ARM

[Sinac]

Za one koji još nisu naišli na vijest:
https://asia.nikkei.com/Business/Tec...uction-for-Mac

Ako će ARM ići u ovom smjeru - poboljšanje performansi uz umjerenu potrošnju el. energije - bit će to veliki udarac i Intelu i AMD-u. Pri tome ne mislim samo na consumer segment računala već i na serverske platforme. Svakome će biti interesantno da dobije 20% niži račun za struju uz iste performanse.

I baš me zanima što će biti sad nakon prodaje ARM d.o.o. () Nvidiji koja je bila "na čekanju". Nekako mi se čini da bismo mogli uskoro imati desktop kartice s normalnijom potrošnjom el. energije i grafičke čipove za mobilne platforme boljih performansi.

[The Exiled]

Još ih Britanci (malo) koče u totalnom preuzimanju, taman toliko da se i oni namire, ali naposljetku će nVidia vrlo vjerojatno asimilirati ARM u potpunosti, prema planu i programu.

Kaj se serverskog dijela priče tiče, ARM se već danas sasvim solidno nosi s modernim Intel i AMD rješenjima, tj. obje firme već sad imaju razlog za određenu dozu zabrinutosti.

Intel više, nego ostali, jer su ionako izgubljeni u prostoru i vremenu sa standardnim stvarima za koje nVidia i AMD konkuriraju, dok ARM dolazi kao još jedan u nizu.

[The Exiled]

Ampere roadmap update: Switching to in-house CPU designs, 128+ 5nm cores in 2022

For people not familiar with Ampere, the company was founded back in 2017 by former Intel president Renée James, notably built upon a group of former Intel engineers who had left along with her to the new adventure. Ampere’s next generation "Mystique" Altra Max is the next product on the roadmap, and is targeted to be sampling in the next few months and released later this year. The design relies on the same first generation Arm Neoverse N1 cores, at the same maximum 250W TDP as a drop-in replacement on the same platform, however with an optimised implementation that now allows for up to 128 CPU cores – 60% more cores than the first iteration of Altra we have today, and double the amount of cores of competitor systems from AMD or Amazon’s Graviton2. Today’s big reveal comes in regard to the microarchitecture choices that Ampere is going to be using starting in their next generation 2022 “Siryn” design, successor to the Altra Max, and relates to the CPU IP being used.

Starting with Siryn, Ampere will be switching over from Arm’s Neoverse cores to their new in-house full custom CPU microarchitecture. In fact, Ampere explains that what the move towards a full custom microarchitecture core design was actually always the plan for the company since its inception, and their custom CPU design had been in the works for the past 3+ years. The company wouldn’t comment on the exact status of the Siryn design right now – on whether it’s been taped in or taped out yet, but they do retierate that they’re planning customer sampling in early 2022 in accordance to prior roadmap disclosures.

The “Mystique” code-named Altra Max design will be characterised in that it’s able to increase the core-count by 60% versus the current generation Altra design, all while remaining at and below the same 250W TDP. the relative positioning against the Altra Q80-30 and the EPYC 7742 would indicate that the new chip would easily surpass the performance of even AMD’s latest EPYC 7763. Ampere actually discloses the SKU model name being used for the comparison, which is the "Altra Max M128-30" – meaning for the first time we have confirmation that all 128 cores are running at up to 3GHz clock speed, which is impressive given that we’re supposed to be seeing the same TDP and power characteristics between it and the Q80-33.

Izvor: AnandTech i ServeTheHome

[Sinac]

Malo nagađanja oko Apple M2 chipa:

https://www.tomsguide.com/news/apple-m2-chip

[The Exiled]

Rumours of an ARM IPO emerge as nVidia's acquisition hits regulatory hurdles

nVidia's ARM acquisition may face delays, as the GPU giant's facing more regulatory scrutiny than expected. nVidia expected their ARM acquisition to be closed by March 2022, though that end date now seems incredibly optimistic given reports of nVidia's delayed paperwork. The Telegraph has reported that ARM is considering an IPO (Initial Public Offering) as a backup plan should their deal with nVidia falls through. ARM has denied the report, as Simon Segars, Arm’s chief executive, stating that “There is no IPO planning going on at Arm, we are 100pc focused on closing this transaction.” Should nVidia fail to get the approval of all regulators, their ARM acquisition deal will fall through. While nVidia plans to acquire ARM before March 2022, the company has the option to extend its deal's deadline until September 2022. From there, nVidia can either renegotiate their deal and try again or put an end to their acquisition plans.

