Add Google Cloud TPU to get_peak_flops#3900
Open
wconstab wants to merge 1 commit into
Open
Conversation
get_peak_flops covers NVIDIA, AMD, Intel, and AWS accelerators but has no
TPU entry, so TPU runs hit the A100 fallback (312 TFLOPS) and report MFU
inflated by ~3.7x (e.g. TPU v7 peak is 1153.5 TFLOPS/device). This adds a
TPU section keyed on the device name reported by torch_tpu's
TpuDeviceModule.get_device_name() ("TPU v7", "TPU v5p", ...).
Values are dense BF16 MXU peak per device as the XLA/PJRT runtime exposes
it: one TensorCore on v7, one chip on single-core/megacore generations.
That is the per-rank granularity MFU divides by, so no per-core division is
applied. Sources: cloud.google.com/tpu/docs.
meta-cla
Bot
added
the
CLA Signed
fegin
approved these changes
Jul 10, 2026
tianyu-l
approved these changes
Jul 10, 2026
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
In preparation for TPU enablement...
Values are dense BF16 MXU peak per device as the XLA/PJRT runtime exposes it: one TensorCore on v7, one chip on single-core/megacore generations. That is the per-rank granularity MFU divides by, so no per-core division is applied. Sources: cloud.google.com/tpu/docs.