CPU, APU & Chipsets - Page 148

AMD might have had a good year in 2017 with its new Zen-based line up of CPUs through Ryzen, Threadripper and EPYC, but this wasn't enough to impress OEMs from building machines round Ryzen.

Now we have Dell EMC's CTO, John Roese, saying that "Intel is the big player, AMD is the second player. There's enough diversity between them that there are use cases to have them both in our portfolio, but just the sheer breadth of the Intel processor portfolio is massive compared to even the accelerated AMD world". He's not wrong, but it's still surprising to see.

It's not all doom and gloom about AMD from Dell, as he added: "AMD is doing some interesting things, and by adding them to the portfolio we pick up a few extra areas, but let's be very clear: there is a huge, dominant player in compute semiconductors, and then there is a challenger which is doing some very good innovative work called AMD, but the gap between them is quite large in terms of market share and use-cases. So our portfolio is not going to change in any meaningful way".

AMD will be launching their refreshed Ryzen 7 2000-series later this year, with a new tease leaking out about the Ryzen 7 2700X that it will be packing the same 8C/16T as the current-gen Ryzen 7 1700X, but have turbo clocks of up to 4.2GHz.

AMD's current Ryzen 7 1700X is an 8C/16T processor with a base CPU clock of 3.4GHz and boost CPU clock of 3.8GHz, but the new Ryzen 7 2700X will have base clocks of 3.7GHz, while the boost clocks will not only hit 4GHz but fly a little higher to 4.1GHz.

There are other entries of the Ryzen 7 2700X at 4.2GHz, so we should expect a larger punch in the frequency department. My insider sources have told me there "won't be much performance" increases with the new Ryzen 7 2000-series processors, so we'll have to wait and see how big the difference is between 3.8GHz boost on the 1700X and the purported 4.2GHz on the 2700X.

Intel won't be launching its new Hades Canyon NUC for a little while yet, but that hasn't stopped someone in South Korea from getting their not-so-grubby (because they're wearing gloves) hands-on it. Why is this important? Because Intel's new NUC for the first time in history, is powered by AMD Radeon graphics.

AMD's new Radeon RX Vega M will be inside of the new Hades Canyon NUC, with 4GB of HBM2 @ 1.6Gbps, all in under 100W TDP under the NUC8i7HVK version of the PC, while the NUC8i7HNK features a 65W TDP and lowered CPU/GPU clocks.

The faster NUC is powered by a Core i7-8809G processor at 3.1GHz stock and 4.2GHz Turbo, with 8MB of cache and 100W TDP. Right next to it, is the Radeon RX Vega M GH with GPU clocks of between 1063-1190MHz, with Intel saying that this is the 'Unlocked and VR-capable' NUC. The slower NUC features a knocked down Core i7-8705G processor that is identical to the faster NUC, except its Turbo clock has been reduced by 100MHz and the GPU has been reduced even more down to 931-1011MHz as the Radeon RX Vega M GL is used.

ARM isn't sitting quietly in its fight against the major chipmakers, with the company announcing its new Project Trillium, which Engadget reports as a "combination of hardware and software ingredients designed explicitly to speed up AI-related technologies like machine learning and neural networks".

ARM ML will be much faster at machine learning tasks compared to normal CPUs and GPUs, with 2-4x the real-world throughput. The ML in ARM ML stands for machine learning, while ARM OD will handle object detection. ARM OD can handle "virtually unlimited" subjects in real-time at 1080p 60FPS, and while it is detecting people, it is good at face detection, poses, and even gestures.

On the software side of things there's ARM NN (neural network) that will act as the "go-between" for neural network frameworks like Google's TensorFlow, and the ARM-based processors.

AMD is only just getting their EPYC server processors into systems now, after the announcement and launch of EPYC over 12 months ago now, and we're already hearing news about the 7nm EPYC successor codenamed Rome.

