3 ms·
Very large matrices is >50kx50k (maybe even bigger). Strassen is then just a tiny bit better when implementing both cache and allocation aware.(Tested in Fortra
by rrobukef 10y ago
Very large matrices is >50kx50k (maybe even bigger). Strassen is then just a tiny bit better when implementing both cache and allocation aware.(Tested in Fortran)
At least it's at a non practical size.
- taeric 10y agoIt is always interesting to see just how valuable a "tiny bit" can be, though. Granted, I confess most is almost always navel gazing. Increases in the "best" sorting/map algorithm have probably not lead to any noticeable improvements in modern codebases. Usually, those benefit more from being the low hanging fruit, not most computationally complex. (That is, I'm sure in these large matrix problems, there are other low hanging fruit that will yield better improvements.)