Category: Fujitsu

Keynote: Fujitsu A64fx and Fugaku – A Game Changing, HPC AI Optimized Arm CPU

Posted by – November 2, 2019

Fugaku is the flagship next generation national supercomputer being developed by Riken R-CCS and Fujitsu in collaboration https://www.r-ccs.riken.jp/en/postk/project Fugaku will have hyperscale datacenter class resource in a single exascale machine, with more than 150,000 nodes of sever-class Fujitsu A64fx many-core Arm CPUs with the new SVE (Scalable Vector Extension) with low precision math for the first time in the world, accelerating both HPC and AI workloads, augmented with HBM2 memory paired with each CPU, exhibiting nearly a Terabyte/s memory bandwidth for both HPC and AI rapid data movements. A64fx has demonstrated up to 8 times the performance of mainstream “Platinum” server processors for HPC workloads, and this game-changing performance has been realized through extensive co-design process involving the entire Japanese HPC community resulting in a highly HPC-optimzed CPU architecture. Fugaku also will likely to be the premier big data and AI/ML infrastructure; currently, we are conducting research to scale deep learning to more than 100,000 nodes on Fugaku, where we expect to obtain near top GPU-class performance on each node. The talk will give an overview of A64fx/Fugaku as well as cover some of the software ecosystem efforts to enable high-end HPC / BD / AI.

Source: https://www.youtube.com/watch?v=A9605xIHdew

Fujitsu A64FX Post-K Supercomputer: World’s Fastest Arm Processor

Posted by – September 24, 2018

Fujitsu A64FX is the new fastest Arm processor in the world, built on 7nm it has 2.7 TFLOPS performance per chip suitable for high-end HPC and AI, they aim to create with it the world’s fastest supercomputer with it by 2021. A64FX is the first processor using the new Armv8-A Scalable Vector Extension (SVE) to accelerate a wide range of large-scale scientific computing, including deep learning. Fujitsu is working closely with Linaro to enrich the Arm HPC ecosystem. A64FX will be featured in the post-K computer, a supercomputer being developed by Fujitsu and RIKEN as a successor to the K computer, which achieved the world’s highest performance in 2011. The organizations are striving to achieve post-K application execution performance up to 100 times that of the K computer. It offers a number of features, including broad utility supporting a wide range of applications, massive parallelization through the Tofu interconnect, low power consumption, and mainframe-class reliability.

You can watch Fujitsu’s keynote at Linaro Connect here

Fujitsu Post-K ARM Supercomputer, Exascale by 2021

Posted by – November 20, 2017

Fujitsu is developing a very powerful ARM processor for its Post-K exascale supercomputer, to have a much wider impact on the HPC market than just a single system. Riken, Japan’s largest and most prestigious scientific research institute, will be the recipient of the Post-K system. This HPC optimized ARM processor design is being done in collaboration with ARM integrating SVE (Scalable Vector Extension), extending the vector processing capabilities associated with AArch64 (64bit) execution in the ARM architecture, enabling implementation choices for vector lengths that scale from 128 to 2048 bits, enabling High Performance Scientific Compute featuring advanced vectorizing compilers to extract more fine-grain parallelism from existing code to reduce software deployment effort. SVE also supports a vector-length agnostic (VLA) programming model that can adapt to the available vector length. When the Post-K Supercomputer is ready, which may be around 2020-2022, and if it lives up to its near-exascale performance promise, it will be eight times faster than today’s most powerful supercomputer in the world, China’s Sunway TaihuLight. The Post-K system will be used to model climate change, predict disasters, develop drugs and fuels, and run other scientific simulations. The Fujitsu Post-K ARM processors are likely to be 10nm FinFET chips fabricated by TSMC, and will feature high-bandwidth memory and the Tofu 6D interconnect mesh that was developed for the original K Supercomputer.

Filmed in 4K60 at Supercomputing 2017 in Denver using Panasonic GH5 ($1999 at Amazon.com) on firmware 2.1 (aperture priority, AF continuous tracking) with Leica 12mm f1.4 ($1297 at Amazon.com) with Sennheiser MKE440 stereo shotgun microphone ($325 at Amazon.com), get $25 off renting cameras and lenses with my referral link at https://share.lensrentals.com/x/wWbHqV