Tensor Processor Unit (TPU)
- is an AI acceleratorASIC developed by Google specifically for neural networkmachine learning, particularly using Google’s own TensorFlow software
- Google began using TPUs internally in 2015, and in 2018 made them available for third party use, both as part of its cloud infrastructure and by offering a smaller version of the chip for sale
TPU Use Case
The current generation of TPUs are not really great for training networks, they’re more focused on executing predictions with them after training
TPU vs CPU
A TPU is a co-processor, it cannot execute code in its own right, all code execution takes place on the CPU which just feeds a stream of micro-operations to the TPU
TPU vs GPU
The main difference is that TPUs are cheaper and use a lot less power, and can thus complete really large prediction jobs cheaper than GPUs, or make it simpler to use prediction in a low-latency service.