Fujitsu’s Fugaku and A64FX Take Arm to the Top with 415 PetaFLOPs

See also closely related <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=23601098" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=23601098</a> from yesterday.

This year seems to be the year of big serious ARM performance.<p>- AWS Graviton 2 is actually a viable option in the public cloud.<p>- Apple starts transitioning their workhorses to ARM.<p>- First place in the TOP500, apparently not due to custom magic or GPUs, but due to <i>already existing&#x2F;standardized</i> ARM vector extensions.<p>ARM is becoming such a household name that Fujitsu appears to have traded getting a paid ARM license over the royalty free SPARC instruction set they used to work with. I hope POWER can survive this onslaught, as SPARC surely has no hope of survival after this.

Did Apple convince these guys and TSMC (re 5nm production) to make their big ARM announcements the same day ARM Macs are announced?<p>If it&#x27;s not great marketing then it&#x27;s a heck of a coincidence.

And on that note, A64FX software stack also makes use of McKernel an an OS alternative.<p><a href="https:&#x2F;&#x2F;github.com&#x2F;RIKEN-SysSoft&#x2F;mckernel" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;RIKEN-SysSoft&#x2F;mckernel</a><p><a href="https:&#x2F;&#x2F;www.sys.r-ccs.riken.jp&#x2F;ResearchTopics&#x2F;os&#x2F;mckernel&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.sys.r-ccs.riken.jp&#x2F;ResearchTopics&#x2F;os&#x2F;mckernel&#x2F;</a>

The thing that surprised me most is that when compared to Summit, the previous #1 powered by POWER9 cores and Nvidia GV100 accelerators, Fugaku seems to show the exact same power efficiency.<p>&gt; The new Fugaku supercomputer is bigger than Summit in practically every way. It has 3.05x cores, it has 2.8x the score in the official LINPACK tests, and consumes 2.8x the power<p>So were the Nvidia accelerators what drove up Summit&#x27;s efficiency?

I remember back in university when I first read about the ARM business model, I naively thought it will always be the second fiddle to intel x86. At that time x86 had such a strong hold that for my young brain it was unimaginable that anyone else will capture the CPU market apart from embedded systems.<p>The year 2020 is the conclusion of a long push by ARM and now I am wondering can RISC-V will be taking a similar path.

It appears that the Fujitsu A64FX is planned for Cray as well as its own A64FX machines, the PRIMEHPC FX1000 and PRIMEHPC FX700 models.<p><a href="https:&#x2F;&#x2F;www.hpcwire.com&#x2F;2019&#x2F;11&#x2F;12&#x2F;cray-fujitsu-both-bringing-fujitsu-a64fx-based-supercomputers-to-market-in-2020&#x2F;" rel="nofollow">https:&#x2F;&#x2F;www.hpcwire.com&#x2F;2019&#x2F;11&#x2F;12&#x2F;cray-fujitsu-both-bringin...</a><p>&gt;The new HPE-Cray system, part of the Cray CS500 lineup, will employ the Fujitsu A64FX Arm-based processor with Arm Scalable Vector Extensions (SVE) and second-generation high-bandwidth memory (HBM). Named as customers in today’s release are Los Alamos National Laboratory, Oak Ridge National Laboratory, RIKEN Center for Computational Science, Stony Brook University, and University of Bristol. Cray and Fujitsu said they will be exploring engineering collaboration, co-development, and joint go-to-market strategies to meet customer demand as supercomputing extends into the exascale era.

Are we getting to the point where we GPUs might be replaced by larabee style general purpose CPUs?

How does it fare in terms of energy efficiency? Or is this not a concern at all?