The rumors on Rome are coming from China with WCCFTech picking it up, with the next-gen Rome CPUs packing a huge 64C/128T of performance for one SKU. AMD will reportedly have two different EPYC dies, with:

The picture above is reprotedly the Zeppelin die with 12 cores per die, and four dies totaling 48 cores. The rumor is that AMD will have a second offering with 16 cores per Zeppelin die, bringing the total CPU cores to a huge 64. If AMD does pull this off, Intel will be pushing out bricks, from you know where.

AMD had a huge year in 2017 with the successful announcements and launches of both Ryzen and Ryzen Threadripper, quickly stealing sales from Intel and causing Chipzilla to react in ways we've never seen before.

But the tide is turning and now we're seeing that as of November 2017, Intel sales are totally dominating, and that two CPUs in particular are doing major damage to AMD. Intel's Core i5-8600K and Core i7-8700K are both demolishing AMD thanks to new data from Minefactory.de.

These numbers show CPU sales from February last year, right through to January this year. You can see that for the first 9-10 months of 2017 that Intel's flagship Core i7-7700K was the CPU of choice for most enthusiasts, and that the only time that changed was in December 2017 and January 2018, when everything flicked over to the new Coffee Lake-based Core i7-8700K.

Out of nowhere, we have a new company that we'll be keeping a close eye on; Ampere. Ampere has just launched a new 64-bit ARM server processor that will compete against products released by Qualcomm, HP Enterprise, and many others.

ARM is an important cog in Ampere's new machine, as it is capable of great power efficiency and lower cost of ownership. Until now, the biggest issue with ARM-based systems is that there isn't a massive list of applications that will work with ARM processors, but that will soon change.

Ampere's first processor rocks a custom core ARM v8 64-bit at up to 3.3GHz, up to 1TB of memory and a 125W TDP. Ampere is fabless, and will have production of these new ARM processors sometime in 2H 2018.

AMD will be having a huge year this year, at least in their CPU department, with the release of their refreshed Ryzen CPUs and new 400-series chipsets in a couple of months. But now news is breaking of their upcoming Starship/Mattise CPUs that will be released late this year and early 2019.

HWINFO is now recognizing a bunch of new CPUs, AMD's upcoming 400-series chipset, and even Intel's upcoming Ice Lake-SP platform. AMD's next-gen Starship CPU is going to be a monster, offering a huge 48C/96T of CPU power rocking the refreshed Zen+ or Zen 2 architecture.

Here's what HWINFO32 and HWINFO64 v5.72 supports:

AMD is in the middle of a class action lawsuit from investors over its not-so-good comments about Meltdown and Spectre, but there was worse news last week when AMD machines were being bricked by Windows 10 security updates meant to help, not harm.

The latest updates come in the form of Windows 10 Creators Update (version 1703) and Windows 10 Anniversary Update (version 1607). These cumulative updates push the Creators Update systems to build 15063.877 (support KB4057144) and Anniversary Update PCs to build 14393.2034 (support KB4057142).

The big star here is KB4073290 that will fix the bricked AMD devices on the Windows 10 Fall Creators Update, so grab that if you've been sitting there twiddling your thumbs with your fresh new Ryzen/Threadripper-powered gaming beast.

We all know that the Meltdown and Spectre security holes have caused headaches for Intel and AMD, but now AMD has been hit with a class action lawsuit that seeks to "recover damages for Advanced Micro investors under the federal securities laws".

Rosen Law Firm is a global investor rights firm, where a class action lawsuit has been kicked off on "behalf of purchasers of the securities of Advanced Micro Devices". The lawsuit details:

"Defendants during the Class Period made materially false and/or misleading statements and/or failed to disclose that: (1) a fundamental security flaw in Advanced Micro's processor chips renders them susceptible to hacking; and (2) as a result, Advanced Micro's public statements were materially false and misleading at all relevant times. When the true details entered the market, the lawsuit claims that investors suffered damages".