Apple is likely to use Google’s Tensor Processing Unit (TPU) to train its AI system, Apple Intelligence.
🟦 Apple Uses Google’s TPU for AI Training: Featured as an Alternative to NVIDIA
Apple has announced that the artificial intelligence model that forms the basis of its AI system “Apple Intelligence” has been pre-trained on a TPU processor designed by Google. This has been noted as a sign that big tech companies are looking for alternatives to NVIDIA’s expensive GPUs. Apple doesn’t name Google, but says it can improve efficiency and scalability by using 8192 TPUv4s for AI training.
🟦 Why did I choose Google’s TPU?
NVIDIA’s expensive GPUs dominated the high-end AI training chip market, and supply continued to struggle to keep up with rising demand. That’s why many tech companies, including Apple, are looking for cheaper and more accessible alternatives. Developed by Google, the Tensor Processing Unit (TPU) is a high-performance integrated circuit focused on machine learning that is available in the cloud, making it flexible and cost-effective.
🟦Summary
Apple’s adoption of Google’s TPUs for AI training has been touted as an alternative to NVIDIA’s GPUs. By utilizing Google’s TPU, Apple can improve the efficiency and scalability of AI training.
With the spread of NVIDIA’s GPU-independent new training method, it is expected that the competition in AI development will further intensify. This move could have a significant impact on the tech industry as a whole.
