Diagram
A decision tree that checks matrix dominance, custom operations, branching and precision needs, and scale to decide between Cloud TPU and CPU or GPU.
Workload Patterns That Fit and Do Not Fit TPUs
Workload patterns that fit TPUs are jobs with steady, large-scale math where training data flows from storage such as Cloud Storage to the TensorFlow training process running on the TPU. Jobs with large matrix computations benefit most from MXUs, HBM, and slices.
Models well suited for TPUs include:
Domain 6 of 6
27 practice questions
Practice this DomainLast updated