor industry, rather than finished devices.

While Intel isn't citing any specific reasons in their decision to wrap up development of new NUC PCs, given the contracting PC market and the intense rivalry therein, we wouldn't be surprised if Intel was being rocked by the same market forces that have been putting a squeeze on other PC OEMs. Intel has already reduced its focus on NUCs in the recent years, never offering Performance versions of its 12^th and 13^th Generation NUCs – and we cannot say that that those machines were missed by the audience. Meanwhile, Intel's enthusiast-grade Extreme NUCs have evolved to be more like fully-fledged desktops rather than compact systems, getting farther and farther away from the NUC's tiny roots. And while the add-in card form-factor used by the NUC Extreme lineup has always looked promising, it is unclear whether they have even been a success for Intel.

Ultimately, as Intel has continued to shed and shutter non-core businesses, it is not entirely unexpected that Intel is axing its NUC program. In its place, the company is urging its OEM/ODM partners – whose bread and butter is designing and selling complete systems – to continue producing and innovating on compact machines for the small office/home office market, business clientele, and gamers. This leaves Intel free to refocus on the highly lucrative chip manufacturing business, as CEO Pat Gelsinger has made a priority over the past couple of years.

Intel's NUC will not be the first business divested by Intel in the recent years. To focus on development on leading-edge CPUs, GPUs, and other lucrative products, Intel left the NAND memory and SSD businesses, axed Optane SSDs, ceased development of notebook models, and even sold its prebuilt server business to MiTAC.

from AnandTech https://ift.tt/2J6H9ul
via IFTTT

This gamers secret to playing Valorant at 1200 FPS is AMDs X3D processors

https://ift.tt/TAryu7g

Good lord, those are some high framerates. Content creator @connorjaiye, best known for playing the fast-paced competitive shooter Valorant, put together a custom gaming PC with the help of AMD - and was able to net upwards of 1,200 fps in his favorite shooter.

The most fascinating part of the whole endeavor is that Connor didn’t even do this using the most high-end parts available; he used a previous-gen Radeon RX 6950 XT as his graphics card of choice, with a relatively bog-standard B650 motherboard from MSI pulling the build together.

I built a PC that gives me 1000FPSFull video below 👇#valorant #ad #amdryzen pic.twitter.com/9qPZN2xi4KJuly 5, 2023

See more

The secret sauce? It’s the AMD Ryzen 7 7800X3D processor, the latest next-gen CPU from AMD to offer Team Red’s nifty 3D V-cache technology. This allows for cache memory to be ‘stacked’ on the CPU die, producing far more memory for the CPU to instantly draw on without needing to access your system RAM.

With this, Connor was able to achieve a stable 900+ frames per second in Valorant, with peaks in excess of 1,200. It’s the sort of framerate that would’ve been unheard of just a few years ago, and a testament to the gaming capabilities of AMD’s X3D chips - even though they typically lose out to Intel’s competing CPUs in non-gaming workloads.

Is 3D V-cache the new frontier for PC gaming?

I already knew that AMD’s X3D chips were an exceedingly solid choice for custom gaming PCs, but this is probably the most impressive showcase of the technology to date. Intel’s i9-13900K might still rule the roost in terms of overall performance, but it’s hard to recommend anything but the 7800X3D as the best processor for gaming right now.

Sure, the CPU shouldn’t get all the credit; the RX 6950 XT was the most powerful card from AMD’s RX 6000 generation, although you can snap one up for around $580 / £585 / AU$900 right now, making it a lot cheaper than either AMD’s current flagship Radeon RX 7900 XTX or Nvidia’s competing powerhouse the RTX 4090.

