![]() |
![]() |
#1 |
Apr 2020
11000111012 Posts |
![]()
The latest revision to CADO includes improvements to the lattice reduction routine, including the addition of AVX-512 code for what I believe is the first time. There is a speedup of a few percent in sieving even on non-AVX-512 systems, so it should be worthwhile for everyone to upgrade - though of course it's possible there are bugs in the new code that are yet to be discovered.
I don't have access to an AVX-512 system myself, but I'd be interested to see how big the improvement is for those who do. |
![]() |
![]() |
![]() |
Thread Tools | |
![]() |
||||
Thread | Thread Starter | Forum | Replies | Last Post |
Maximum core count support in latest Prime95 | avinashkv94 | Information & Answers | 3 | 2021-09-09 15:26 |
Support Organizations | xilman | Soap Box | 3 | 2017-04-27 08:13 |
5+ GPU support | TheMawn | GPU Computing | 3 | 2014-07-13 02:31 |
I know this isn't an athletic support, but... | Skrek | Information & Answers | 3 | 2012-08-27 15:43 |
Support AVX | Unregistered | Information & Answers | 5 | 2011-07-05 17:12 |