1brc

1brc copied to clipboard

Running `bin/analyze measurements.txt` , I noticed that an identifier was included twice, with differing values for `min/mean/max`. From running `bin/analyze measurements.txt > out.txt` and then running the below python script...

immywing

parsing string without branching

1

Now parsing the int without explicit branch. In my tests this speeds up from (13.52 +- 1.18)s down to (13.07+-0.87)s measured across 15 runs. This is a speed-up of 3.39%....

wutzi15

bug: beyond the bound of chunk

1

``` // this assumes the file ends in a newline... const char *end = (char *)memchr(&data[chunk_end], '\n', chunk_size) + 1; ``` `data[chunk_end]` is over the limit, the end of the...

bbqz007

https://theartincode.stanis.me/008-djb2/ Maybe faster? @dannyvankooten https://r2p.dev/b/2024-03-18-1brc-go/

Grabber

I ran your code on Dual EPYC 9354 128 threads machines

3

``` // https://github.com/gunnarmorling/1brc/discussions/46 #include #include #include #include #include #include #include #include #include #include #include // Capacity of our hashmap // Since we use linear probing this needs to be at...

lehuyduc

What's the memory bandwidth of your machine?

3

Curious about the memory bandwidth of your machine. If you can, I'd be interested in the results of the C++ program below. Compile with full optimization of course. #include #include...

kennethchiu

runxiyu

Why not SIMD?

1

Perhaps it's a silly question, but why not SIMD? I think it would actually make the challenge much more interesting combine with the no external deps. Even because unless you...

danielealbano

Inconsistent output

1

I was looking to test the performance impact of several parts of the program and was setting up some tests to ensure I didn't break anything while testing. However, I...

QazCetelic