Define software interface, because Adreneline as an interface is miles in modernity and responsiveness compared to nvidia control panel+geforce experience.
The word you are looking for is features, because the interface is not what AMD is behind on at all.
Hence its a featureset. CUDA has a more in depth feature set because Nvidia is the leader and gets to dictate where compute goes, this in turn has a cyclical feedbackloop as devs use CUDA which locks them more and more into the ecosystem. Its a self inflicting problem till one bows out, and it wont be Nvidia.
It forces AMD to have to play catchup and write a wrapper that converts CUDA into OpenCL because the devs wont do it.
Ai is the interesting situation because when it came to the major libraries (e.g pytorch, tensorflow), they already have non Nvidia backends, and with microsofts desire to get AI compute to every pc, it makes more sense for them to partner with AMD/Intel due to the pc requiring a processor, while an nvidia gpu in the pc is not guaranteed. This caused more natural escape from requiring CUDA. If a project requires an Nvidia gpu, it rolls back that it was a small dev who programmed with CUDA for a feature and not the major library.
If we’re nitpicking about AMD: another thing I dislike about them is their smaller presence in the research space compared to their competitors. Both Intel and NVIDIA throw money into risky new ideas like crazy (NVM, DPUs, GPGPUs, P4, Frame Generation). Meanwhile, AMD seems to only hop in once a specific area is well established to have an existing market.
For consumer stuff, AMD is definitely my go-to. But it occurs to me that we need companies that are willing to fund research in Academia. Even if they don’t have a super good track record of getting profitable results.
deleted by creator
Define software interface, because Adreneline as an interface is miles in modernity and responsiveness compared to nvidia control panel+geforce experience.
The word you are looking for is features, because the interface is not what AMD is behind on at all.
deleted by creator
Hence its a featureset. CUDA has a more in depth feature set because Nvidia is the leader and gets to dictate where compute goes, this in turn has a cyclical feedbackloop as devs use CUDA which locks them more and more into the ecosystem. Its a self inflicting problem till one bows out, and it wont be Nvidia.
It forces AMD to have to play catchup and write a wrapper that converts CUDA into OpenCL because the devs wont do it.
Ai is the interesting situation because when it came to the major libraries (e.g pytorch, tensorflow), they already have non Nvidia backends, and with microsofts desire to get AI compute to every pc, it makes more sense for them to partner with AMD/Intel due to the pc requiring a processor, while an nvidia gpu in the pc is not guaranteed. This caused more natural escape from requiring CUDA. If a project requires an Nvidia gpu, it rolls back that it was a small dev who programmed with CUDA for a feature and not the major library.
deleted by creator
If we’re nitpicking about AMD: another thing I dislike about them is their smaller presence in the research space compared to their competitors. Both Intel and NVIDIA throw money into risky new ideas like crazy (NVM, DPUs, GPGPUs, P4, Frame Generation). Meanwhile, AMD seems to only hop in once a specific area is well established to have an existing market.
For consumer stuff, AMD is definitely my go-to. But it occurs to me that we need companies that are willing to fund research in Academia. Even if they don’t have a super good track record of getting profitable results.