TPU transformation: A look back at 10 years of our AI-specialized chips

Re:TPU transformation: A look back at 10 years of our AI-spe... by walterbell on Saturday August 03, 2024 @09:46PM (#41150608)

Apple Intelligence uses Google TPUs instead of GPUs.

[ Parent | on HN ]

Re:TPU transformation: A look back at 10 years of our AI-spe... by bigcat12345678 on Saturday August 03, 2024 @10:23PM (#41150754)

That's something not surprising, given JG and Ruoming's Google stint.

Google is going to dominate LLM ushered AI era. Google has been AI first since 2016, they just don't have the opening. Sam, as inapt at engineering, just has no idea how to navigate the delicate biz & eng competitions.

[ Parent | on HN ]

Re:TPU transformation: A look back at 10 years of our AI-spe... by j16sdiz on Saturday August 03, 2024 @10:21PM (#41150751)

If you read all their paper, they use a mix of them

[ Parent | on HN ]

Re:TPU transformation: A look back at 10 years of our AI-spe... by dlewis1788 on Sunday August 04, 2024 @10:09AM (#41153663)

For training, yes, but no indications on inference workloads. Apple has said they would use their own silicon for inference in the cloud.

[ Parent | on HN ]

Re:TPU transformation: A look back at 10 years of our AI-spe... by walterbell on Sunday August 04, 2024 @07:34PM (#41157033)

Plus the Apple "Neural Engine" which has shipped on millions of iPhones for local inference.

[ Parent | on HN ]