Giulio Ermanno Pibiri

Hi @rob-p, a small update on this matter: PTHash benchmark with and without options `-march=native -mbmi2 -msse4.2`. # Without -march=native -mbmi2 -msse4.2 {"n": "10000000", "c": "7.000000", "alpha": "0.990000", "minimal": "true",...

These are the results for 100M keys: # With -march=native -mbmi2 -msse4.2 {"n": "100000000", "c": "7.000000", "alpha": "0.990000", "minimal": "true", "encoder_type": "compact-compact", "num_partitions": "1", "seed": "1234567890", "num_threads": "1", "external_memory": "false",...

I also made the following experiment: # With -mbmi2 -msse4.2 but without -march=native {"n": "100000000", "c": "7.000000", "alpha": "0.990000", "minimal": "true", "encoder_type": "compact-compact", "num_partitions": "1", "seed": "1234567890", "num_threads": "1", "external_memory":...

Hi @adamant-pwn and thanks for this! I quickly skimmed through the commits and the code seems really well done. So, as far as I see, you use a template for...

Hi @adamant-pwn, thanks for your willingness in testing this out! You're right, sorry for the confusion, I was referring to `__m256i`. In general, it would be very nice to be...

Hi @adamant-pwn, thank you very much for the update! Nice that the generalization does not compromise anything for k

Hi @adamant-pwn, wow that's very cool! So, if I'm understanding correctly, you use some template magic to let the compiler define recursively the bitwise operators working over two halves. Sweet....

> And I do hope that it is genuine and not a result of some bug in the code that misses something or not accounts for something when k goes...

Hi @adamant-pwn, ok that makes sense! I think this is the correct result. I would still perform a check to see if all kmers are found, etc. (just add --check...

I think you did everything right. The output of GGCAT is some lz file that should be decompressed and given to SSHash as input. The point is that there must...