Christian Buchner

I removed a round bracket, instead of adding one. How's that? 2014-04-13 21:05 GMT+02:00 brian [email protected]: > [image: LGTM] http://www.lgtm.in/i/c2z1 > > [image: :+1:] http://www.lgtm.in/u/c2z1 > > — > Reply...

Maybe we can get the best of both worlds and compile for both architectures (compute_10 and sm_20/sm_21) in parallel, so Legacy users can keep using this kernel. does it make...

This could be accomplished by placing sleep intervals of a carefully controlled length into the computation. Unfortunately while CUDA runs a kernel, no other application can access the GPU. It's...

the Windows specific code has an added CTRL-C handler that tries to initiate an orderly shutdown. You could try to install a SIGBRK / SIGINT handler for Linux that does...

the higher the -L , the longer it runs. use -D to see what it's doing... 2014-02-06 14:07 GMT+01:00 zerowind168 [email protected]: > command is follow [image: :+1:] > cudaminer -s...

what version was tested? a tagged release or the current master? 2014-04-11 6:45 GMT+02:00 utahman [email protected]: > 1. Start CUDAminer in auto mode (-l auto) > 2. After 8-10 seconds...

I don't know your GPU model, I don't know which cudaminer builds you are comparing. All I know is that I've removed the legacy kernel support for very old GPUs...

I get that too occasionally. Probably the cudaGetLastError() is not cleared before entering the autotune algorithm, making it terminate early. 2014/1/12 dchokola [email protected] > Got a backtrace. WARPS_PER_BLOCK is 0...

Is the problem gone now? I have made some code changes after running into this problem myself... 2014/1/16 Jamone Kelly [email protected] > Any idea how to fix this yet? >...

Hi, it might not be the full API, but rather a subset. But it's not the first thing I am going to add: https://bitcointalk.org/index.php?topic=167229.msg5436210#msg5436210 2014-03-01 22:49 GMT+01:00 Valcorb [email protected]: >...