Should the European Commission receive nVidia's paperwork in September, regulators could still take six months or more to decide whether or not to approve the deal. This makes nVidia's original March 2022 deadline seen incredibly optimistic. nVidia does not have the power to rush the EU's regulators. As Jonathan Branton, head of competition at law firm DWF, puts it: “No regulator will allow itself to be pushed into a corner and having to make a decision in a timeframe it’s not happy with.” While nVidia's approval from the EU might delay proceedings, China may prove to be a larger hurdle to overcome. China is infamous for allowing approvals for large mergers to take from one year to 18 months, a factor that could push nVidia's ARM acquisition past their extended September 2022 deadline. Should ARM be acquired by nVidia, the company will fall under US ownership, a prospect that China is wary of. Beyond that, there are ongoing problems with ARM China, ARM's Chinese joint venture, which has seen the joint venture go rogue as ARM failed to oust the division's Chief Executive, Allen Wu.

Should nVidia acquire ARM, a US-based company, China will face the prospect of further sanctions from the US to limit its use of ARM technology. At this point, China remains a wild card, one that could make or break nVidia's ARM ambitions. As it stands, a lot of people are against nVidia's plan to acquire ARM. For starters, nVidia is an ARM licensee, a factor that would make nVidia a user of ARM's IP and the controller of ARM's direction as a company. Currently, ARM is the Switzerland of the technology industry, an independent entity that licenses its IP to a broad range of customers, including AMD, Apple, Broadcom, Qualcomm, Samsung, and many more. nVidia ownership of ARM will allow nVidia to push ARM's development plans into areas that primarily benefit nVidia's hardware, potentially allow nVidia to withhold IP from licensees until after nVidia implements it (giving nVidia an unfair advantage), or increase their license costs to squeeze the margins of its competitors. While nVidia has the money and expertise to push ARM forward, nVidia's ownership of ARM remains a contentious issue. Beyond the issues raised in this article is the UK's loss of ARM to US ownership and the potential mass exodus of ARM licensees as nVidia's competitors seek to create a non-nVidia alternative to ARM's architecture.

Izvor: OC3D

[Sinac]

Nvidia se možda neće dokopati ARM Ltd.
Ne znam o kakvoj se nacionalnoj sigurnosti tu radi, više mi sliči na to da se netko u UK sjetio da ARM ima veliki potencijal.

https://www.reuters.com/technology/u...ws-2021-08-03/

[The Exiled]

Ma da, ono kaj smo pričali, još da se i Britanci namire u toj nVidia + ARM priči i to je to.

[The Exiled]

Will AMD get back into ARM server chips?

There was a bit of a kerfuffle this week when it looked like AMD was changing its position a little bit on whether or not it would get back into designing and selling server chips based on the Arm architecture. The funny thing is that the world has changed around AMD, and a lot of it for the good, as its EPYC X86 server processors have attained around ten percent market share (by shipments), but its position on Arm servers has not really changed. AMD’s top brass has said all along that if customers wanted Arm chips, it would make them. Well, maybe AMD’s position, internally, has changed just a little now that the server market has changed so much since the summer of 2015 when AMD stopped its “SkyBridge” X86-Arm shared socket effort, launched in 2014 with much fanfare. Prior to that announcement, AMD’s Opteron Arm server chips based on its homegrown “K12” core were expected in 2016 and were pushed out to 2017, and they were SoCs that combined Arm cores and GPUs, much as the “Seattle” Opteron A110 was. Seattle fizzled because it was too underpowered to do anything useful in the datacenter. So AMD started shutting things down, killing off SeaMicro microservers and its interconnect, killing off Arm server chips, so it could just hunker down on the EPYC server CPU. Because, frankly, the Arm server chip demand never materialized.

AMD has plenty of interesting options when it comes to Arm server chips. First of all, as Cavium, which is now part of Marvell, proved so well and as AMD proved with its Seattle Arm chips, it is not all that hard to do a global replace and put Arm cores in the same place where an Octeon NPU core or an X86 CPU core is in a processor design. The uncore stuff can be recycled, and if you are an Arm licensee as AMD is, you can take some of the stuff in the guts of the core — branch predictors, caches and cache hierarchies, vector units, other accelerators, memory controllers, and peripheral controllers and reuse these elements in an Arm server chip. So if a hyperscaler (not likely a public cloud since no one has a lot of Arm server workloads outside of the hyperscalers) doesn’t want to do an Arm server chip design all by its lonesome, they could hire AMD to do it.

