David Edelsohn <dje....@gmail.com> writes: > Why is it early to use POWER10 instructions? Various BLAS libraries > have been updated to use POWER10 instructions.
I'm not so familiar with the family of power processors, and I don't know which versions are the most common in today's running machines. But in general, I think it makes sense to first target the oldest power version which (i) is in wide use, and (ii) where we can get significant speedup compared to the C implementation. That's perhaps Power9 or Power7? (It may vary per algorithm, depending on which processor features are important to gain that speedup for each algorithm). And then we can have additional code for newer versions, when that brings significant *additional* speedup over the first version, and let the fat machinery figure out which implementation to use. And of course, also depending on contributor's judgement on what's useful and interesting. Regards, /Niels -- Niels Möller. PGP key CB4962D070D77D7FCB8BA36271D8F1FF368C6677. Internet email is subject to wholesale government surveillance. _______________________________________________ nettle-bugs mailing list -- nettle-bugs@lists.lysator.liu.se To unsubscribe send an email to nettle-bugs-le...@lists.lysator.liu.se
