Rendered at 20:30:46 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
yu3zhou4 12 hours ago [-]
Adding a support for new hardware to PyTorch is actually quite convenient. I did that with WebGPU using the same PrivateUse1 mechanism TorchTPU used. Every hardware has its own slot and identifier, and when you want to add a support for a new one without merging it into PyTorch, PrivateUse1 works essentially like plug-in slot
So Google is basically admitting PyTorch/XLA on TPUs didn't work — TorchTPU looks like them rebuilding what should have worked day one. Its hard to run production ML on a toolchain engineers can't trust, no matter how fast the silicon is.
guipsp 4 hours ago [-]
This does still look like it uses XLA, fwiw.
in-silico 21 hours ago [-]
This is great to see.
I did trained some research models using the existing PyTorch/XLA on TPUs, and it was a mess of undocumented behavior and bugs (silently hanging after 8 hours of training!).
If anyone is trying to use PyTorch on TPU before TorchTPU is released, you can check out the training pipeline that I ended up building to support my research: https://github.com/aklein4/easy-torch-tpu
isusmelj 10 hours ago [-]
Is it just me, or does it feel like everyone now uses AI to write any kind of blog?
These parts here somehow trigger me:
- Enter TorchTPU. As an engineering team, our mandate was to build a stack that leads with usability, portability, and excellent performance.
- Engineering the TorchTPU Stack: The Technical Reality
- Eager First: Flexibility Without Compromise
- The breakthrough, however, is our fused eager mode.
- The Road Ahead: 2026 and Beyond
I have mixed feelings about this. On one hand, we all seem to be using the same tools and converging to the same style. On the other hand, if we all use the same models with the same system prompts, we might lose a lot of creativity and diversity in online content.
pjmlp 8 hours ago [-]
More like, everyone now has KPIs and OKRs attached to AI for their quartal appraisals.
If you don't, I envy you.
thevillagechief 8 hours ago [-]
This right here is exactly right. If it hasn't come for you yet, it's coming.
amelius 5 hours ago [-]
Yeah but everyone also uses AI to summarize blogs for them.
PunchTornado 10 hours ago [-]
it's just sad
Reubend 21 hours ago [-]
Sounds good, but my main question is: is this a fork, or a new backend they're building in (like MPS)?
yu3zhou4 12 hours ago [-]
They write that they use PrivateUse1, so it’s a custom out-of-tree backend
musebox35 16 hours ago [-]
I attended the related session at Next’26 yesterday. From my understanding it is a new backend and they will release the torch tpu source on github in one or two months. It will not support all ops initially but they are moving fast. Still for a while torchax is mature enough to run torch models on tpus by translating to jax.
noracists 17 hours ago [-]
Very excited for this.
MASNeo 14 hours ago [-]
Now all that’s missing is an actual chip that can be purchased. Any ideas?
sterlind 12 hours ago [-]
I'm thinking of picking up some used Gaudis from eBay. they're pretty TPU-like. but other than oddball hardware like that it's just the GPU duopoly and proprietary bespoke stuff the hyperscalers have made for themselves.
shit, maybe China will start selling Huawei Ascend chips internationally.
20 hours ago [-]
immanuwell 12 hours ago [-]
pitch basically boils down to 'just change one line and it works' which sounds too good to be true, but if they actually pull it off at 100k-chip scale, that's genuinely a big deal
https://github.com/jmaczan/torch-webgpu
I did trained some research models using the existing PyTorch/XLA on TPUs, and it was a mess of undocumented behavior and bugs (silently hanging after 8 hours of training!).
If anyone is trying to use PyTorch on TPU before TorchTPU is released, you can check out the training pipeline that I ended up building to support my research: https://github.com/aklein4/easy-torch-tpu
These parts here somehow trigger me:
- Enter TorchTPU. As an engineering team, our mandate was to build a stack that leads with usability, portability, and excellent performance.
- Engineering the TorchTPU Stack: The Technical Reality
- Eager First: Flexibility Without Compromise
- The breakthrough, however, is our fused eager mode.
- The Road Ahead: 2026 and Beyond
I have mixed feelings about this. On one hand, we all seem to be using the same tools and converging to the same style. On the other hand, if we all use the same models with the same system prompts, we might lose a lot of creativity and diversity in online content.
If you don't, I envy you.
shit, maybe China will start selling Huawei Ascend chips internationally.