|
WHY
IT’S IMPORTANT
Google's tensor processing units (TPUs) are one of the few
viable alternatives to the advanced AI chips made by Nvidia,
though developers can only access them through Google's Cloud
Platform and not buy them directly.
Google plans to offer the Arm-based central processing unit
(CPU) called Axion via Google Cloud. The company said it has
superior performance to x86 chips, and general-purpose Arm chips
in the cloud.
KEY QUOTE(S)
"We're making it easy for customers to bring their existing
workloads to Arm," said Mark Lohmeyer, Google Cloud's vice
president and general manager of compute and machine learning
infrastructure. "Axion is built on open foundations but
customers using Arm anywhere can easily adopt Axion without
re-architecting or re-writing their apps."
CONTEXT
Rival cloud operators such as Amazon.com and Microsoft have
built Arm CPUs as a way of differentiating the computing
services they offer. Google has built other custom chips for
YouTube, AI and its smartphones but had not built a CPU.
Broadcom has partnered with Google on prior generations of the
TPU chips. Google declined to comment on whether it used a
design partner for Axion and Broadcom's involvement with the TPU
v5p.
BY THE NUMBERS
The Alphabet subsidiary said the new TPU v5p chip is built to
run in pods of 8,960 chips, and can achieve twice the raw
performance as the prior generation of TPUs. To help ensure the
pod runs at optimal performance, Google uses liquid cooling.
The Axion chip offers 30% better performance than
"general-purpose Arm chips, and 50% better performance than
current generation x86 chips produced by Intel and Advanced
Micro Devices.
WHAT’S NEXT
Google plans to use Axion to power services such as its YouTube
Ads in Google Cloud "soon." The TPU v5p is generally available
via Google's cloud on Tuesday.
(Reporting by Max A. Cherney in San Francisco; Editing by
Marguerita Choy)
[© 2024 Thomson Reuters. All rights
reserved.]
This material may not be published,
broadcast, rewritten or redistributed.
Thompson Reuters is solely responsible for this content.
|
|
