*Official Alienware M18x AMD CrossFire Discussion Thread (re: Dell/AW Conference Call)*

Yea you have the apollo cards as well. Are you having fan issues?
Edit: nvm I just saw your other post. I'm at a loss here... Dell's gonna have to figure this one out. My wife's M17xR3 has the same vendor as well and hers has no issues. I'd still like to see some more screenshots from people that are having fan issues. I'm curious if these are all from the same vendor and batch or not. That should help narrow things down a little more. For all we know, we could have bad motherboards.

My new cards and fans will be here tomorrow afternoon so I'll make sure and report back with the results.

 

Who knows, it could be a motherboard problem. I have two other M18x's here that are due to be sent back tomorrow but I could swap in the 6990s in them to see if the problem persists. That would help narrow it down by a lot.

 

This has been an interesting discussion. I'm showing my ignorance here, but it seems to me that if the problem was with the chips themselves, the overheating issue would be consistent. When I talked with tech, I was advised that the UEFI handled the gpu fans. I read earlier in this post that the VBIOS can be edited to change fan thresholds. I probably misunderstood the response from tech.

So maybe bad VBIOS? What about faulty temp sensors? Does the VBIOS talk to an onboard fan controller to adjust speeds? Maybe fan controller is faulty? When my machine was overheating, GPU-Z indicated fans at 30% just before thermal shutdown. I couldn't hear any fan activity just before the machine shutdown.

I ordered my machine on 9/26. That seems to be consistent with others with the problem. I have seen so many of my symptoms in the posts from fellow members.

I cleaned my hard drive(s), installed my personal Win 7 with stock drivers except: Intel chipset, Bigfoot, and 11.9 from AMD. I have not had a thermal shutdown since. Previous to that, I shutdown a number of times watching a video on YouTube. I haven't loaded up the machine since the problem started in fear of melting something crucial in the machine. I'll probably be back tomorrow to tell you that I overheated after a cold boot in the morning. So I'll say it now

I have a replacement machine on the way. It is taking a bit longer than one would expect for it to get out of production. So perhaps the delay for the machine replacement ties into the delay in getting cards replaced by tech support.

I take my hat off to Dell for taking good care of their customers. Also props to Mr. Fox for reaching out to Dell. I think though, that if there is an issue with the graphics system, they should be a little more forthcoming about it; without having to be prodded by this forum. I realize that a company is not going to wash its laundry in public, but judging from the number of posts, it is not an isolated, random problem.

 

Lets wait the Dell's fix on the issue but I definitely don't want my 1 week old machine taken apart

 

Alienware tech support is unaware of any problem. I just called in, described my problem. Tech remotely accessed my system, ran msinfo32, bench-marked with Heaven DX11 and monitored temps. He noticed my system bios was on A03 and updated to A04.

The thermal shutdown did not occur during his call, it had occurred 15 minutes b4. And I since rebooted then called Dell, all was fine during call. LOL, the tech thinks the bios update will fix problem but remarked if problem happens again, I should call during an extended off period, to replicate condition.

He also noted that the problem is not in their trouble database.

We need to make dell aware of this issue in every way possible. (call in guys) Doesn't seem like the dell community rep here in this forum has informed the rest of Dell. And if you happen to reboot b4 calling Dell they won't see an issue.

 

Thats the entire reason for this thread. Thats why were are putting together an agenda. We're going to be on a conferance call with the sole Alienware graphics engineer. Things are not going to get done right now! They will be looked into though. We're going straight to THE man who can help. Louis.

Tech support reps only know what thier script tells them. They know the basics. Kinda like geek squad agents. I highly doubt that any issue they run into gets passed along up the chain.

 

Again, there is 'something' called "Post-Sell Service" and here, again, we the end users, experiment the lack (& assurance] of a "Total [& transparent] Quality Control"...

 

At least 3. I have an older set of 6970M's and the vendor is Dell. Also the Clevo's. Neither have issues in my M18x.

 

oh...they SEE the issue...the problem is IF it can be easily fixed. that's the problem...then it has to go through testing or get ridiculed like things are now.

