2 ms·
There are no viable alternatives if you require high performance or even just the software to actually work properly.
by andersa 3y ago
There are no viable alternatives if you require high performance or even just the software to actually work properly.
- jvanderbot 3y agoYes - this is a competition problem. If there were viable alternatives, then interoperability wouldn't be something we really talk about. Instead it'd be abstraction layers to run on the chip-specific runtimes. And then it's just CUDA trying to "beat" the alternative.
- bee_rider 3y agoTrying to best Nvidia’s language on Nvidia’s hardware seems pretty hopeless. I wonder if that’s the wrong abstraction layer? There exists stuff like CUBLAS, which is of course using CUDA under the hood, but it also is something like a BLAS. Maybe as the AI/ML world keeps developing people will be grow more stubborn about sticking to frameworks. We probably just need a couple rounds of people getting burned by vendor lock-in, I guess.