How “latency numbers everybody should know” decreased from 1990–2020

Re:How “latency numbers everybody should know” decreased fro... by throwaway81523 on Thursday March 03, 2022 @06:19PM (#30548426)

> Can you explain how to square that with the link saying compressing data hasn't gotten any faster since 2005?

Compression may be bottlenecked by ram (L2 or L3 cache) access speed, or maybe even dram (main memory) access speed.

[ Parent | on HN ]

Re:How “latency numbers everybody should know” decreased fro... by gallabytes on Thursday March 03, 2022 @06:22PM (#30548451)

That wouldn't make sense - compression has very cache-friendly access patterns, and would benefit greatly from the observed improvements in memory bandwidth.

[ Parent | on HN ]

Re:How “latency numbers everybody should know” decreased fro... by throwaway81523 on Thursday March 03, 2022 @09:02PM (#30549701)

That surprises me to hear. I would expect it to jump around in ram a lot. And at higher compression settings, some compressors use a lot of ram, more than will fit in cache.

[ Parent | on HN ]

Re:How “latency numbers everybody should know” decreased fro... by ______-_-______ on Thursday March 03, 2022 @06:28PM (#30548512)

I have an i7-3770k and an i9-12900k sitting here. Do you know of a benchmark I could run myself that would give me the same results on both CPUs?

[ Parent | on HN ]

Re:How “latency numbers everybody should know” decreased fro... by formerly_proven on Friday March 04, 2022 @05:07AM (#30552717)

sleep 1?

[ Parent | on HN ]