Well!
Here we are, 2024, and Apple is now using Google’s TPUs (Tensor Processing Units) for its Artificial Intelligence software.
Apple said on Monday that the artificial intelligence models underpinning Apple Intelligence, its AI system, were pretrained on processors designed by Google, a sign that Big Tech companies are looking for alternatives to Nvidia when it comes to the training of cutting-edge AI.
Apple’s choice of Google’s homegrown Tensor Processing Unit (TPU) for training was detailed in a technical paper just published by the company. Separately, Apple released a preview version of Apple Intelligence for some devices on Monday.
https://www.cnbc.com/2024/07/29/apple-says-its-ai-models-were-trained-on-googles-custom-chips-.html?
TPUs were first introduced in 2015 and were used for internal processing. Fast forward to today and TPUs are commercially available through Google’s cloud infrastructure. TPUv1 had TOPS (Tera Operations Per Second) of 23 while the more recent TPUv5p has TOPS of 459. They’ve come a ways! While TPUs aren’t going to make an appearance in an AI desktop or homegrown server anytime soon you should fully expect their increasing use.
They offer a good deal of performance while helping to break NVIDIA’s monopoly on AI processing power. Which is welcomed by everyone surely.
NVIDIA cannot be allowed to continue with business as usual.