4870 512mb by ASUS, i think its just a bug but i dont knowremo wrote:knapisch which card oo? wtf:D
got 1.1-1.2kkk hash/s whit 4870 1gb not oced
but.. b4 i got arroung 200m and more with my q9550 @3.4
now i only get 100m? bug?

Moderator: BarsMonster
4870 512mb by ASUS, i think its just a bug but i dont knowremo wrote:knapisch which card oo? wtf:D
got 1.1-1.2kkk hash/s whit 4870 1gb not oced
but.. b4 i got arroung 200m and more with my q9550 @3.4
now i only get 100m? bug?
Do you use latest 8.12 driver?knapish wrote:BarsMonster wrote:This is defenetly a leader
it it still working if the cal failed? cause i get kinda high also![]()
![]()
![]()
http://3.14.by/forum/download/file.php?mode=view&id=32
BarsMonster wrote:Do you use latest 8.12 driver?knapish wrote:BarsMonster wrote:This is defenetly a leader
it it still working if the cal failed? cause i get kinda high also![]()
![]()
![]()
http://3.14.by/forum/download/file.php?mode=view&id=32
It said falling back to CPU, but I've stripped CPU part of brook as it took 5 minutes to compile
So this might be the reason.
I should start bundling a faq with the programCatalyst® Version 08.11
Suddenly I am all confusedBarsMonster wrote: Yes, core clock(shader domain) increase BarsWF performance linearly.
I seemrb wrote:Suddenly I am all confusedBarsMonster wrote: Yes, core clock(shader domain) increase BarsWF performance linearly.Did you mean to write core clock, or shader clock in that sentence (they are 2 different things) ?
Yeah I was just confused when twobombs reported that overclocking the core clock increased BarsWF's perf (which is not normal, only the shader clock should impact performance). Anyway he now editted his post to make it clear he overclocked the shaders as well, that explains everything.BarsMonster wrote: I see
Well, shader clock is the most important
But one can do some benchmarks![]()
Actually, I need to update that screenshot with the new values we got after a few optimizations on the software and the hardware. Updated the cooling system and did some overclocks.Spaztikdude wrote:Nope, that's nowhere near the record.
http://blog.red-database-security.com/2 ... cer-barswf
3611.81 MHash/sec
Hmm, could you try latest videodriver? AFAIK it was fixedThe second core of the gtx295 didnt show up due to stupid vista requirement to have a real monitor plugged so cuda recognizes the gpu
You need to disable SLI for both of them to work in CUDA. Once you do that you'll get 1600 Mhash/s instead of 900 Mhash/s.dJe781 wrote:Edit : Erm, I guess that 700 Mhash/s means that both of my GPUs are working
intel i7 plus gtx295...s0l wrote:what are your specs dJe781?
Users browsing this forum: No registered users and 1 guest