 

Indeed if Bill says they are working on a solution, they are working on a solution. Everything he has said and done to represent himself and Dell here has been completely open and honest as far as anyone can tell.

John is right, if they rush a fix and it doesn't work people will be doubly upset and claim its Doubly Dell's fault, even if they are just trying to make people who are demanding a fix 'now' happy.

 

this is my bios pic:

http://i957.photobucket.com/albums/ae55/FredFlint/bios_pic.png

 

Agree: better a 'slow' (but secure) fix than a premature one... That's the reason I have so many sustainable hopes in the so far proven seriousness of Dell (and in the M18x2... ) [ besides the still apparently lack of a Transparent Total Quality Control]

 

I'm going to put the 6990a in another M18x to see ifI can narrow down the cause. If it turns out to be my motherboard I'll just swap them out.

 

My rig built end of july I 7 2720 xfire 6990's. No problems so far with overheating even when playing crysis 2 full maxed out. Was ordered from the Netherlands. Following this thread with great interest!

 

Lets push this back to the topic - specifically, if you are having issues with your AMD configuration or have questions relating to performance with your Dell AMD cards, post away.

Please refrain from spinning off into other areas as this will make Fox's job of collecting info more difficult - he already has 12 pages to review .

Thanks everyone.

 

DISREGARD THIS FOX

Can you post a video/thread about how much paste you use? I've seen everywhere from a LOT to very little used but I can't argue much if you're slashing temps 20 degrees O.O

 

My cards have Dell IDs.

 

An update on my laptop, it has been shipped back from the depot and it is currently on its way, it seems the motherboard was replaced as noted by the list of components that were repaired/replaced, which consisted of:

C9XMR - Board

So, we will see if this is indeed a motherboard issue, at least in my case anyways, but I have seen that it could possibly be a motherboard issue in this thread.

 

6990s are in the other M18x I have on hand. I'm typing from it right now, I'm about to run some furmark tests to see if its a motherboard issue and answer some lingering questions.


Test 1:
HDD used: The original one with Dell software and catalyst drivers.
Result: Fans come on during cold boot!

Test 2:
HDD used: my SSD with Catalyst 11.10 v3 + Cap 4
Result: Fans come on during cold boot!


Test 3: Final test
HDD used: SSD + Catalyst 11.10 v3 + Cap 4
Result: NO COLD BOOT BUG!


It's definitely the motherboard that is defective. So those of you that have the cold boot bug will need to get your mainboards replaced.

 

One update on the issue on my side, may be relevant to the topic. If I leave my notebook powered down for more than a few hours sometimes it doesn't boot when it is connected to adaptor for the first try. When I press the power button, keyboard lights up and screen back-light turns on, a brief fan spin-up then it cuts power. I have to press the power button again and than it powers-up okay.

It didn't power up at all when it was hooked up to the adaptor right after the delivery (had to boot on battery then connect the adaptor) but I haven't had this bad again after the 1. day.

 

nice testing joker !

 

I am having my 1 week old m18x sent back to dell and they are sending me another brand new one to replace it. So when I receive that I will let you guys know if the new systems being shipped out this month have the problem or not.

The tech guy I have been dealing with says he will continue to send me new systems until I get one that works out of the box. Me and him have dealt with replacing parts in my old m17x for the past 3 months and thats the last thing we want to do lol. In the meantime I will keep getting new systems to see if they work and hold onto my m17x until then

 

Not to be too picky, but I wonder what would happen if you swapped the cards back into the original mobo, Joker. Maybe reseating for the swap did something?

 

I put them back in the original and it has the cold boot bug as expected.

 

Nice testing, it makes sense, since the Compal EC is on the motherboard.

 

What is compal EC? Can it cause cold boot issues? can it be flashed/upgraded/fixed by us or Dell only?

 

Well this is good and bad news. I'm STILL waiting on my replacement cards and fans after over 2 weeks... my warranty has since expired, and I'll need a new motherboard to fix the problem. I've got my hammer handy in case Dell decides to shaft me on this one.

 

so is there a bad batch of motherboards since not every1 has this issue? or maybe different types of motherboards that AW uses?

