Jeff King <peff@xxxxxxxx> writes: > On Fri, Aug 21, 2020 at 01:51:53PM -0400, Jeff King wrote: > >> That value was determined experimentally in 2012. I'm not sure of the >> exact reason it's different now (modern processors are better at >> parallelism, or modern git is better at parallelism, or the original >> experiment was just a fluke). But regardless, I can get on the order of >> a two-to-one speedup by bumping the thread count. See the final patch >> for timings and more specific discussion. > > After writing a response elsewhere in the thread, it occurred to me that > a good candidate for explaining this may be that our modern sha1dc > implementation is way slower than what we were using in 2012 (which > would have been either block-sha1, or the even-faster openssl > implementation). And since a good chunk of index-pack's time is going to > computing sha1 hashes on the resulting objects, that means that since > 2012, we're spending relatively more time in the hash computation (which > parallelizes per-object) and less time in the other parts that happen > under a lock. Believable conjecture that is. You could benchmark again with block-sha1 on today's hardware, but because the performance profile with sha1dc is what matters in the real world anyway... Thanks.
