Topic: Problem with new mining/minting rig: Biostar TP67XE motherboard (Read 1372 times)

Ok, great.

I agree. Since then I removed the 5970 and has been running the board with 2x 5850s and having no problems. I still didn't hear back from Biostar so I will install windows to do the bios upgrade and try again.

Ouch, that sounds scary and laborious looking at the number of screws on the 5970, but I may give it a shot if I cannot find another motherboard that can handle it. I actually previously got it to work on one of my other computers that has the Asus P5NSLI motherboard with the nForce 570 chipset. The 5970 was running stable around 82C degrees. The Biostar has the 590 chipset, but I get crashes sometimes immediately when the temperatures are not even 65C!

No, I don't have the crossfire cable installed as they are incompatible GPUs. I also disabled the internal crossfire on the 5970. I learned my lesson the last time I got that 5970 to work (which was painful, too)

Thanks for the suggestions..

The Coolermaster PSU should be sufficient.

I have a feeling it's either a cooling issue with the 5970 or the Biostar board doesn't enjoy the 5970 and 5850 together. It's worth looking for a bios update for the motherboard.

If you're feeling brave, you can always remove the 5970, take off the cooler, wipe off the existing thermal tape / paste. Put on a layer of arctic silver 5. Make sure the cooler makes close contact with both GPUs. This should bring temps down a couple of degrees over time.

As for the 5970 + 5850, are you running crossfire on them? This may be an issue if that's the case.

Thanks for the quick answer, zerokwel!

Ok, I did some of those before, but this time I followed your advice and tried all combinations. The 5850 by itself works fine. Even the 5970 works if I only use one of its GPUs. The moment I start the second one it crashes. Sometimes I still get crashes even with the first GPU.


Yes, the memory were in the red slots and as you suggested, I tried using either stick, but no difference.

I also changed the software: I downgraded to SDK 2.1 and updated pyopencl and phoenix, etc. Using SDK 2.1 seemed to improve the stability, but I still cannot completely prevent the crashes.

Next, I will try replacing the 5970 with another 5850. And try a different PSU.

I forgot to mention one thing about the motherboard: I got it as an "open box" item from newegg. They still have a few more of those, so it's not one person returning them I guess. That may be the main reason for my problems There are two issues that I still couldn't resolve from the Biostar website:
1) The bios is not the most up-to-date one, but the update instructions in the manual doesn't seem to work for this mobo: pressing F12 during POST messages doesn't do anything.
2) The mobo has a 10-phase power display, and one of the leds is off. I have no clue what this means. It's not explained anywhere. It's some crazy overclocking debug signal provided by Biostar.
I contacted the manufacturer about these, but didn't get any response yet.

Best thing to do is go step by step.

does it do it when you mine on the 5970 or the 5850. or does it do it on both

Have you tried unplugging one of the cards does it still crash

tried just the other card does it still crash

booting on only 1 stick of ram one stick might be bad and causing the crashes ?. try booting on one then the other.

Just something to get you started.

there are quite a few things that it could be. so its best to check for hardware issues first.

also the memory is installed in the red slots?

Hi, I've been patiently waiting for my new computer parts to arrive, but I was quite dissappointed when it started crashing right away when I start mining on more than one GPU

My system is:
Biostar TP67XE socket 1155 motherboard
Intel i5 2500K CPU
Patriot 2x 2GB RAM
CoolerMaster 1000W PSU

GPUs:
1x Sapphire HD5970
1x Gigabyte HD5850

Software:
64-bit Linuxcoin 0.2a on a persistent USB stick
Phoenix r86 and r100
ATI SDK 2.4

It's supposed to be a dream system, but even when I run phoenix on a single GPU for a few minutes, I get the familiar CPU lockup errors that I know from overclocking GPUs in my other systems. But in this system, it is happening with normal clocks and a cool running (60-70C) cards. So it feels like either a motherboard or PSU problem, unless you can tell me that there is something obviously wrong with my hardware/software setup.

I will test some more different software/hardware scenarios now. Also downloaded Ubuntu Natty Live.

Any suggestions will be greatly appreciated!!!