Processors - Page 65

AMD will be launching the world's first processor with 3D V-Cache technology next month, with the introduction of its new Ryzen 7 5800X3D processor, but it will be joined by another 9 processors that AMD will unveil in the next month.

The news is coming from Wccftech, which reports that AMD is going to be rolling out 10 new SKUs in total: 6 from the Ryzen 5000 series and 4 x from the Ryzen 4000 series. Starting at the top, we'll see AMD introduce the new Ryzen 7 5800X3D, as well as the AMD Ryzen 5700X/5600/5500/4600G/4500/4100, and then the AMD Ryzen 5700/5100/4700.

AMD's new Ryzen 4000 series will be updated, packing the Zen 2 Renoir-X CPUs -- the same CPUs, but they'll be dropping their built-in Vega GPU and enjoying some CPU clock speed changes -- not much new there. We should see these new AMD Ryzen CPUs as the last wave of AM4-based processors, as the company will be shifting into its new AM5-based socket with DDR5 memory support later this year.

Apple just announced its new M1 Ultra SoC, which combines two of their M1 Max SoCs together for quite the mobile beast... so much so that the CPU and GPU are significantly faster than their desktop counterparts.

The new Apple M1 Ultra chip has a 20-core CPU, packing 16 high-performance cores with 4 high-efficiency cores -- joined by a 64-core GPU -- and 128GB of unified memory, a 2x more powerful media engine, and 800GB/sec of memory bandwidth.

CPU-wise, the M1 Ultra is 3.8x faster than the highest-end Intel Core i9-based Mac (discontinued), and up to 60% faster than the 28-core Mac Pro which packs an Intel Xeon W processor. On the GPU sde of things, the M1 Ultra is 4.5x faster than the 27-inch iMac, and a huge 80% faster than the highest-end Mac Pro which packs an AMD Radeon Pro W6900W graphics card.

MSI has just announced that it has recalled its new MAG CORELIQUID 240R and MAG CORELIQUID 360R AIO CPU coolers.

The company only just launched its new MAG CORELIQUID 240R and MAG CORELIQUID 360R AIO CPU coolers in the last month or so, but now it is recalling them after the company received feedback from users that were experiencing a "drop in heat dissipation efficiency" when using the AIO coolers.

MSI has discovered in their preliminary investigation that a "small portion of the liquid coolers produce sediment that can cause a blockage. This, however, will not cause any damage to your system since the processors are equipped with a protection mechanism against overheating. MSI cares deeply about our users' overall experience and we simplified our processes to provide faster product replacement services".

AMD's new Ryzen 7 5800X3D processor is out in the wild, the very first consumer-focused CPU with 3D V-Cache technology. But... it comes at a cost.

The new CPU has been torn apart and that gorgeous die shot has been shown, our first look at the Ryzen 7 5800XD chip. One of the most unfortunate things about AMD's new Ryzen 7 5800X3D processor is that it doesn't look like it has any OC (overclocking) support, whatsoever.

A post on the Chiphell forums by 'brother number one' said: "Information that can be disclosed so far: 5800X 3D does not support overclocking". But why? The 64GB of additional cache through the 3D V-Cache means that CPU clocks are lower on the 5800X3D, and the additional 3D-stacked cache gets hot underneath, and could reduce the CPU clocks too far -- or even damage the CPU.

Apple has been going from stride to stride with its in-house M1 Max SoC already offering game-changing performance for its size, and now the company is jamming two of them together by introducing the M1 Ultra SoC.

The new Apple M1 Ultra SoC literally has two M1 Max SoCs combined together with Apple's newly-announced UltraFusion -- UltraFusion is an innovative packaging architecture that interconnects the die of each M1 Max chip to create a single, monster Apple M1 Ultra SoC -- impressive to see.

Inside, the M1 Ultra packs a huge 114 billion transistors (that's 114,000,000,000) which is the most ever in a personal chip. Apple has up to 128GB of high-bandwidth, low-latency unified memory that sits next to a 20-core CPU, 64-core GPU, and 32-core Neural Engine.

AMD's new Ryzen 7 5800X3D processor will be launching on April 20, with a retail price of $449 -- at this price, it replaces the original price of the Ryzen 7 5800X -- and it's a helluva lot faster.

