Weird: 3x 6990 but Phoenix mines only one, distributing the load

[Forp]

Hi there.

Proudly upgraded my 2x 6990 / Windows to a 3x 6990 / Debian. :-)

Now I get a weird situation: I am getting the same performance as with ONE 6990. :-(

This is catalyst 11.7, SDK 2.4 on a fresh Debian. clinfo, atconfig and flrxinfo see all 6 GPUs. Phoenix / poclbm see all 6 devices. I started 6 miners for 6 different pool workers and I get...a hash rate of 590 MH/s.

All 3 cards get warm, atitweak reports that all 6 GPUs are working, albeit at a utilization of 30%. If I start additional miners, situation stays the same.

When I start the miners, the first two miners run at 300 MH/s and on 2 GPUs. Adding miners (to the remaining 4 free GPUs) has them run slower and slower, evenly distributing the load on all 6 GPUs but never reaching more performance than I usually get on one 6990 card.

It's probably not the motherboard or the CPU, since on Windows the TWO 6990 nicely produced 1,2 GH/s.

After 1 day tweaking and twisting ... I am a bit desperate ... and could use some help ... please ...

[1714004747]

1714004747

[1714004747]

1714004747

[1714004747]

1714004747

[1714004747]

1714004747

[tysat]

Try out Catalyst 11.6?  I heard of other people having issues with 11.7.

[Jack of Diamonds]

How much juice does your power supply have?

3x6990's in switch position 2 will take ~1500 watts, that's before the CPU, motherboard, harddrive, ram, possible fans

[Forp]

How much juice does your power supply have?

3x6990's in switch position 2 will take ~1500 watts, that's before the CPU, motherboard, harddrive, ram, possible fans

That's plural. TWO power supplies adding up to 1650 watts. So it should be fine.

However, I got the rig running now under Windows 7. Looks like the 11.8 preview driver is indeed able to accommodate 6 GPUs (the one I used before did not - windows complained about the driver of the third 6990 having insufficient ressources). And I get 1600 MH/s and am quite happy :-)

Also it looks like 11.8 preview leads to a thermal improvement. I am running at 5 - 8 deg celsius less with the same hash rate per GPU, if compared to earlier drivers.

The sad side, of course, is, that something I did not achieve under Linux now is working fine under Micro$oft. To me, this is rather irritating 

[CanaryInTheMine]

I assume under linux, each instance of your miner pointed to the correct GPU number?

if you are getting 1600 from 3 6990s, it's a little low there at 266 Mhashes per GPU...  you should be getting 400 Mhashes per GPU with very modest overclocking.

[Forp]

I assume under linux, each instance of your miner pointed to the correct GPU number?

How do I do that correctly?

What I did was set DISPLAY to :0, turn off xfire and then set DEVICE=0 (1, 2, 3, 4, 5).

Is that the way to do it?

Because...it sounds like a quite plausible explanation that I was only using ONE card. And if I addressed them with atitweak incorrectly as well...I also might be getting the funny performance figures.

if you are getting 1600 from 3 6990s, it's a little low there at 266 Mhashes per GPU...  you should be getting 400 Mhashes per GPU with very modest overclocking.

Yes. But that's fine for me. I am still grilling my steak on the stove and not on the GPU :-)

I am mildly underclocking the cards, a bit intentional. Moreover, it is summer here and when I did some modest overlocking the cards locked up on me.