zlacker

If you replace CUDA -> x86 and NVIDIA -> Intel, you'll see a familiar story which AMD has already proved it can work through.

These were precisely the arguments for 'x86 will entrench Intel for all time', and we've seen AMD succeed at that game just fine.

>>jvande+(OP)
And indeed more than succeed, they invented x86_64.

>>ianlev+t
And indeed more than succeed, they invented x86_64.

If AMD invented the analogous to x86_64 for CUDA, this would increase competition and progress in AI by some huge fraction.

>>jvande+(OP)
Transmetta was Intels boogey-man in the 90s.

>>jvande+(OP)
> These were precisely the arguments for 'x86 will entrench Intel for all time', and we've seen AMD succeed at that game just fine.

... after a couple decades of legal proceedings and a looming FTC monopoly case convinced Intel to throw in the towel, cross-license, and compete more fairly with AMD.

https://jolt.law.harvard.edu/digest/intel-and-amd-settlement

AMD didn't just magically do it on its own.

>>ianlev+t
x86_64's win was helped by Intel's Itanium misstep. AMD can't bank on Nvidia making a mistake, and Nvidia seems content with incremental changes to CUDA, contrasted with Intel's 32-bit to 64-bit transition. It is highly unlikely that AMD can find and exploit a similar chink in the amor against CUDA.

>>jvande+(OP)
If that's the model, it sounds like the path would be to burn money to stay right behind NVIDIA and wait for them to become complacent and stumble technically, creating the opportunity to leapfrog them. Keeping up could be very expensive if they don't force something like the mutual licensing requirements around x86.

>>sangno+q7
If they're content with incremental changes to CUDA then it doesn't cost much to keep updated compatibility and do it as quickly as any users actually adopt changes.

>>stcred+r3
Only works if NVidia misteps and creates the Itanium version of CUDA.

>>pjmlp+fe
You don't think someone would welcome the option to have more hardware buying options, even if the "Itanium version" didn't happen?

>>stcred+r3
Why "if only". Intel had been around forever when AMD showed up. CUDA isn't unassailable

>>jvande+(OP)
The difference is that AMD's CPUs are designed to implement the x86 and x86-64 ISA, so there is no loss of performance. In contrast, AMD and NVIDA's GPU instruction sets and architectures are not the same, and to get top performance out of these architectures code needs to be customized for them.

If you slap a CUDA compatibility layer on top of AMD, then CUDA code optimized for NVIDIA chips would run, but would suffer a performance penalty compared to code that was customized/tuned for AMD, so unless AMD GPUs were sold cheap enough (i.e. with low profit margin) to mitigate this loss of performance you might as well buy NVIDIA in the first place.