ugrep 7.8.3 vs FFF - the Fuzzy showdown #554
Sanmayce
started this conversation in
Show and tell
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Hi Robert,
a text search aficionado here.
Wanted to have the Fastest Fuzzy Finder for my GUI search tool, so I asked Grok AI to propose the fastest such rival tools, so here comes the speed dance-off.
// Testmachine: My 'Suprapig' Dell Precision 7560, i7-11850H (16 threads), 4x32GB DDR4 3200 MT/s Rank: 2, GCC v15.2.1
// L1d: 384 KiB (8 instances)
// L1i: 256 KiB (8 instances)
// L2: 10 MiB (8 instances)
// L3: 24 MiB (1 instance)
// Haystack Size : 64.1 GiB, 'grokipedia_scrape.ndjson'
// Tested in Performance Mode, with highest niceness;
// Note1: Needle (18 bytes) "Sylvester Stallone";
// Note2: Needle (27 bytes) ""Nastassja Aglaia Nakszynski"";
The benchmark is fully reproducible:
I use the freely downloadable Grokipedia (64.1 GiB) as a haystack:
https://huggingface.co/datasets/htriedman/grokipedia-v0.1-dump/resolve/main/grokipedia_scrape.ndjson?download=true
https://github.com/user-attachments/files/30639626/FFF_vs_ugrep.tar.gz
I am in dismay how your single-threaded AVX512 ugrep outperformed my FFF in such a definitive way, still cannot process it, I was so sure FFF was superior - a cold shower for me, hee-hee.
All reactions