21)
Message boards :
Number crunching :
OpenCL WU test on OS X 10.11.5
(Message 1789726)
Posted 23 May 2016 by Gianfranco Lizzio Post: What revision you compiled from, BTW? I compiled the SoG version Rev. 3463. Gianfranco |
22)
Message boards :
Number crunching :
OpenCL WU test on OS X 10.11.5
(Message 1789717)
Posted 23 May 2016 by Gianfranco Lizzio Post: MB OpenCL apps continue to return incorrect values ​on OS X 10.11.4 and 10.11.5. I did the test with both SoG version and the classic one, always getting the same result . Then I compiled myself the OpenCL apps with the same result! I tested the apps using the test WU downloaded from Lunatics . But all this does not happen with the CUDA apps and MB CPU apps that returns the correct values ​​during the test. I have kept all the files obtained during the various tests, so i can send the result if necessary. Gianfranco |
23)
Message boards :
Number crunching :
GBT ('guppi') .vlar tasks will be send to GPUs, what you think about this?
(Message 1786756)
Posted 11 May 2016 by Gianfranco Lizzio Post: Assuming you changed maxregisters up to 64 as you stated elsewhere, try dialling it back to 32. I'm using maxrregcount=64 in El Capitan without any problem. Gianfranco |
24)
Message boards :
Number crunching :
GBT ('guppi') .vlar tasks will be send to GPUs, what you think about this?
(Message 1786742)
Posted 11 May 2016 by Gianfranco Lizzio Post: Running 'guppi' 3 at a time on my i7 4770 using avx optimized code i take 42-45 min. Running 1 at time on my GTX 960 using Petri's cuda code i take 17-18 min. This means that i7 with 'guppi' is 20% more efficient than GTX 960 and uses 36% less power! Gianfranco |
25)
Message boards :
Number crunching :
GBT MESSIER031 work on GPU NVIDIA
(Message 1781723)
Posted 23 Apr 2016 by Gianfranco Lizzio Post: The interesting thing of the overflows is that the result are not simple spikes but triplets. |
26)
Message boards :
Number crunching :
GBT MESSIER031 work on GPU NVIDIA
(Message 1781656)
Posted 23 Apr 2016 by Gianfranco Lizzio Post: Good news the Berkeley servers started the distribution of guppi work unit on CUDA NVIDIA GPU. |
27)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1773057)
Posted 21 Mar 2016 by Gianfranco Lizzio Post: Hmmm, so you're using a LapTop with an External video card? No Tom...I'm using a self-built Macintosh with performance comparable to a Mac Pro 2013. |
28)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1773017)
Posted 21 Mar 2016 by Gianfranco Lizzio Post: If you're seeing inconclusives directly as a result of switching maxregcount, then you have Cuda kernels failing silently (and so looking quick) due to launch restrictions in Drivers and hardware..... The most likely explanation seems to be the use of Nvidia Beta Driver. But there is nothing to do because they are the only ones available. |
29)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1772774)
Posted 20 Mar 2016 by Gianfranco Lizzio Post: Increasing the maxrregcount value increase inconclusive results. Now I'm trying to find a good compromise between speed of execution and inconclusive results. |
30)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1772613)
Posted 19 Mar 2016 by Gianfranco Lizzio Post: Jason the same thing happens with your code, placing maxrregcount = 64 and testing with the reference work unit present on Lunatics I register an increase in performance of 3.6%. The performance increase is modest compared with 21% of Petri code, but still present. |
31)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1772600)
Posted 19 Mar 2016 by Gianfranco Lizzio Post: http://setiathome.berkeley.edu/result.php?resultid=4800007576 with maxrregcount=32 http://setiathome.berkeley.edu/result.php?resultid=4800260714 with maxrregcount=64 and the same AR 0.415 Performance increase 19,8%. This result are on my GTX 960 with gencode 35,50 and 52. [/b] |
32)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1772597)
Posted 19 Mar 2016 by Gianfranco Lizzio Post: Okay, I just searched the Xbranch folder from the last build and found; Tom, I recompiled the code with maxrregcount = 64 and the results are very promising. The client seems faster for average AR of about 1 minute, however, the use of the CPU is always close to 100%. I will do other tests and I will let you know. |
33)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1772083)
Posted 17 Mar 2016 by Gianfranco Lizzio Post:
Petri i replaced this part of stock code in yours but the result is the same 100% cpu is using. |
34)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1771505)
Posted 14 Mar 2016 by Gianfranco Lizzio Post: Tom after ./compile ... the problem is in analyzeFuncs.cpp in the client folder. You have to append the follow line code #include <fft8g.h> after #endif // USE_IPP It works successfully for me. |
35)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1761657)
Posted 2 Feb 2016 by Gianfranco Lizzio Post: Anyone using the older versions might want to try the newer versions and see if they are any better. CPU Intel(R) Core(TM) i7-4770K @3.70GHz (running 8 instances of SETI= AVX Build 3352 Vs AVX Build 3366 Build 3352 Run time: 1 h 54 min 52 sec CPU time: 1 h 48 min 7 sec VLAR=0.010316 Build 3366 Run time: 1 h 45 min 40 sec CPU time: 1 h 40 min 10 sec VLAR=0.010306 Build 3366 is 8,6% faster! |
36)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1758271)
Posted 22 Jan 2016 by Gianfranco Lizzio Post: Computer use with CUDA app is much more fluid unlike OpenCL where I experience slowdowns in the animations on the screen. Someone else has noticed these slowdowns using OpenCL? |
37)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1758222)
Posted 22 Jan 2016 by Gianfranco Lizzio Post: TBar the setiathome_x41zi_x86_64-apple-darwin_cuda65 app is 50% more faster compared with the MBv8_8.05r3346_nvidia_ssse3_x86_64-apple-darwin and uses 25% CPU against 33% the OpenCL app. It's a very good result. OK it's the same for me. For high angle range the computing time is the same as the OpenCL app. |
38)
Message boards :
Number crunching :
I've Built a Couple OSX CUDA Apps...
(Message 1758153)
Posted 22 Jan 2016 by Gianfranco Lizzio Post: TBar the setiathome_x41zi_x86_64-apple-darwin_cuda65 app is 50% more faster compared with the MBv8_8.05r3346_nvidia_ssse3_x86_64-apple-darwin and uses 25% CPU against 33% the OpenCL app. It's a very good result. |
39)
Message boards :
Number crunching :
Solarix 10 on x86
(Message 100802)
Posted 18 Apr 2005 by Gianfranco Lizzio Post: How fast is the seti_boinc client running under Solaris 10 (x86 machine) compared with Linux or Windows? |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.