It is also other features such as GitHub workflow, releases, integration with other tools, webhooks etc. that makes it useful.
If it is unified memory, CPU can access the result of GPU processing without copying it to CPU memory (theoretically)
Way back in time, I used delta encoding for storing posting list (inverted index for search index). I experimented with using GPUs for decoding the posting list. It turned out that, as another reply mentioned copying…
Very true, but in recent years feature development has taken precedence over efficiency. VP of whatever says hardware is cheap, software engineers are not.
It is also other features such as GitHub workflow, releases, integration with other tools, webhooks etc. that makes it useful.
If it is unified memory, CPU can access the result of GPU processing without copying it to CPU memory (theoretically)
Way back in time, I used delta encoding for storing posting list (inverted index for search index). I experimented with using GPUs for decoding the posting list. It turned out that, as another reply mentioned copying…
Very true, but in recent years feature development has taken precedence over efficiency. VP of whatever says hardware is cheap, software engineers are not.