Overall, chances are Connor’s build wasn’t that expensive - at least, not by gaming PC standards. I ran the numbers, and buying from Newegg in the US you’d be spending around $1,250 for his listed components - then using PCPartPicker, I calculated the total build cost assuming 32GB of DDR5 RAM, a 1TB SSD, and a liquid cooler for the CPU, and it ran to just shy of $1,700 (about £1,350 / AU$2,550)

If I was looking to achieve similar framerates, I could probably do it for a bit less - in fact, if you wanted to build a budget version of Connor’s framerate-blitzing PC, you could use the newly-released AMD Ryzen 5 5600X3D along with a cheaper graphics card and other lower-spec components to get that sweet sweet 3D V-cache goodness for a little over a thousand bucks. Be warned, though - the 5600X3D is a limited-edition chip only available from Micro Center in the US, which is kind of a bummer.

from TechRadar: computing components news https://ift.tt/agBcUiE
via IFTTT

El Capitan Installation Begins: First APU-based Exascale System Shaping Up For 2024

https://ift.tt/IAqcYsm

Lawrence Livermore National Laboratory had received the first components of its upcoming El Capitan supercomputer and begun to install them, the laboratory announced on Wednesday. The system is set to come online in mid-2024 and is expected to deliver performance of over 2 ExaFLOPS.

LLML's El Capitan is based on Cray's Shasta supercomputer architecture and will be built by HPE, just like two other exascale systems in the U.S., Frontier and Aurora. Unlike the first two exascale machines, which use a traditional discrete CPU plus discrete GPU configuration, the El Capitan supercomputer will be the first one based on AMD server-grade APUs that integrate both processor types in to a single, highly connected package.

AMD's Instinct MI300A APU incorporates both CPU and GPU chiplets, offering 24 general-purpose Zen 4 cores, compute GPUs powered by the CDNA 3 architecture, and 128 GB of unified on-package HBM3 memory. AMD has been internally evaluating its Instinct MI300A APU for months, and it appears that AMD and HPE are now ready to start installing the first pieces of hardware that make up El Capitan.

According to pictures released by the Lawrence Livermore National Laboratory, its engineers have already put a substantial number of servers into racks. Though LLNL's announcement leaves it unclear whether these are "completed" servers with production-quality silicon, or pre-production servers that will be filled out with production silicon at a later date. Notably, parts of Aurora were initially assembled with pre-production CPUs, which were only swapped out for Xeon CPU Max chips over the past couple of months. Given the amount of validation work required to stand-up a world-class supercomputer, AMD and HPE may be employing a similar strategy here.

"We have begun receiving & installing components for El Capitan, first #exascale #supercomputer," a Tweet by LLNL reads. "While we are still a ways from deploying it for national security purposes in 2024, it is exciting to see years of work becoming reality."

When it comes online in 2024, LLNL is expecting El Capitan to be the fastest supercomputer in the world. Though with its full specifications still being held back, it's not clear how much faster it is on paper compared to the 2 EFLOPS Aurora – let alone real-world performance. Part of the design goal of AMD's MI300A APU is to exploit additional performance efficiency gains that come from placing CPU and GPU blocks so close together, so it will be interesting to see what the software development teams programming for El Capitan can achieve, especially as they get their software further optimized.

LLNL's El Capitan is expected to cost $600 million. The system will be used nuclear weapons simulations and will be crucial for the U.S. national security. It replaces Sierra, a supercomputer based on IBM Power 9 and NVIDIA Volta accelerators, and promises to offer performance that is 16 times higher.

Gallery: El Capitan Installation Begins: World's First 2 ExaFLOPS Supercomputer

from AnandTech https://ift.tt/cWtHkGv
via IFTTT

AMD To Release Limited Run Ryzen 5 5600X3D for $230 Micro Center Exclusive

https://ift.tt/U5efp7S

With AMD now several months post release of their current AM5 platform, I figured that the sun was setting on AMD’s previous-generation AM4 platform. But, it would seem, AM4 is going to get one last hurrah, as AMD is preparing to release a new chip for the platform: a V-cache equipped, hex-core Ryzen 5 5600X. And while the chip itself is notable in a couple of ways, what’s likely going to end up better remembered is the unusual launch of the chip, with it being released as a limited volume Micro Center exclusive on July 7^th.

from AnandTech https://ift.tt/EK0y1lm
via IFTTT