I have suspected for a while that the Optimized Cuda Client is slower than
the official one and did this unit earlier
http://setiathome.berkeley.edu/workunit.php?wuid=692713367
Whilst it did finish first
My device is running at 1890mhz (800core 2538 ram)
His is running at 1728mhz
Both are Nvidia G92's
What's everyone's opinion on this
the official one and did this unit earlier
http://setiathome.berkeley.edu/workunit.php?wuid=692713367
Whilst it did finish first
My device is running at 1890mhz (800core 2538 ram)
His is running at 1728mhz
Both are Nvidia G92's
What's everyone's opinion on this
Huh? They are both optimized clients just different version from what I can see. Thats why they are so close in performance.
Not sure, perhaps most of the optimizations have gone into the standard seti client for the old cores (G80/G92 etc) and you only see optimizations in the 200 and 400 cores?