| | | AMD and Nvidia could soon get another major cloud chip challenger By Mayank Sharma 2 days ago
Google Cloud is pushing ahead with building its own chips
Google has made further progress in its move to create custom chips that can help ramp up its cloud computing infrastructure by hiring a former Intel veteran to head up the new division.
Google has a long history of building its own chips, to boost the performance of the technology to meet its demands, whether it be the Tensor Processing Units in 2015, or the Video Processing Units in 2018, or the security-centric OpenTitan in 2019.
“The future of cloud infrastructure is bright, and it’s changing fast. As we continue to work to meet computing demands from around the world, today we are thrilled to welcome Uri Frank as our VP of Engineering for server chip design,” wrote Google Fellow and VP of systems infrastructure Amin Vahdat, in a blog post.
System on a chip
In addition to announcing Frank’s arrival, Vahdat also shared details about Google’s plans to tackle the demand for the increasing cloud workloads.
He argues that while they’ve been focused on putting together computing components on the motherboard into an optimized system, this approach was no longer feasible.
“Instead of integrating components on a motherboard where they are separated by inches of wires, we are turning to Systems on Chip (SoC) designs where multiple functions sit on the same chip, or on multiple chips inside one package,” shared Vahdat.
Frank will head Google’s efforts into the SoC realm which they argue will help them achieve the levels of latency and bandwidth that are required to efficiently tackle the workloads of the day, while at the same time reduce power requirements and running costs.
Although Google hasn’t shared any further details about its SoC plans, we can assume it’ll probably license one from Arm, which might be interesting as it is one of the big tech companies that are protesting against Arm’s acquisition by Nvidia. |
|
