3 ms·
I dunno. The section "Latency Tolerance, High Throughput and Little’s Law" has many applications in programming. Ever need to scale your cache, or size a connec
by cratermoon 3y ago
I dunno. The section "Latency Tolerance, High Throughput and Little’s Law" has many applications in programming. Ever need to scale your cache, or size a connection pool? Little's Law.
- abhi9u 3y agoThank you for reading! :)
- cratermoon 3y agoThanks for writing this. I note you mention that GPUs are massively parallel, and I think it would strengthen the article to add a paragraph or two discussing what kinds of loads lend themselves to GPU computing. The sort of "embarrassingly parallel" things like linear algebra vs inherently serial example of numerical analysis.
- abhi9u 3y agoYes, that's a good feedback. Few other people also mentioned this. I was trying to save space, to spend time more on the "how things work" part. But I see your point that some motivation behind parallelization would make it a more interesting read.