 

I would say that's a good guess. My motherboard is defective, no doubt about it in my mind. I plan to have AW send me one and I'll install it myself--I don't trust the techs around here, they can spectate while I do it.

 

Tonight an Alienware tech has ordered a replacement motherboard and the Crossfire 6990s to be put in next Monday at my house.
I know who the techs are now, because there are only two of them, and I have had them stop by my home/work many many times.

Anyhow...My question is....Should I take the new/refurbished motherboard and GPUs, or should I sit it out and wait for..... for something or other?
I'm not sure if this M18x has new or refurbished parts in it due to that it is a replacement of my old M18x.

How can one tell if something is new or refurbished? I once learned how to read the manufacturing date on a circuit from a circuit board design course I took in the early 90's. Just looking at a RAM chip two seconds ago, I am not able to find what I used to find to know that manufacturing date.

So...should I take the new boards, or wait for something that resolves the no fan/shut down issue because I do not want refurbished parts? And how do you know if something is refurbished or not?


Thanks a bunch ahead of time for anyone who answers my questions.


Best,
Mike

 

Honestly, a refurbished mainboard is just as good as a new one and likely is new anyway. They just have to mark it refurbished by law from what I've been told.


Edit: For those that have the cold boot bug, remove your bottom panel and battery and look at the motherboard near there. You will see a label that has the revision of the mainboard, mine is as follows: Revision G2 1142. I'm betting that those of you with the same problem have a Revision G. The working motherboard is a Revision E (if I remember correctly).

 

My replacement M18x arrived on 9/27. It had intermittent issues from day 1 with crashing to the desktop during gameplay (Crysis, Left4Dead 2, Bulletstorm, Modern Warfare 2, etc.). Some days the system seemed to function normally and other days it would not. The problem, as it first started, was that my screen would sometimes go dim and I would get a series of system tray pop-up messages stating "Display driver amdkmdap stopped responding and has successfully recovered." I would have to reboot to regain full screen brightness. After rebooting, Crossfire would often be disabled and greyed-out in CCC. Upon discovering this, I found that launching CCC and disabling/re-enabling Crossfire would often restore screen brightness and temporarily stop the error messages short of rebooting. This was occurring with Dell video drivers to begin with. I tried 11.9 and 11.10 preview and the problem never changed. This never happened when running on the Intel IGP and I initially assumed it was an AMD driver issue.

During the first weekend of October, I did a clean Windows installation and tried swapping RAM modules to see if there was another cause. The problem was no different.

Around October 16, the screen dimming and the "Display driver amdkmdap stopped responding and has successfully recovered" error message happened numerous times and I could not get the problem to stop by rebooting. I got frustrated with that situation and switched to the IGP.

The next morning everything was fine when I first booted on the IGP. I later switched to the AMD discreet graphics and attempted to reboot. Instead of the system restarting, it started giving 8 long beeps and would not boot. There was no BIOS splash screen. I removed the AMD GPU modules and my system would boot again. I have also tried to install the 6990 modules individually in the M18x and in my M17x R2 and it will not POST with either of the cards installed in either system.

Like a number of other M18x owners, I have been waiting for replacement parts for an extended time and running my M18x on the HD 3000. I have not been given an estimated date that replacement parts will become available.

 

I wish in next gen systems would be a premium thermal paste job, the quality of from factory paste can be described as "poor", because GPU temp easily goes above 90C in all games and above 100C in Metro2033... (at temps near 100 stuttering can be seen)

I ran Furmark like 5150Joker did on the previous page, and of course I had to interrupt it.. I read people could run it qute long after repaste




Uploaded with ImageShack.us

 

Just checked my fans & temps from cold boot. (first boot of the day after being off all night).

I switched on my external and "disabled" the internal display (Fn&F6 projector only).

Absolutely no fan activity upon switching on. Temps on the primary rose to 65c before fans kicked in to cool them back down. It will now cycle like this if I left it. The only thing that makes a difference is when i put the cooler fans on full (they are quiet so that doesnt bother me).

