News for hackers. Stuff that matters.

A comment on Apple's On-Device and Server Foundation Models.

Re:Apple's On-Device and Server Foundation Models by ra7 on Monday June 10, 2024 @06:46PM (#40640131)
> Our foundation models are trained on Apple's AXLearn framework, an open-source project we released in 2023. It builds on top of JAX and XLA, and allows us to train the models with high efficiency and scalability on various training hardware and cloud platforms, including TPUs and both cloud and on-premise GPUs.

Interesting that they’re using TPUs for training, in addition to GPUs. Is it both a technical decision (JAX and XLA) and a hedge against Nvidia?

[ Parent | on HN ]
Re:Apple's On-Device and Server Foundation Models by m-s-y on Monday June 10, 2024 @07:48PM (#40640666)
They’d be silly not to hedge. Anyone, in fact, would be silly. It to hedge. On pretty much everything.
[ Parent | on HN ]
Re:Apple's On-Device and Server Foundation Models by anvuong on Monday June 10, 2024 @09:39PM (#40641360)
Jax was built with TPUs in mind, so it's not surprising that they use TPUs
[ Parent | on HN ]
Re:Apple's On-Device and Server Foundation Models by gokuldas011011 on Tuesday June 11, 2024 @01:29AM (#40642692)
"Use the best tool available"
[ Parent | on HN ]
Re:Apple's On-Device and Server Foundation Models by flakiness on Tuesday June 11, 2024 @05:40AM (#40644297)
They hired people nearby. Conveniently there is a small town called Mountain View.
[ Parent | on HN ]