|
January 27, 2014, 12:57:01 AM |
|
I finally got off my ass and started mining a little more then a month back with my 7950. Using GUI miner, I found the best thread-concurrency for me was 13,000. This was independent of intensity or number of gpu threads. At night i would run 13000 with an intensity of 18 and get 600 h @ 80 C. (increasing the intensity to 20 only gave me 12 more hashes) While using my computer id run it at 13000 2 threads and an intensity of 12. I had no issues. Then i got a 280x ( i have a 1000w psu). I found out GUI miner can't run 2 cards at the same time. When I tried, both hash rates would run low and fluctuate strangely. I never let it run for more than a few minutes. Running the 280 on it's own in GUIminer i found the best thread-concurrency was 40,000 and i would run it at an intensity of 20 24/7 ( i was getting 660 h). I would have temps ranging from 78-82. I didn't do much testing on my numbers because it's due to go into it's own private rig soon anyway.
Then i started using cgminer so i could run both cards. I found the numbers (thread-concurrency, intensity) i had to use to even get it up and running were completely different. I finaly got both cards hashing at a good steady rate (at and below 80c) and a few hours after that my screen crashed. nothing but green and blue vertical lines. I was still able to carry a conversation on Skype with the responsless screen. I think i fried my 7950...
I pulled it out and put the 280x in my primary x16 slot and fired up GUI miner. Now the optimum thread-concurrency for the 280 is 13000. The same as it was for the 7950... I don't know what to make of this and im afraid to start mining again because i don't want to fry my other card. Does any know what is going on?
Here is my config file for cgminer. This is what i was running when the 7950 fried.
"scrypt" : true, “kernel” : “scrypt”, "intensity" : "13, 20", "worksize" : "256, 256", "gpu-threads" : "2,1", "gpu-engine" : "1000, 1100", "gpu-memclock" : "1250, 1500", "thread-concurrency" : "10000, 40000", “gpu-fan” : “40-100", “temp-cutoff” : “90,90", “temp-overheat” : “82,82", “temp-target” : “80,80",
|