While that would be interesting, what we think would actually be far more interesting is for AMD to pull a sequel to the SkyBridge effort and create a common socket for X86 and Arm server chips that is completely neutral about what kind of chip is plugged in. This is a real value for those who want flexibility, and if it really wanted to hedge its bets, it could do server CPU designs that had a lot of the components between X86 and Arm — and maybe someday RISC-V — be as common as possible.

Izvor: The Next Platform

[Dr. Strange]

Autor Sinac

Nvidia se možda neće dokopati ARM Ltd.
Ne znam o kakvoj se nacionalnoj sigurnosti tu radi, više mi sliči na to da se netko u UK sjetio da ARM ima veliki potencijal.

https://www.reuters.com/technology/u...ws-2021-08-03/

I bolje, ARM bi trebao biti neovisan, pružati licencu svima po jednakim uvjetima, a znamo kakva je Nvidia i koliko prljavo znaju igrati, nisu se bezveze svi živi pobunili i 99.9% da to neće uspjeti.

[Sinac]

Autor Dr. Strange

I bolje, ARM bi trebao biti neovisan, pružati licencu svima po jednakim uvjetima, a znamo kakva je Nvidia i koliko prljavo znaju igrati, nisu se bezveze svi živi pobunili i 99.9% da to neće uspjeti.

Da, stvari se i dodatno kompliciraju za Nvidiu.
https://www.bloomberg.com/news/artic...tough-scrutiny

Jedino mi nije jasno kako to da se nitko nije sjetio koliko ARM zapravo vrijedi i koju sve pozadinu ima cijela priča oko ARM-a dok Nvidia nije uletila s planom za kupnju.

[The Exiled]

Ampere Altra Max M128-30

The past month we have started our testing of Ampere's Altra Max M128-30, the company's new 128 core server processor, and in this article today are our initial benchmarks of this promising chip for high core count servers in both 1P and 2P configurations tested. With Ampere Altra Max the flagship processor is currently the M128-30 that has a 3.0GHz maximum clock speed for its 128 cores. Ampere Altra Max processors continue to use the Neoverse N1 cores, support eight channels of DDR4-3200 ECC memory, 128 lanes of PCIe Gen4, and other features in common with Altra. Like the Q80-33, the M128-30 does maintain the same 250 watt TDP rating. Ampere Altra Max processors are a drop-in upgrade to Ampere Altra with an updated BMC and system firmware.

Across a wide range of tests, the Ampere Altra Max M128-30 2P is proving to be very competitive with the AMD EPYC 7763 2P in workloads that scale well and make full use of ARMv8 capabilities. Given the Ampere Altra Q80-33 2P was already outperforming the Xeon Platinum 8380 2P Ice Lake server, it's to no surprise Ampere's lead there has widened even more substantially with now being able to offer 128 cores per socket. It's incredible to see the competitiveness of Ampere Altra Max in testing ARM Linux servers since the days of Calxeda and how far not only the hardware but also the ARM software ecosystem as a whole has come over the past 10~15 years in being able to truly take on x86_64 in the server space.

Izvor: Phoronix

[Starsylph]

Čovječe, 128 jezgri, kada će to postati mainstream.

[The Exiled]

Apple's M1 Pro, M1 Max SoCs investigated: New performance and efficiency heights

Apple didn’t talk much about core performance of the new M1 Pro and Max, and this is likely because it hasn’t really changed all that much compared to the M1. We’re still seeing the same Firestrom performance cores, and they’re still clocked at 3.23GHz. The new chip has more caches, and more DRAM bandwidth, but under ST scenarios we’re not expecting large differences. Against the competition, the M1 Max either has a significant performance lead, or is able to at least reach parity with the best AMD and Intel have to offer. The chip however doesn’t change the landscape all too much.