The news is coming from Disclosuzen, who is reporting the April 20 release + $449 price. As for the CPU itself, it is the first consumer CPU that rocks 3D V-Cache technology. This sees the Ryzen 7 5800X3D rocking a huge 100MB of cache, a huge increase from the Ryzen 7 5700X that sits below it in the 2022 updated CPU family from AMD, as that chip has just 36MB of cache.

3D V-Cache is a huge L3 cache that sits on top of the compute chiplet, where AMD is expecting around 15% more performance thanks to the stacked memory, which means the new Ryzen 7 5800X3D processor is actually faster than AMD's higher-end Ryzen 9 5900X in gaming. Not only that, but the new 5800X3D beats the Core i9-12900K in gaming which is what really matters.

AMD is about to unleash its beefy new Ryzen Threadripper PRO 5000 series CPUs which will be led by the flagship Ryzen Threadripper PRO 5995WX processor... and man, it's going to be a beast.

In some new leaks, we have five different Ryzen Threadripper PRO chips which will see AMD having the new flagship Ryzen Threadripper PRO 5995WX processor will feature 64 cores and 128 threads, at up to 4.5GHz. The new Ryzen Threadripper PRO 5000 series CPUs will be based on the Zen 3 architecture, just like the desktop Ryzen 5000 series CPUs.

AMD will reportedly have the Ryzen Threadripper PRO 5995WX, Ryzen Threadripper PRO 5975WX, Ryzen Threadripper PRO 5965WX, Ryzen Threadripper PRO 5955WX, and the Ryzen Threadripper PRO 5945WX. All of these new Zen 3-powered CPUs will have a default 280W default TDP.

AMD will have its exciting new Ryzen 7 5800X3D processor in the hands of enthusiasts, benchmarkers, and gamers by the end of the month as the first CPUs have rolled out of the factory.

In a new tweet, industry insider "Greymon55" tweeted that AMD's new Ryzen 7 5800X3D processor "has been shipped and will probably be on the market by the end of this month". As for the Ryzen 7 5800X3D processor and what makes it so exciting is its new 3D V-Cache technology.

3D V-Cache is a huge L3 cache that sits on top of the compute chiplet, where AMD is expecting around 15% more performance thanks to the stacked memory, which means the new Ryzen 7 5800X3D processor is actually faster than AMD's higher-end Ryzen 9 5900X in gaming. Not only that, but the new 5800X3D beats the Core i9-12900K in gaming which is what really matters.

Intel's upcoming mobile mega-thread CPU has been teased, with partners of Intel already receiving a new platform codenamed "Alder Lake-HX".

The new Alder Lake-HX test sample CPUs are labeled as "8+8" which would be an indicator that the new Alder Lake-HX chip that's being tested has 8 Performance, and 8 Efficient cores (thus, 16 cores in total). Right now, Alder Lake-P and Alder Lake-U processors are limited to 14 cores, but the new Alder Lake-HX chips look to be competing with the Zen 3-based Ryzen 6000 HX series CPUs from AMD.

Intel's new 16-core mobile Alder Lake-HX processor will be aiming for the "Muscle" segment of laptop users, with between 45-55W TDP. The codename "Alder Lake-HX" processor that we have here is just a single SKU, and should end up being the Core i9-12980HX... but there could be other SKUs that we'll see in the coming months.

Intel's current flagship 12th Gen Core "Alder Lake" CPU is the Core i9-12900K with 16 cores and 24 threads, but the company has just teased its new flagship 13th Gen Core "Raptor Lake" CPU with 24 cores and 32 threads that drops later this year.

The company showed off an engineering sample of its next-gen Raptor Lake CPU, packed with 24C/32T of processing power with Intel's hybrid architecture and blend of Performance and Efficient cores. Intel said that its next-gen consumer CPU will have 8 Performance cores and 16 Efficient cores, meaning there's twice as many "small" cores than current Alder Lake CPUs.

Intel's new Raptor Lake CPUs will be made on its in-house Intel 7 process technology, the same Intel 7 process that is used for the current-gen Alder Lake CPUs. The new Intel 13th Gen Core "Raptor Lake" CPUs will be socket compatible with LGA1700, something Intel also confirmed -- but it didn't add if you'll need a BIOS update -- something I'm sure we'll hear more about in the coming weeks.

Intel has confirmed its next-gen 2024 processor series, with the tease of its 15th Gen Core "Arrow Lake" CPUs at its recent Investors Meeting 2022.

