5 Gpu rig problem
-
-
and ad more memory?? ;)
-
I know it holds 5 cards, but does it support 5?
My board holds 7 cards but will only support 4 at any one time.The memory is loads.
-
Thanks all for help. My psu is coolermaster 1500w, don’t remember exact model as I’m in bed already. Mobo is seing 5 GPU’S, and I can customise them via Trixx. But problem Is only when cgminer kicks in. Will try add rams tomorrow.
-
cgminer 3.1 is the version I prefer, 4gb of ram is loads.
-
Thanks, will try it. :)
-
think i have mentioned this before,
5gpu’s isnt a great idea. especially not in windows, and especially not if you want 24/7 mining.used to have it before but stopped due to random cgminer crashes etc.
-
This is bad … 24h later I’m still in same poopoo :'(
Have tested many amd catalyst versions, SDK versions, have turned on and off some setting in bios, downloaded 10 :d cg mienr versions ( well 5 more likely), added extra rams, moved the last gpu that was on board to powered riser, shouted on my three guinea pigs ( they hate me ) and NOTHING SEEM TO WORK ( at least caps working :D )Still same crap : unplug, plug back gpus = 5 visible, start cgminer with .exe without any .conf or .bat = working ( but you know how good hash is at default). S, W = fresh .conf file . Adjusting to “normal” values = bang, and either freeze or cgminer off. And 1 of Gpu’s also - not visible anymore …
Would go with 4 Gpu’s option BUT, have 2 extra Vapor x left and cannot build 1 more rig (even when 1 more Silverstone PSU is otw) because its sold out, and not sure do I want another headache fro running 2x290 and 2x7970 :((( Damn it!
So now managed to get 5gpu’s back, started cgminer with lower then x4 gpu’s settings and got : Amd driver stop responding - and pc wnet frozen bananas …
Weird, not the 5th gpus not disapearing, but I cannot start cgminer, as it freezes it self, I trying to enter pool addreess - and getting that : amd driver stop responding …
Well finally, after 5 or so crashes with cgminer 5th gpu has disappeared. 4gpus running at 750 hash wash… Will blame Mobo and give up. Thanks for trying to help. -
Doubt it, as 4x R9 280x Vapor x rig is using 990w-1010w according to my watt meter. GPU’s lightly underclocked with Sapphire Trixx to 1087v.
Will see how Silverstone 1500w PSU behave when it arrive. -
Solved!
After morning meeting with my tech. department girls, we came to conclusion that cable was no good ! (look first picture)
So what we done is :
Reinstalled windows 8.1, disabled all protection shit they offering on, by default, turned off updates and so on …
Catalyst 13.9, latest ADM SDK, Cgminer, Cgwatcher, TrixxAnd since my missing Gpu happens to be in PCI-E 1 slot ,so I did research about it. Eventually there is solution to force windows to “monitor” that slot constantly, which can be achieved by “short circuit the slot”, which works as many useres confirms. But however it scares me to burn down da house, with such experiments :)
So all it came up with adding non powered riser to PCI-E 1 slots instead of powered that was there since rig beginning. And now all seems work as charm, apart that 1 gpu is Little slower than the rest four. But honestly - I don’t care atm :DP.S. 5x Sapphire R9 280x Vapor x uses 1336W, that’s according to my socket meter readings. Its under volted by Trixx to 1112v. I believe it will take a Little less in the future after I receive more non powered risers and increase the distance between cards, and reduce heat (so fan’s will draw less).
Thanks again for all your help guys .
[attachment deleted by admin]