UnconfirmedUnverifiedRating 3/5
No atom-level source-text verification is available.
Fact check
In an Amazon demo, a model learned at the same rate on Trainium as on a standard chip.
Appears in
Verification ledger
- Checked by
- verification engine
- Atom checks
- 3/3
Verification held
This claim is held pending public source text we can check it against.
It will be re-verified once the cited source text is available.
Other claims in this story
- ConfirmedTorchNeuron enters public beta on 20 October at the PyTorch Conference in San Jose.
- ConfirmedTorchNeuron is open source under the Apache 2.0 licence.
- ConfirmedTorchNeuron supports Trainium2 and Trainium3 chips.
- ConfirmedThe first release does not yet support some PyTorch features, such as dynamic input shapes.
- ConfirmedAbout 22,000 PyTorch tests are skipped in the first release, according to the PyTorch blog post by Amazon engineers.
- ConfirmedAmazon says developers can target Trainium by changing a single device setting in existing PyTorch code.
- UnconfirmedAmazon says its custom chip business has passed a $25 billion annual revenue run rate.
- UnconfirmedAmazon was reported to be in talks to buy Decart, a TorchNeuron launch partner.
This fact check is published as schema.org ClaimReview structured data.