TorchTPU: Native PyTorch Backend for TPUs, 52% MFU Claimed

ax-trend issue #3377 · October 20, 2026

TorchTPU overview: Google and Meta co-development, Apache 2.0, 52% MFU

At PyTorch Conference 2026 on October 20, TorchTPU shipped as open source. It is a native PyTorch backend for TPUs, and the core claim is that existing PyTorch model and training code keeps running as-is. Google's vice presidents Mark Lohmeyer and Bill Jia introduced it on the official Google Cloud blog (October 20). A PyTorch-paradigm introduction ran on the pytorch.org blog the same day.

One-line summary for developers: Google and Meta's PyTorch teams co-developed TorchTPU over the project Google started in mid-2025. Marktechpost confirms the license is Apache 2.0 (October 20). The design explains itself as a bill for the four problems with PyTorch/XLA from 2020. The second news is that Meta uses it in production: its stated position is that it "cannot be locked into a single hardware architecture."

The Design Concept: Eager First, Plus a Compile Path

device = "tpu"  # was "cuda"
model = build_model().to(device)
for batch in loader:
    out = model(batch.to(device))
    loss = criterion(out)
    loss.backward()

The Numbers: Qwen 3 8B at 52% MFU via TorchTitan

TorchTPU numbers: Qwen 3 8B 52% MFU, near-linear scaling to 8,192 devices
Performance: Qwen 3 8B at 52% MFU, near-linear scaling to 8,192 devices

52% MFU is Google's announcement figure, not independent verification. Keywords to watch afterward are MFU on longer sequences and the variance of scaling efficiency by model size.

Four Problems With the Previous Path: PyTorch/XLA (2020)

PyTorch/XLA problems: invasive source edits, severe compile latency, no eager support, distributed primitives behind
PyTorch/XLA: four problems, and TorchTPU positioned as the replacement direction

The official announcement lists the existing path's issues, and they double as the design rationale for the new backend:

TorchTPU claims to solve all four: "eager first, compile as a production switch" (blog highlights, October 20).

The Ecosystem Story: vLLM, SGLang, LiteRT, One-Line Hugging Face Deployment

TorchTPU ecosystem: vLLM, SGLang, llm-d, TorchTitan, Ray, LiteRT
The ecosystem: backends and frameworks assembled

What reads bigger than the performance number is the ecosystem story. On the serving side sit vLLM-TorchTPU and SGLang-TorchTPU. Integration with llm-d, TorchTitan, Ray, TorchRec, VeRL, and Diffusers was announced alongside. llm-d 1.0 (#3369) shipped at the same conference.

# Hugging Face — deploy to a TPU in one line
from transformers import AutoModelForCausalLM
model = AutoModelForCausalLM.from_pretrained(name, device_map="tpu")

Extent of Meta's use is undisclosed: Meta production use is likely limited to a subset of workloads rather than the whole stack. Public ranking for the Ironwood TPU is not confirmed. TorchTPU is real open source (Apache 2.0, with review by community contributions). As an out-of-tree backend, its long-term maintenance participatory model is an open question.

Practical Takeaways: Selection Criteria for Tool Builders

Sources

PerformanceAX Trend

FAQ

What is TorchTPU?

TorchTPU is a native PyTorch backend for TPUs that shipped as open source at PyTorch Conference 2026 on October 20. Google and Meta's PyTorch teams co-developed it over the project Google started in mid-2025, and the license is Apache 2.0. Its core claim is that existing PyTorch model and training code keeps running as-is.

What code changes are needed to run a model on a TPU?

The migration story stays minimal: move the device from .to("cuda") to .to("tpu") and leave the rest essentially untouched. Eager execution comes first, with a compile path provided alongside for production through torch.compile. Hugging Face Transformers v5 also deploys models to TPUs with device_map="tpu" in one line.

What performance does TorchTPU claim?

The announcement claims up to 52% MFU on a Qwen 3 8B training run via TorchTitan, and near-linear scaling to 8,192 Ironwood TPUs. The 52% MFU figure is Google's announcement figure, not independent verification.

Who uses TorchTPU in production?

Meta is already a production user, and its LLM, advertising, and recommendation workloads have reportedly run on the new backend. The extent of Meta's use is undisclosed. Production use is likely limited to a subset of workloads rather than the whole stack.