Date: Sat, 11 Aug 2012 04:05:22 +0400 From: Solar Designer <solar@...nwall.com> To: john-dev@...ts.openwall.com Subject: Re: Current -fixes GPU formats vs TS On Sat, Aug 11, 2012 at 01:59:14AM +0200, Lukas Odzioba wrote: > I wasn't able to reproduce this bug that way (running xsha first) but > it I was able to get 100% self test fails during ./john -test=0. > This speedup is weird what is c/s, 3% seems to be more than error > dispersion. I have no idea what is the source of this speedup. Maybe > some powersaving mechanisms inside cuda impelentation or device > itself. > > On gtx470 I get 466->470 after adding memset. Here's a guess: maybe the sequential writes to that memory - all at once - prevent cache line loads from it later, when we make partial writes first (different portions by different threads). Alexander
Powered by blists - more mailing lists
Confused about mailing lists and their use? Read about mailing lists on Wikipedia and check out these guidelines on proper formatting of your messages.