What’s more interesting than ST performance, is MT performance. With 8 performance cores and 2 efficiency cores, this is now the largest iteration of Apple Silicon we’ve seen. Apple does have 2 additional cores versus the 8-core 11980HK or the 5980HS, the performance advantages of Apple’s silicon is far ahead of either competitor in most workloads. Again, to reiterate, we’re comparing the M1 Max against Intel’s best of the best, and also nearly AMD’s best (The 5980HX has a 45W TDP). The performance differences here are just insane, and really showcase just how far ahead Apple’s memory subsystem is in its ability to allow the CPUs to scale to such degree in memory-bound workloads.

Izvor: AnandTech

[Sinac]

Ne znam koliko vas je naletjelo na ovu vijest, s obzirom na potrošnju el. energije ne bi me čudilo da od toga nešto i bude:

Microsoft and AMD are reportedly developing an Arm processor for laptops

https://www.notebookcheck.net/Micros....568434.0.html

[The Exiled]

Apple announces M1 Ultra: Combining two M1 Maxes for Workstation Performance

Apple unveiled the fourth and final member of the M1 family of Apple Silicon SoCs, the M1 Ultra. Aimed squarely at desktops – specifically, Apple’s new Mac Studio – the M1 Ultra finds Apple once again upping the ante in terms of SoC performance for both CPU and GPU workloads. And in the process, Apple has thrown the industry a fresh curveball by not just combining two M1 Max dies into a single chip package, but by making the two dies present themselves as a single, monolithic GPU, marking yet another first for the chipmaking industry. Apple has bonded two M1 Max dies together on to a single chip, with all of the performance benefits doubling their hardware would entail.

At the heart of the new M1 Ultra is something a bit older: the M1 Max. Specifically, Apple is using two M1 Max dies here, and then bonding them together to form a massive amalgamation of 114B transistors. On the CPU front, this means Apple now offers a total of 20 CPU cores. This is comprised of 16 of their performance-focused Firestorm cores, and 4 of their efficiency-focused Icestorm cores. With 32 GPU cores, M1 Max was already setting records for a monolithic, integrated GPU. And now Apple has doubled things to 64 GPU cores on a single chip.

The secret ingredient that makes this all possible – and which Apple has been keeping under wraps until today – is that M1 Max has a very high speed interface along one of its edges. An interface that, with the help of a silicon interposer, allows two M1 Max dies to be linked up. Apple calls this packaging architecture UltraFusion, and it’s the latest example in the industry of 2.5D chip packaging. Thanks to UltraFusion, Apple has become the first vendor to ship a chip that transparently combines two otherwise separate GPUs. In particular, the company is touting that the M1 Ultra’s GPU performance exceeds that of NVIDIA’s GeForce RTX 3090, which at the moment is the single fastest video card on the market. And furthermore, that they’re able to do so while consuming a bit over 100 Watts, or 200 Watts less than the RTX 3090.

Izvor: AnandTech

[Sinac]

Prvi benchmark pokazuje snagu M1 Ultra procesora.

https://www.macrumors.com/2022/03/08...ked-benchmark/

Ovo uzmite s rezervom dok ne potvrdi barem još jedan sajt. U svakom slučaju, impresivne brojke.

[stitrek]

Kupnja ARM-a sa strane Nvidije ipak nije prosla te je obustavljena, jedna od posljedica su veliki broj otkaza koji slijede u ARM-u.

[Sinac]

Autor stitrek

Kupnja ARM-a sa strane Nvidije ipak nije prosla te je obustavljena, jedna od posljedica su veliki broj otkaza koji slijede u ARM-u.

Ne znam odakle si iskopao tu informaciju, ARM se nije ništa unazadio zbog toga. ARM i dalje živi od proizvodnje procesora (GPU, CPU, network...) i licenci koje naplaćuje. I nadalje se svi veliki igrači motaju oko ARM-a, Apple, Nvidia, Microsoft...

https://www.nextplatform.com/2022/02...m-partnership/

[udarnik60]

Pa nije teško provjeriti tu informaciju.

https://www.theguardian.com/business...e-uk-us-nvidia

Samo kao da lika to zanima. U jednom danu je pokušao nabiti postove za tržnicu pa je imao nešto za reći u svakoj temi....


Sent from my SM-A715F using Tapatalk

[Sinac]

Autor udarnik60

Pa nije teško provjeriti tu informaciju.

https://www.theguardian.com/business...e-uk-us-nvidia

Samo kao da lika to zanima. U jednom danu je pokušao nabiti postove za tržnicu pa je imao nešto za reći u svakoj temi....


