Author

Topic: Seemingly random drop in half of my mhash. Any help? (Read 1764 times)

member
Activity: 94
Merit: 10
Yea, the only thing was that i was running one instance of guiminer and even checked my process and apps to make sure. I have done the half and half like a dummy before, but i really have no idea what happened this time :p
full member
Activity: 210
Merit: 100
Not for hire.
Definitely done the half and half thing by accident as well.  Its a very odd thing to troubleshoot and seems insurmountable until the "duh" part hits you in the face...  Glad you fixed it, whatever it was!
sr. member
Activity: 378
Merit: 250
Most mining software will allow you to mine to more than one pool with the same card, at the same time.  Whether or not you meant to, my guess is that was somehow happening, which was why you were see half your hash rate.  Maybe one of those deepbit miners was still running in the background somewhere when you fired off the mtred miner.

Can't speak for deepbit connection problems.  Never used them, never will. :p
member
Activity: 94
Merit: 10
Im not entirely sure what you mean. I did try restarting various times trying just different workers on deepbit, then i added a worker for mtred and it worked. However even now ill get connection errors after a few hours of mining on deepbit and it stops my miner. The mhash issue is no longer a problem though.
sr. member
Activity: 378
Merit: 250
My gut feeling is that the 'switch' wasn't really switching and you were mining to both pools at once.  I've seen that happen before.
member
Activity: 94
Merit: 10
I'm not entirely sure exactly what it was, but switching pools for deepbit to mtred corrected the mhash issue. Now i can switch back to my original deepbit pool and have it work just fine too.
sr. member
Activity: 336
Merit: 250
Don't rush to help anyone  Grin i already figured it out myself.

what was the problem?
member
Activity: 94
Merit: 10
Don't rush to help anyone  Grin i already figured it out myself.
member
Activity: 94
Merit: 10
if it helps i know its not a reporting error on guiminers behalf because the hashrate deepbit has been reporting is the same as guiminers.

Any help would be really appreciated.
member
Activity: 94
Merit: 10
So I had my 5830x2 set-up running smoothly for about 5 months, OC'd and everything at about 520-560mhash and running from guiminer. I never had any trouble overclocking or going between different clocks until last night when i switch my hdd out of my 5830 rig. However now, both cards are ranging between 50mhash and 200mhash as a rare high. At one point guiminer told me both were getting 30,000 ghash, i wish  Cry

I replaced the hdd i had that was running win7 with another hdd that was running ubuntu natty narwhal lts. I did the hdd switch because i've been trying to set up a second rig but ubuntu and the hardware from my backup rig were having a bit of a disagreement.

When i had the ubuntu hdd booted up in my 5830x2 machine everything seemed like it was working fine and i was trying to set up a miner that would actually work on it. I tired the how-to guide here to no avail then tried other methods, ultimately trying out BAMT which seemed like it was working up until it didnt work. BAMT was the only place that i edited anything regarding my gpu; in the config folder i uncommented the lines that set each gpu clock at the very bottom and reset the numbers so that i'd get the same mining performance as before. After some frustration and a few hours of not being able to properly start mining with ubuntu, i switched the hdd back to the original win 7. Thats when i first notied that my cards were not reaching their previous clockrate.

I tried a few things already such as rebooting with just the win7 hdd in, restarting guiminer, changing flags, a whole lot of afterburner tweaking and thats about it. When i changed the clock on afterburner it did show that mhash would drop and raise with the bar, but it each gpu is capped at 160-200ish mhash even when im clocking at 890mhz.

Sorry for a wall of text but thanks ahead of time if you read through it.
Jump to: