Message boards :
Number crunching :
Original Titan, scoring poorly, need help
Message board moderation
Author | Message |
---|---|
Send message Joined: 31 Aug 11 Posts: 20 Credit: 528,735,480 RAC: 2,465 |
I have been using Radeon 280X's, and they all can complete one Milkyway@home v1.02 task in less than 24 seconds. I have acquired an original GeForce Titan, and the double precision according to Wikipedia should be at LEAST 30% more than the 280X, but instead of <24 seconds, a single task requires 220-390 seconds, as well as a significant increase in CPU time. Are the opencl_nvidia tasks not as optimized as the opencl_amd_ati tasks? Is there some configuration file I need to optimize? Running multiple MW tasks raises the card's utilization from mid 70% to 99%, but the length of the individual tasks likewise increases (as expected). Anyone have a solution to make a Titan run as it should on MW? |
Send message Joined: 31 Aug 11 Posts: 20 Credit: 528,735,480 RAC: 2,465 |
I've found an option to enable more cores to be used for double precision, but that didn't help. My research so far is that perhaps the massive double precision processing is only available in CUDA format, not in Open_CL tasks. Thoughts? |
Send message Joined: 31 Aug 11 Posts: 20 Credit: 528,735,480 RAC: 2,465 |
Well, perhaps this thread will prove somewhat useful to someone else in the future. More info can be found below: http://forums.anandtech.com/showthread.php?t=2446534 |
Send message Joined: 24 Dec 12 Posts: 6 Credit: 2,720,018,100 RAC: 0 |
I have been using Radeon 280X's, and they all can complete one Milkyway@home v1.02 task in less than 24 seconds. The problem with the TITAN's running MW tasks is that each task will only take up about 14% of the GPU and you will need to install an app_info file telling the computer to run 7 or 8 MW tasks concurrently. It is your decision how many should run concurrently at any time however by doing that you will find the TITAN card can produce competitive results and achieve superior GPU optimization. Try not to engage the double precision ("DP") feature in the NVIDIA control panel as while the does give better DP performance numbers, it also slows down the clock speed and negates the DP benefit for MW tasks. |
©2024 Astroinformatics Group