AI Hypercomputer
Reference implementations, benchmarks, recipes, and all things Google Cloud AI Hypercomputer
Popular repositories Loading
-
-
gpu-recipes
gpu-recipes PublicRecipes for reproducing training and serving benchmarks for large machine learning models using GPUs on Google Cloud.
Repositories
Showing 10 of 23 repositories
-
- maxdiffusion Public
- ml-goodput-measurement Public
- accelerator-agents Public
- accelerator-microbenchmarks Public
TPUMS (TPU Microbenchmark Suite) is a modular, high-fidelity benchmarking and profiling framework designed to evaluate the compute, memory, and interconnect performance of Cloud TPUs using JAX.
- tpu-codesign Public
Hardware-grounded model architecture design principles, MoE co-design patterns, low-precision numerics, and parallelism guides for training and serving on Google Cloud TPUs.
- gpu-recipes Public
Recipes for reproducing training and serving benchmarks for large machine learning models using GPUs on Google Cloud.
- google-cloud-mldiagnostics Public