Loading…
Linaro Tech Days has ended
Linaro Tech Days will feature some of the content from our cancelled event Linaro Connect Budapest remotely. These will be held via Zoom Webinar and livestreamed to Youtube. Joining information is available in each session description. If you cannot view the session please make sure you are registered and logged in.


PLEASE NOTE: ALL SESSION TIMES ARE LISTED IN UTC


Tuesday, March 24 • 09:00 - 09:25
LTD20-100K Keynote: A64FX: Power efficient design for HPC and AI

Log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
To join this session live please go to:

Description:
Fujitsu developed the A64FX processor that is optimized for HPC (High Performance Computing) and the AI(Artificial Intelligence).
A64FX is used in Fugaku supercomputer which is a massively parallel supercomputer and a successor to K-computer.
A64FX is based on Fujitsu microarchitecture, as used in our SPARC64 and mainframe processor development.
It provides the world's top-class flops/watt, computing performance and memory bandwidth. A64FX Prototype took 1st place in the Green500 November 2019.
To achieve high performance and low power consumption in a wide range of actual applications, we collaboratively work with RIKEN (co-design).
For ISA (Instruction Set Architecture), Fujitsu chose to adopt the Armv8-A with SVE (Scalable Vector Extension) to best position the Fugaku to utilize and contribute to a broader user base.
Fujitsu has collaborated with Arm as a lead partner and contributed to the development of SVE of the HPC and AI extension for the Armv8-A.
I would explain the processor features from the standpoint of microarchitecture and ISA.

Speakers
avatar for Takekazu Tabata

Takekazu Tabata

Staff engineer, Fujitsu Limited
Takekazu Tabata is an engineer at Fujitsu. His research interests include microprocessor design. Especially, he is in charge of  high-performance core microarchitecture design and instruction set architecture.



Tuesday March 24, 2020 09:00 - 09:25 UTC
Track 1 [Tuesday]