Sent from my SM-A715F using Tapatalk

OK, iako ja mislim da će uskoro objaviti da ipak neće otpuštati radnike, pogotovo što će i UK i EU i Intel pogurati proizvodnju čipova na europskom području. Ako se ne varam, ARM je trenutno jedini dizajner čipova na europskom području koji ima značajniju ulogu. S obzirom da će transport s azijskog područja i nadalje biti s velikim problemima, kao i sama proizvodnja koja će uvelike ovisiti o tome kako će se Kina politički ponašati u odnosu na Rusiju, bit će još velikih previranja po tržištu.

[The Exiled]

ARM Announces the Cortex-X3, Cortex-A715 CPU Cores and Immortalis-G715 GPU

Arm is announcing its new flagship Immortalis GPU today, its first to include hardware-based ray tracing on mobile. As PCs and the latest Xbox Series X and PS5 consoles are all gradually moving toward impressive ray-traced visuals, Immortalis-G715 is designed to be the Arm’s first GPU to deliver the same on Android phones and tablets. Built on top of Mali, a GPU that’s used by the likes of MediaTek and Samsung, Immortalis is designed with 10–16 cores in mind and promises a boost of 15 percent over the previous generation premium Mali GPUs. Arm sees Immortalis as the start of a transition to ray tracing on mobile following its success with the 8 billion Mali GPUs that have shipped to date.

Arm also has an update to its main Mali line with the Mali-G715. This GPU includes variable rate shading (VRS) to boost gaming performance and energy savings on mobile. VRS essentially renders the parts of a scene in a game that require more detail, so details in the background don’t need as much rendering power. Arm’s move to support hardware-based ray tracing on its GPUs is a significant step for mobile Android gaming. Ray tracing is currently limited to powerful GPUs that are typically found in gaming PCs or the latest Xbox Series X and PS5 consoles.

Izvor: TechPowerUp i The Verge

[The Exiled]

nVidia details Grace Hopper CPU Superchip design: 144 cores on 4N TSMC process

nVidia's Grace CPU is the company's first CPU-only Arm chip designed for the data center and comes as two chips on one motherboard, totaling 144 cores, while the Grace Hopper Superchip combines a Hopper GPU and the Grace CPU on the same board. Among the most important disclosures, nVdiia finally officially confirmed that the Grace CPUs use the TSMC 4N process. TSMC lists the "N4" 4nm process under its 5nm node family, describing it as an enhanced version of the 5nm node. nVidia uses a specialized variant of this node, dubbed '4N,' that is optimized specifically for its GPUs and CPUs. The Grace CPU has 16 dual-channel LPDDR5X memory controllers, working out to 32 channels that support up to 512 GB of memory and up to 546 GB/s of throughput. nVidia says it selected LPDDR5X over HBM2e due to multiple factors, like capacity and cost. Meanwhile, LPDDR5X provides 53% more bandwidth and 1/8th the power-per-GB compared to standard DDR5 memory, making it the better overall choice.

There are really two reasons nVidia is building its own CPU. First, nVidia wants something that is a more efficient co-processor to the company’s GPUs in large systems. nVidia owns the hardware/ software stack much like Apple, so this is an attempt to become a truly full-stack provider. nVidia does not need Intel/ AMD’s new AI accelerators, it has its own. Likewise, it can build a chip with the right performance attributes to augment a GPU, rather than duplicating other parts. The second reason is more obvious now. nVidia needs to de-couple itself from Intel and AMD. Both are competing with nVidia in HPC and AI. At the same time, the Hopper H100 is waiting for Sapphire Rapids to launch before the DGX H100 can launch. As a result, we are expecting a mid-to-late Q1 2023 debut for Hopper. Given that the chips exist, (see below) nVidia needs a CPU solution so it can sell its GPUs.

