Hello
I have a bunch of very large files that can be processed in parallel. Can I do this efficiently using seperate cores of a GPU. What kind of performance enhancement can I expect?? Will I eventually be limited by file I/O.
Saleem
Hello
I have a bunch of very large files that can be processed in parallel. Can I do this efficiently using seperate cores of a GPU. What kind of performance enhancement can I expect?? Will I eventually be limited by file I/O.
Saleem
Whether you’ll be bound by file I/O depends on how compute intensive is the processing and how much parallelism can you extract. “Processing large files” can mean literally anything :)
How large ?
Are they binary or do they need to be parsed (e.g. text to float) ?
What is your hardware please