Jump to content
EwinRacing Flash Series Gaming Chairs
Brian

Require assistance from those that own M18x - GPU 2 temp issues

Recommended Posts

Could someone with 6970M Xfire run furmark multiGPU 1.82 (with postfx, displacement mapping and xtreme burn mode selected) for at least 10 minutes and have two instances of GPU-Z open and report back what the differences in temperatures are between GPU 1 and 2? My GPU 2 is reporting 5-9C higher temps despite having just repasted both GPUs. In fact, I went back in and opened the system a second time and repasted GPU 2 and noticed the same behavior. After leaving the system half assembled, I physically pressed down on GPU 2's heatsink while furmark + GPUZ were running and noticed the temps for GPU 2 drop and fall in line with GPU 1. Furthermore, the idle temps of GPU 2 are also quite a bit higher than GPU 1.

This can mean one of two things:

1. My system is flawed and needs a replacement.

2. Design of GPU 2 sink is flawed and needs a retention mod to correct it.

This behavior was observed before repasting and after repasting.

Share this post


Link to post
Share on other sites

Thanks justice. JT and others that got yours, could ya test it out and let me know? Thanks

Share this post


Link to post
Share on other sites

Did you see any real difference in temps before and after repasting?

Are both GPU fans running at the same speed?

Share this post


Link to post
Share on other sites

Both GPUs are running cooler after repasting. I went ahead and a retention mod on GPU 2 by using the plastic washers + c clips I pulled from an old M17x-R2 HDD caddy I had and GPU 2 still hovers 5C hotter than GPU 1. When I press on GPU 2 now, it doesn't drop the temperature so it means the retention mod is doing its job. I guess the heatsink design itself seems to be inefficient for GPU 2. I still would like confirmation from others that have the M18x.

Share this post


Link to post
Share on other sites

Currently returning my m18x, after about 10min furmark on stock speeds , gpu2 goes to 90C with its sub temps going to above 100C. To hard to mess with the GPU's repasting and such + don't really want to bother fiddling with a brand new hardware. Going to re-order when new GPU's are released hopefully soon, maybe then engineering will address this.

Edited by littleone562

Share this post


Link to post
Share on other sites
Currently returning my m18x, after about 10min furmark on stock speeds , gpu2 goes to 90C with its sub temps going to above 100C. To hard to mess with the GPU's repasting and such + don't really want to bother fiddling with a brand new hardware. Going to re-order when new GPU's are released hopefully soon, maybe then engineering will address this.

How long did you let it run for?

Share this post


Link to post
Share on other sites
hi Brian question, how is this different from what I posted in this thread about my m17x-R1?/ link http://forum.techinferno.com/hwinfo32-discussion/319-normal.html That system is running a 40 degree difference between GPU's and answer i got from TI membership is that is normal.

Your logs were in Farenheit not Celsius. I only saw Marc briefly answer your question. At best, there should normally only be 2-5C difference between GPU's when under load.

Share this post


Link to post
Share on other sites

Are both GPUs 100% loaded during the test?

What does HWiNFO32 say? Post a screenshot (also with fan speeds), use latest build with M18x support.

You can also open graphs for both GPUs to see how the load/temperatures behaves during the test.

If both are symmetricaly loaded, but temps are different it's most probably a heatsink issue.

Share this post


Link to post
Share on other sites
Are both GPUs 100% loaded during the test?

What does HWiNFO32 say? Post a screenshot (also with fan speeds), use latest build with M18x support.

You can also open graphs for both GPUs to see how the load/temperatures behaves during the test.

If both are symmetricaly loaded, but temps are different it's most probably a heatsink issue.

Hi Martin, this is what I'm seeing with the system half open and furmark running at full blast (all options selected with 1280x1024). This is after adding new paste to both GPU's and doing a retention mod to GPU 2 (i added a c clip + plastic washer to GPU 2 to increase tension on the core).

post-5-14494988003282_thumb.jpg

Share this post


Link to post
Share on other sites

Hm, both ~99% loaded and same clocks.. IMO a cooling issue - heatsink, thermal dissipation or a GPU chip with bad thermal characteristics..

Share this post


Link to post
Share on other sites

Even in celcius Brian the difference between GPU's is close to 17 degrees, so it is not a localize problem to the m18x, it seems to be the norm. I will run Medal of honor for 10 min and then post the hwinfo numbers.

