Why are so many of the GPU threads even considering AMD when AMD can't even into CUDA?
AMD is actually bringing in CUDA support soon™.
Vulkan is better
>>54546256
AMD can into OpenCL. OpenCL on Nvidia is hell.
Also AMD can into ATS, Nvidia can't.
>>54546287
But OpenCL is inferior
>>54546300
>implying implications
>>54546287
Is it that bad? Trying to figure out which platform would be more practical for physics modeling at home.
They should fall so Nvidia they can price up their shit.
Sounds like heaven.
Specifications for M970W
memory size: 3.5gb + 0.5gb,
>3.5 + 0.5
what? should i get a 950 instead?
>>54547265
a 980 seems better.
>>54546256
most people don't need cuda
>>54547265
GTX 1070
Why are so many of the GPU threads even considering Nvidia when Nvidia can't even into hardware async compute?
>$CURRENT_YEAR
>NO ASYNC COMPUTE
>STILL GIMPED PERFORMANCE ON DX12
Nvidicucks back to >>>/v/ please
>>54548382
This is a technology board though
>>54546256
I fucking wonder OP I fucking wonder why the whole world has not started using CUDA. With the ginormous 275 apps no one uses you would think people would go nvidina till I die.
>>54548682
Cuda is a meme tho