Seti is 10 times faster with Cuda, why not RNA ?

[Michael H.W. Weber]

Because the types of calculations are completely different. The current RNA World applications do not profit from GPU-based computations and assigning these to GPUs would just be a waste of your (increased) electricity costs. However, future RNA World applications will quite likely profit from GPUs and we will see tha we uitilzed these when the time has come.

Michael.

[The Knighty NI]

Only just come back online due to puter problems outside of my control. Or more likely due to me not thinking the situation through properly first Lol. E.G. putting too much power through my PSU and burning it out along with the MB Grrrr. Yeah give me a pointy hat with DUNCE written on it Lol.

Sadly CUDA has only one speed. MAX according to the graphics card you have installed. OK yes they shrub a lot of units and build a lot of credits for credit ho'es. However, if you want some control over what you do ... ... ...

Coming from an old school I like the fact that I have control over my puter and can tweak my CPU to do more or less according to my wishes.

Currently I have a Phenom 1100T 6 core black edition rated at 3100mhz running at 3629.98mhz per core on this project. Over the next 28 days I am expecting to shrub up about 200k credits for this project.

Oh by the way I am still wanting to build a server with 48 Opteron 2100mhz cores to run this project.

When built I will see how much it can be tweaked to maximise output
Greg

[mxplm]

Sadly CUDA has only one speed. MAX according to the graphics card you have installed.

I think NVidia and ATI are working on some kind of acpi/speedstep for GPUs. I remember some kind of cooperation with Berkely University, so BOINC will probably be able to use such a new API shortly after it's available.

[Michael H.W. Weber]

...guys, you are a bit off topic with this in the RNA World subforum, I think. If you like to discuss general topics concerning CUDA, please feel free to open a thread elswhere in this forum:

Michael.