Share this post


Link to post
Share on other sites

I spoke with Alienware tech support for 2 hours on the phone without a proper resolution. I asked for a system exchange and was told that it is normal GPU behavior for GPU 2 to run 15C higher than GPU 1. Their engineer claims that Crossfire works by putting the full load on GPU 1 and only then shifts it to GPU 2 when needed (this goes contrary to the way AFR works). I asked then why does GPU 2 run so hot if that is the case? They couldn't provide an adequate answer. Instead they told me I could choose the parts replacement route or ask for a refund. When I opted for a refund, i was told to call back and speak to teh proper department and that I would not be guaranteed a waiver of the restocking fee. This was probably the poorest example of Dell customer service I have had to deal with in a long time and I will be including it in our up coming article. Stay tuned.

Share this post


Link to post
Share on other sites

Yes the graphs are very accurate and in realtime it was perfect for adjusting the fan control respin setting. WBabt007 I'm sorry for the brief answer I did believe there would be more input from the forum members. Thank you Brian for the proper answer. So his R1 needs a repasting and an extensive tension mod?

Edit: I'm sorry to hear that Bryan. I was told there systems been down two days ago and they couldn't credit my account at the time and sent a expedite request that should be done in 24 to 48 hours. I've been involved with my case for more than a month and a half and still haven't placed the order yet. Long waits on the phone endless processes to go through to get anything done. It's been a lousy treatment for my investment.

Edited by mw86
  • Thumbs Up 1

Share this post


Link to post
Share on other sites

don't call tech support when your system is within the 21 days, call customer care and all you need to tell them is your not happy with the laptop.

  • Thumbs Up 1

Share this post


Link to post
Share on other sites

Brian, according to that screenshot the fan on GPU2 is running at much lower speed than GPU1. Can you audibly check the fans when loading to see if GPU 1 is running noticably higher than GPU2?

If that is the case, could be a dying fan. If the fan is OK, then it is a bios issue and needs an update.

What I would do is try run a few wires from GPU2 fan to GPU1 fan header to test if the fan runs at the same speed, or just swap the fans around. I would say thats your issue.

-Ash

Share this post


Link to post
Share on other sites

Nice catch Ash those are very possible. Maybe all of them are suffering from this and need the gpu fan replaced and or need fan control updated.

Share this post


Link to post
Share on other sites

How do you know that GPU2 fan speed is much slower, do you judge based on the GPU-Z reported fan speed (or HWiNFO32 fan speed under the GPU)? If yes, then forget this reading - it's invalid, because the GPU doesn't control the fan here. Just like on other Alienware models I had to implement the model-specific fan speed reading via EC, so use only those values for fan speeds. Unfortunately for M18x I don't know how to read both fan speeds yet. I can read 2 fans only now: CPU and GPU (probably GPU1). I'm not sure if Compal offers the GPU2 fan speed read, I could try something - make a special build to test if I can get the GPU2 fan speed too. Let me know if you want to give it a try.

Share this post


Link to post
Share on other sites

Yeah, I figured as much Mumak, that's why I asked him to audibly check, because if he has done a retention mod the only problem left would be a dodgy fan.

-Ash

Share this post


Link to post
Share on other sites

The fan is okay because it eventually kicked into high after I ran furmark at 1080p. Mumak's program also puts it into full blast so the problem seems to be something else.

Share this post


Link to post
Share on other sites

Have you tried changing crossfire drivers? It could be a crossfire volt control issue or something?

-Ash

Share this post


Link to post
Share on other sites

ok having a hard time locating furmark 1.8 multi gpu addition, I DL 1.6.5 and it seems to work with both GPU's

temp never got over 80c, and going to try again with GPU Z logging the temps.

Share this post


Link to post
Share on other sites

Hmm GPU 2 is sitting at idle at 64c and gpu 1 49 degrees idle. hmm ill do more testing once i figure out how to use GPU-Z effectively.

Share this post


Link to post
Share on other sites

okay i ran it for 10 mins and gpu 1 sat at about 77 degrees stable and gpu bounced between 79 and 80 seems to be running fine for me

3xO42.jpg

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use. We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.