Wednesday, May 13th 2020

Fujitsu Completes Delivery of Fugaku Supercomputer

Fujitsu has today officially completed the delivery of the Fugaku supercomputer to the Riken scientific research institute of Japan. This is a big accomplishment as the current COVID-19 pandemic has delayed many happenings in the industry. However, Fujitsu managed to play around that and deliver the supercomputer on time. The last of 400 racks needed for the Fugaku supercomputer was delivered today, on May 13th, as it was originally planned. The supercomputer is supposed to be fully operational starting on the physical year of 2021, where the installation and setup will be done before.

As a reminder, the Fugaku is an Arm-based supercomputer consisting out of 150 thousand A64FX CPUs. These CPUs are custom made processors by Fujitsu based on Arm v8.2 ISA, and they feature 48 cores built on TSMC 7 nm node and running above 2 GHz. Packing 8.786 billion transistors, this monster chips use HBM2 memory instead of a regular DDR memory interface. Recently, a prototype of the Fugaku supercomputer was submitted to the Top500 supercomputer list and it came on top for being the most energy-efficient of all, meaning that it will be as energy efficient as it will be fast. Speculations are that it will have around 400 PetaFlops of general compute power for Dual-Precision workloads, however, for the specific artificial intelligence applications, it should achieve ExaFLOP performance target.
K SuperComputer
Source: Riken
Add your own comment

12 Comments on Fujitsu Completes Delivery of Fugaku Supercomputer

#1
londiste
How is the performance of supercomputers measured for comparison? I mean, even the tidbit says 400 PFLOPS for DP workloads and ExaFLOP for AI.

Looking at the Green500 list:
www.top500.org/green500/lists/2019/11/
- A64FX is fairly obvious - a lot of ARM cores with dedicated hardware AI units.
- NA1 runs bunch of Xeons whose purpose is to support custom PEZY-SC2 processors (MIPS-based).
- AiMOS, Satori and Summit are using a bunch of PowerPC machines to support Nvidia Volta accelerators.
- The rest of the beginning of that list is Xeon or PowerPC CPUs with a lot of Nvidia Volta/Pascal accelerators.

Edit:
Performance is Linpack, specifically High Performance Linpack (www.netlib.org/benchmark/hpl/) with the best problem size for specific supercomputer.
Posted on Reply
#3
R-T-B
metalslaw
But can it run Crysis?
No. You need a monitor for that.
Posted on Reply
#4
steve360
Fugaku sounds like a word for Japanese kama sutra.
Posted on Reply
#5
windwhirl
steve360
Fugaku sounds like a word for Japanese kama sutra.
Actually, it's just another name for Mount Fuji.
R-T-B
No. You need a monitor for that.
:roll::roll::roll::roll:

Just take the gold, man.
Posted on Reply
#6
lexluthermiester
metalslaw
But can it run Crysis?
No. Crysis is not compiled for ARM.
Posted on Reply
#7
R-T-B
lexluthermiester
No. Crysis is not compiled for ARM.
I mean pretty sure this could just brute force emulate it. Seriously, the bigger issue would be system frame latency, not FPS. It'd probably be horrendous.
Posted on Reply
#8
The Quim Reaper
R-T-B
No. You need a monitor for that.
..All 'Supercomputers' have an interface terminal with a monitor.
Posted on Reply
#9
lexluthermiester
R-T-B
I mean pretty sure this could just brute force emulate it. Seriously, the bigger issue would be system frame latency, not FPS. It'd probably be horrendous.
Good point.
The Quim Reaper
..All 'Supercomputers' have an interface terminal with a monitor.
Yeah, but that is little more than a data relay. It's not a direct interface that anyone could game on.
Posted on Reply
#10
Moofachuka
wow, 150 thousand Athlon 64 FX CPUs...
Posted on Reply
#11
lexluthermiester
Moofachuka
wow, 150 thousand Athlon 64 FX CPUs...
Wow, you're missing a few things...
Posted on Reply
#12
DeathtoGnomes
lexluthermiester
Wow, you're missing a few things...
wait for it...
Posted on Reply
Add your own comment