On: Matrix multiplication inches closer to mythic goal

[colanderman]

> for small problem size that's exactly what they are

O notation does not apply to small problems. It is strictly concerned with asymptotic behavior.

[anon_tor_12345 → colanderman]

you're completely missing the point (again). the point is that complexities (even asymptotic) of sequential algorithms don't apply to parallelizations of those algorithms.

[aquadrop → anon_tor_12345]

Only for infinite number of parallelized processors?