Izvor: Tom's Hardware i ServeTheHome

[The Exiled]

ARM is the new RISC/Unix, RISC-V is the new ARM

When computer architectures change in the datacenter, the attack always comes from the bottom. And after more than a decade of sustained struggle, Arm Ltd and its platoons of licensees have finally stormed the glass house – well, more of a data warehouse (literally) than a cathedral with windows to show off technological prowess as early mainframe datacenters were – and are firmly encamped on the no longer tiled, but concrete, floors. The success of the proprietary minicomputers spawned a new breed of server makers, using computationally efficient RISC processors and operating systems that hewed to the Unix standard to various degrees that started off in scientific workstations in the mid-1980s. Intel expanded and hardened the X86 architecture, which was not initially as capacious or reliable as the RISC chips made by Sun, HP, Data General, ARM, and others. It didn’t need to be as computing on the Web was more distributed. What the Web-scale world needed was for datacenter compute to be was cheaper, and the X86 architecture was certainly cheaper than the RISC/Unix machines of the time. The relative affordability and absolute compatibility of the X86 architecture literally transformed the world.

And now the Arm CPU architecture that is in our tablets and smartphones and that took over as the embedded controller of choice from PowerPC, which displaced the Motorola 68K – and all four CPU architectures famously used in Apple machines – is pulling the X86 off that peak even as server volumes and revenues keep growing. If you ever wonder why Intel chief executive officer Pat Gelsinger is so hot to trot with opening up the Intel foundry business, this is why. Because, as we have been saying for more than a decade now, when Arm comes into the datacenter, another profit pool is going to be eliminated and the money is just going to end up in someone else’s pocket. Arm will very quickly become the CPU of choice on the clouds and among the hyperscalers unless Intel radically cuts prices on Xeon SP CPUs. With a 30 percent to 40 percent price/performance advantage for equivalent performance for a cloud instance, why would you deploy on an X86 instance instead of an Arm instance? Perhaps you buy a Xeon SP – and a back generation one at that – because AMD is sold out of Epyc CPUs and that is all you can get? Our point is that this is not just about mobile affecting servers.

John Cocke created a Reduced Instruction Set Computing architecture to try to make CPUs more efficient in the experimental 801 system in the 1970s, which ended up being used in IBM mainframe controllers and not much else. And along with the help of luminaries like David Patterson, Complex Instruction Set Computing (it wasn’t called until RISC made a distinction necessary) that prevailed in all CPU designs was replaced by the RISC method. Even the venerable X86 is a RISC computer at heart that is pretending to be a CISC machine, and has been for decades. We would not be surprised if IBM is, under the covers, doing the same thing with its System z mainframe motors.

The scale of Arm server deployments is on the order of hundreds of thousands of units against a few million units per quarter, but the Arm slice is growing fast. And the success of Arm is not being driven by performance per watt in the strictest sense, as was the drive early on with Arm server designs from a decade ago. This time, it is about an Arm CPU having equivalent or better performance than an X86 CPU, and much lower cost – and about the hyperscalers, cloud builders, and telcos having the option of customizing an Arm CPU design and having their own relationships with the few remaining advanced process chip foundries. This is all about IT organizations controlling their own fate.

Izvor: The Next Platform i Phoronix

[Sinac]

Sad, ovoj situaciji kada se mijenja cijena el. energije, ispast će da će korisnici itekako početi voditi računa o potrošnji uređaja. Šteta što je Apple s M1 i M2 procesorima jedini koji je nešto značajnije napravio na desktop segmentu.

[The Exiled]

AMD-ov Zen 3+ Rembrandt se dosta dobro nosi s Apple M2 čudima tehnike, tak da, ako AMD ikada odluči taj SoC zapakirati u desktop APU, bog-bogova.

Apple M2 vs. AMD Ryzen 7 Pro 6850U performance in nearly 200 benchmarks

Izvor: Phoronix

It was a nice improvement in performance-per-Watt with Rembrandt and further extends the AMD power efficiency wins over Intel Alder Lake P.

In addition to the power efficiency, most compelling with the Ryzen 7 PRO 6850U on Linux were the RDNA2-based Radeon 680M graphics as a big upgrade over Vega graphics.

[Sinac]

Nadam se da će se AMD upustiti u daljnji razvoj Rembrandt tehnologije. Nije ni čudo da se Nvidia htjela dokopati ARM-a, gledam novu 4xxx seriju grafičkih kartica, je da su brze i moćne, ali to pojede struje da ti odmah mrak padne na oči, i bez redukcije struje.

