EDIT: I just realized that I don't know what you mean by 'FP' projects, so I don't know if my reply is valid. Floating-point?
yes, i believe he is referring to "floating point"...of course i think he meant to specify FP64 (double precision floating point), as opposed to FP32 (single precision floating point).
I thought for sure that einstein@home used double-precision math? I guess I was wrong...
You are right, it does not. AFAIK only MW uses DP.
PrimeGrid GeneferCUDA also requires a double-precision gpu and is very sensitive to any overclocked chips, even factory gpu overclocks. GeneferCUDA doesn't tolerate even the slightest of errors.
The funny thing, i tried and finished one (~30hours), with a (little damaged on one or two memorychips witch are not very heattolerable anymore) overclocked 285GTX. Seems i got luck or something ^^
EDIT: I just realized that I don't know what you mean by 'FP' projects, so I don't know if my reply is valid. Floating-point?
yes, i believe he is referring to "floating point"...of course i think he meant to specify FP64 (double precision floating point), as opposed to FP32 (single precision floating point).
Sure! Extract me plz the best values on eatch card plz per PM or something (there are two values, dont know exactly what to take), dont have the time to fully read such big threads, im only very short online every day at home (going offline for today ;)). thx :)
7750 in OSX 10.8.3 has a mean GPU runtime of 5400 seconds with 1 task and one core free for the GPU. 8400 with two tasks and still one core free. this is with two E5462's at 2.8GHz.
RE: EDIT: I just realized
)
yes, i believe he is referring to "floating point"...of course i think he meant to specify FP64 (double precision floating point), as opposed to FP32 (single precision floating point).
RE: RE: RE: I thought
)
The funny thing, i tried and finished one (~30hours), with a (little damaged on one or two memorychips witch are not very heattolerable anymore) overclocked 285GTX. Seems i got luck or something ^^
DSKAG Austria Research Team: [LINK]http://www.research.dskag.at[/LINK]
RE: RE: EDIT: I just
)
Oops, yep I typed FP and meant DP.
A User had the wish to show
)
A User had the wish to show credits per Day too. I looked into it and showing that now too from 1-3WUs in parallel.
[LINK]http://www.dskag.at/images/Research/EinsteinGPUperformancelist.pdf[/LINK]
Have fun :)
DSKAG Austria Research Team: [LINK]http://www.research.dskag.at[/LINK]
RE: A User had the wish to
)
are you planning on adding the missing GTX 560 Ti, GTX 580, and GTX 670 values that i provided previously in message 123864 to your .pdf file?
Sure! Extract me plz the best
)
Sure! Extract me plz the best values on eatch card plz per PM or something (there are two values, dont know exactly what to take), dont have the time to fully read such big threads, im only very short online every day at home (going offline for today ;)). thx :)
DSKAG Austria Research Team: [LINK]http://www.research.dskag.at[/LINK]
you got PM ;-)
)
you got PM ;-)
7750 in OSX 10.8.3 has a mean
)
7750 in OSX 10.8.3 has a mean GPU runtime of 5400 seconds with 1 task and one core free for the GPU. 8400 with two tasks and still one core free. this is with two E5462's at 2.8GHz.
Has anyone first benchmarks
)
Has anyone first benchmarks of the new HD7790?
Theoretically it should give > 2 results within one h with a power draw of < 80 Watt.
RE: Has anyone first
)
Jeearr have one HD7790 with recent results:
http://einsteinathome.org/host/6014248/tasks&offset=0&show_names=1&state=3&appid=0
It's 20% faster (comparing time for one unit) than his old "CapeVerde" card(HD7770 or HD7750). From timings of HD7770 in pdf file here, I think his HD7790 is working with 2 units simultaneously, and 2u <3200s.