Intel CEO Pat Gelsinger confirmed that Intel's new 15th Gen Core "Arrow Lake" CPUs will succeed the next-gen 14th Gen Core "Meteor Lake" CPUs that are launching in 2023. Before that, we have the 13th Gen Core "Raptor Lake" CPUs launching later this year, which succeed the 12th Gen Core "Alder Lake" CPUs that Intel just launched.

The new Intel 15th Gen Core "Arrow Lake" CPUs will be launching on not one but three different process nodes: Intel 4, Intel 20A, and External N3 technology... but spoiler alert: it'll be TSMC. After Arrow Lake, we have the 16th Gen Core "Lunar Lake" which Tom @ Moore's Law is Dead last year said would KILL the next-next-gen Zen 5 architecture from AMD. The next few years of CPU technology are going to be pretty radical, while Intel simultaneously launches a new GPU to compete with AMD and NVIDIA. Can't wait.

Intel is preparing another flagship Alder Lake CPU with the introduction of the upcoming Core i9-12900KS, which is a heavily-binned 12900K with higher max Turbo CPU clocks.

Intel's new Core i9-12900KS still packs 8 Golden Cove CPU cores and 8 Gracemont CPU cores, which means there's a total of 16 cores (8+8) and 24 threads (16+8) in total. The hybrid architecture sees the P-cores operating at a maximum boost clock of 5.5GHz on 1-2 active cores, and 5.2GHz on all cores. The Gracemont-based E-cores clocks are 3.9GHz with 1-4 cores, and up to 3.7GHz for all cores.

We have a higher base 150W TDP for the new Core i9-12900KS, which is a 25W increase over the 125W TDP on the normal Core i9-12900K. The normal Core i9-12900K has a maximum turbo power rating of 241W, so we can expect an increase on that for the new binned Core i9-12900KS variant.

AMD has been ripping up the HEDT processor market with its Ryzen and Ryzen Threadripper CPUs, refreshed with the newer Ryzen Threadripper PRO processors... but where is Intel? Well, they're about to burst back onto the HEDT processor scene in a big way.

Intel's next-gen Sapphire Rapids-AP Xeon CPUs will rock the company's new Golden Cove CPU cores, and could deploy as an MCM-based design. The current-gen Sapphire Rapids-SP processors use an MCM-based design, with 4 tiles and up to 15 cores (14 cores enabled) each.

The new Intel Sapphire Rapids-AP chips will offer up to 56 Golden Cove CPU cores, with as low as 12 of those CPU cores able to pump above 4GHz. There will be on-chip accelerators that are already on Sapphire Rapids HEDT processors, but we don't know if they will be working or not when the chips are deployed.

Intel's next-gen flagship 13th Gen Core "Raptor Lake" CPU has been benchmarked in Ashes of the Singularity.

The upcoming Intel Core i9-13900K will succeed the 12th Gen Core "Alder Lake" flagship Core i9-12900K, with the Raptor Lake-powered CPU now in testing and is as fast as the 12900K during its ES (engineering sample) phase.

We're looking at a larger 24 cores and 32 threads for the Core i9-13900K, up from the 16 cores and 24 threads on the Core i9-12900K. Ashes of the Singularity isn't up for the task of differentiating between the physical and logical cores (thanks to Alder Lake, and Raptor Lake having a hybrid architecture with P-cores and E-cores).

Imagine a future where you've purchased a brand new Intel Core i9 processor, let's say the Intel Core i9-14900K in a couple of years' time... and then you go to play a new movie in 8K 60FPS (we're in the future, guys).

Blurt. System error. New feature required. Please pay Intel $149.99 to unlock the video codec required to playback your content. This is the future Intel is headed towards, Intel (payable, probably with some shitty Intel NFT or something) Inside.

Well, the future is now old man -- and with Linux 5.18 dropping in Spring 2022 we're seeing Intel rolling out its new Software Defined Silicon (SDSi) in the open-source Linux OS. This allows Intel Xeon CPUs to receive features and updates installed directly to the CPU, after you've purchased it. Intel's upcoming Xeon "Sapphire Rapids" CPUs will be the first processors with this new ability from the company. But man, people aren't happy... and for good reason.