[The Exiled]

SiPearl works with AMD on GPU support for ARM HPC chip

Chip designer SiPearl is working with AMD on software support to enable supercomputing systems that pair SiPearl's high-performance Rhea processor with AMD's Instinct GPU accelerators. SiPearl is the company involved with the European Processor Initiative (EPI) to develop an Arm-based processor for an exascale supercomputer designed and built in Europe, slated for 2023. This latest agreement sees the company enter a business collaboration with AMD towards a joint offering for exascale supercomputing systems using their respective technologies. AMD and SiPearl said they will start by working on the interoperability of AMD's ROCm software stack for developing GPU accelerated applications with SiPearl's Rhea microprocessor. The aim is to ensure that a software ecosystem is in place ready for such applications to run on any future supercomputer based on SiPearl CPUs combined with AMD Instinct GPUs.

This joint work will focus on porting and optimization of AMD's Heterogeneous Interface for Portability (HIP), created to let developers write portable code to target both AMD and Nvidia GPUs, the duo said. Beyond this, AMD and SiPearl said they'll "engage with OEMs adopting the Rhea microprocessor with AMD Instinct accelerators in their designs," which simply means they will work with any hardware company considering using this combination of silicon to build an actual HPC system. The two companies also said they hope to expand the joint work to include European research institutions supporting research modeling tools, which would enable developers to port and optimize their applications for any system using a SiPearl/AMD solution.

Izvor: The Register

TSMC Expansion in Arizona to target 3-nm Node

Apple set to purchase chips from TSMC Arizona plant

[The Exiled]

Apple pushes TSMC to make more advanced chips in US by 2024

TSMC reportedly plans to make 4nm chips at its Arizona manufacturing plant when it goes online in 2024, a significant upgrade from the Asian foundry giant's previous commitment to focus on less advanced 5nm silicon at its first leading-edge US fab. The Taiwanese chipmaker has updated plans for its Phoenix fab at the behest of its customers including Apple, nVidia, and AMD. President Joe Biden is expected to join Apple CEO Tim Cook, nVidia CEO Jensen Huang, and AMD CEO Lisa Su at a ceremony in Arizona next Tuesday where TSMC will make the announcement.

TSMC is also anticipated to confirm at the event that it will construct a second fab at the Phoenix site to make 3nm chips, its most state-of-the-art silicon in production, according to the newsw wire. The company's founder, Morris Chang, said last week that TSMC was weighing the possibility. The developments mean the United States will have more advanced chipmaking capabilities sooner than anticipated. Intel is building out new fabs for advanced chips in Arizona and Ohio that are expected to go online in 2024 and 2025 respectively. Meanwhile, Samsung is constructing a leading-edge plant in Texas set for 2024 operation, and there's possibility of a massive expansion. Increased chipmaking capabilities and capacity have been top priorities of the US government, which seeks to rebalance the world's silicon supply chain and move reliance away from Asian manufacturing hubs by using $52 billion in subsidies approved this year to aid with new fab construction.

There has been a sense of urgency to move more chip production to the West, in part due to concerns that China could invade Taiwan and disrupt the island nation's chip manufacturers, which make up most of the world's advanced chipmaking capacity. The tech world has also been keen to diversify the global silicon supply chain due to disruptions caused by the COVID-19 pandemic. When Cook said earlier this month Apple plans to source chips from an Arizona fab in 2024, he said the goal was to reduce the iPhone maker's reliance on Taiwan fabs, according to an earlier report.

The accelerated ramp-up of TSMC's capabilities in the US is likely to create more pressure for Intel, which plans to compete with the Taiwanese company and other contract chip manufacturers like Samsung through its revitalized foundry business. Last week, we reported that Intel Foundry Services had a major leadership shakeup. Intel is racing to catch up with TSMC and Samsung in making more advanced chips after several years of manufacturing missteps. The company recently said it's on track to surpass its Asian rivals in process performance by 2025. But it may be a challenge for Intel to convince chip design rivals to use the company's manufacturing services.

Izvor: The Register

[Pupo]

Autor The Exiled

Apple pushes TSMC to make more advanced chips in US by 2024

Izvor: The Register

Priprema za možebitno Tajvan sra*e. Iako, sve je to mizerno, jedna(trenutno) tvornica u US a hrpa na Tajvanu i par u Kini.

[The Exiled]

Da, ali čak i obzirom na situaciju TSMC dosta brzo slaže nove pogone, dok eto Intel i dalje cmizdri, mada su dobili 52 milijarde pomoći, a konkurencija im je već prek puta i ne kasne.