AMD Watch posted an update
AMD’s ZenDNN optimisations are entering Intel’s oneDNN library, which supports CPU-based PyTorch inference, TechRadar reports. It is a concrete AI-software link between two chip rivals, beyond the companies’ broader discussions about possible cooperation.
Why it mattersThe practical point is that AMD’s optimisations could be available through software used with Intel processors. TechRadar gives no timing or implementation detail, so this is a useful signal, not a promise of a particular performance gain.
Discuss: Should AI software libraries make it easier to use one chipmaker’s optimisations on a rival’s processors, even if that blurs the lines between their competing platforms?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.