AMD's new EPYC 7773X "Milan-X" processors with 3D V-Cache technology are now out on the market, with some "overclocking" performed on the dual-CPU powerhouse 128-core, 256-thread monster.

A dual-socket SuperMicro motherboard was used, with dual AMD EPYC 7773X "Milan-X" processors with 64 cores and 128 threads each, so with a dual-CPU configuration, we're looking at 128 cores and 256 threads -- as well as AMD's new 3D V-Cache technology -- and now some overclocking.

One of the AMD EPYC 7773X processors was detected as the "100-000000504-04 CPU" which is an OPN core for the AMD engineering sample. This processor in particular has a 2.2GHz base CPU clock and up to 3.5GHz boost CPU clock -- but this engineering sample had its CPU clocks at 100MHz lower -- 2.1GHz and 3.4GHz, respectively. But there's also the special 3D V-Cache technology that dollops a special layer of L3 cache right on top of the compute dies.

Samsung has unleashed its next-gen Exynos 2200 processor with a custom GPU based on the AMD RDNA 2 architecture, and now according to some new benchmarks, we know it beats Qualcomm's new Snapdragon 8 Gen1 chip.

The custom GPU inside of the Exynos 2200 is called the Xclipse GPU, which has been discovered inside of a Samsung model number "SM-S908B" running Android 12 and rocking 8GB of RAM in an OpenCL benchmark. The Xclipse GPU is up against the Snapdragon 8 Gen1-powered Motorola XT2201-2, and beats it even in early form.

The biggest part here is the new custom Xclipse GPU only has a GPU clock of 555MHz -- yeah, just 555MHz and not the previously rumored 1.3GHz+ that was reported -- Samsung's new Xclipse GPU packs 4GB of memory but we don't know if it is dedicated memory, as in 4GB of VRAM, or if its 4GB of shared DRAM.

Samsung has finally, through the flames of drama, officially announced its next-gen Exynos 2200 processor. The company is saying that this new "game-changing" SoC is powered by an AMD RDNA 2-based "Xclipse" GPU.

Samsung's new Exynos 22 is a "freshly designed mobile processor" with an AMD RDNA 2-based Samsung Xclipse GPU, with the "most cutting-edge ARM-based CPU cores available in the market today and an upgraded neural processing unit (NPU), the new Exynos 2200 will enable the ultimate mobile phone gaming experience, as well as enhancing the overall experience in social media apps and photography".

Thanks to the RDNA 2-powered Samsung Xclipse GPU, we have hardware-accelerated ray tracing (RT), as well as variable-rate shading (VRS) -- these technologies have been reserved for PC gamers, and only just recently with the next-gen consoles -- and now Samsung has it in your hands.

NVIDIA is expanding its CPU business with the announcement of a new CPU group in Israel, which will help NVIDIA continue to dominate the HPC market.

It was only in April 2021 that NVIDIA CEO Jensen Huang announced their next-gen ARM-based "Grace" CPU, which will be released in early 2023. The new CPU group will be working with other teams in Israel that are already working on high-speed networking, HPC technologies, and NVIDIA's own Data Processing Units.

NVIDIA CTO Michael Kagan said: "Israel, with its unique wealth of talent, is a key player in the global tech ecosystem, and we are excited to be creating a new CPU group here. We look forward to further growing our local R&D activities both in this area and in our extensive work supporting the local ecosystem through unique programs for startups and developers".

Intel CEO Pat Gelsinger posted a video he made for the end of 2021, providing some of his thoughts on many topics -- but the biggest one was about Intel's new Alder Lake CPUs -- and how AMD is now finished.

Well, kinda -- Gelsinger joined Intel as its freshly-minted CEO in January 2021 in a very tumulous time for the company. In his video, Gelsinger said: "Alder Lake. All of a sudden...Boom! We are back in the game. AMD in the rearview mirror in clients [consumer market], and never again will they be in the windshield; we are just leading the market".

AMD was fighting the good fight with its new Zen 3-powered Ryzen 5000 series desktop CPUs, but Intel really did deliver in strides with its new Alder Lake-powered 12th Gen Core CPUs. Not only are the CPUs the very best you can buy right now, but they also have both DDR5 memory and PCIe 5.0 support. AMD is in the rearview mirror with performance, DDR4 only, and PCIe 4.0 only -- that is, until Zen 4 is here with the Ryzen 7000 series desktop CPUs later this year.