The temps then will idle at around 50c and will NOT increase up to 65c while left idling - in effect meaning my gpu fans arent getting used at all at this point and the cooler is doing the work. The shot below shows idle for 15mins with just the U3 doing the cooling, no system fans at all.....

Dunno if thats good or bad really....part of me thinks not as without a cooler and just on a flat surface, temps would surely cycle up to 65c often and in my opinion, thats too high before any fan activity icks in....dunno what you guys think?

 

Thanks for testing. It would seem your ahead of Dell on identifying the cause of the problem.

Could people having and not having the problem post their board rev. It's easy to find and it will help us get to bottom of this. I'm having the issue.

Here's my mainboard revision:
27617440 00917 43194731L03 REV: F2 1139

 

Just repeated without the external and temps are the same - no system fans, just my cooler doing the work. Disconnect the cooler and the temps rise and fans kick at 65c.

 

Steviejones if I remember correctly your not having this thermal shutdown bug. Mind posting your mainboard rev, PLZ

 

Have problem with fan too (after cold reboot) and my mainboard revision is REV F2 1139.

 

I do NOT have teh thermal shutdown issue ....

Rev F2 1123

 

You ended up getting, what appears to be one of the best 6990 systems out there. Any plans to go 580M?

 

lol the stuff John has been cranking out certainly is tempting, but then thats not really for every day use, its just the very limits to the system.

I still need to push this fully with the new processor in it, John and I had my system running about a 29K vantage with only about an hour and a bit of work and with only the es processor in it. Quite sure it wouldn't match his system, but it was in the neighboorhood at least of those upper tiers ... I still want to break 30k though and a full 9K 3dmark11 (I should have done this in Sept when the school year was slow, we are now getting more and more busy) ...

Short answer though, no.

I was also considering overvolting my cards but I don't want to risk what I have going ... so I probably wont even do that. I think I am ordering another though (another 6990), so if I do, I may play around with it a little more. I just can't see the reason to spend the extra cash, even if I have it.

 

When I get off work today, I will go home and pull the cover off and see what MB I have. I am having the issues so we will see if I have Revision G.

 

so far

torgars has REV F2 1139 <-- has problem
nuroo has REV F2 1139 <-- has problem
5150Joker has REV G2 1142 <-- has problem


5150Joker(2) has REV E? ???? <-- no coldboot thermal shut down.
Shaden has Rev F2 1123 <-- no coldboot thermal shut down.

 

update 1

torgars has REV F2 1139 <-- has problem
nuroo has REV F2 1139 <-- has problem
5150Joker has REV G2 1142 <-- has problem
mert361 has REV F2 1139 <-- has problem


5150Joker(2) has REV E? ???? <-- no coldboot thermal shut down.
Shaden has Rev F2 1123 <-- no coldboot thermal shut down.

 

Seems I am not having any temp. issues at idle on Dell stock pastes etc with A04 and Dell vBios. It is noticeable that I am having problems under load. I would like to see (apart from everyones problems sorted) a bios revision to alter the temps at which the fans kick in. Lower than 65c would be nice.

I shall be "opening her up" again within the next day or so after I have recieved my replacement memory modules from Kingston (can you beleive both of my x2 8gb kits had errors on 'em! - RMA'ing both kits and getting a nice bump to the 1866 stuff with a free 8gb flash drive too so Im happy with the outcome).

Once it arrives, I will install and get my mainboard revision info and post back with it.

 

Rev. F2 1136 here with the problem.

 

hummm, my board is an f2 1139.....this is interesting.

 

update 4

Nicholaus.rossi has Rev. F2 1136 <-- has problem
torgars has Rev. F2 1139 <-- has problem
nuroo has Rev. F2 1139 <-- has problem
mert361 has Rev. F2 1139 <-- has problem
djskorch has Rev. F2 1139 <-- has problem
5150Joker has Rev. G2 1142 <-- has problem


5150Joker(2) has Rev. E? ???? <-- no coldboot thermal shut down.
Aikimox has has Rev. E2 1116 <-- no coldboot thermal shut down.
Shaden has Rev. F2 1123 <-- no coldboot thermal shut down.

 

OK, I will jump on the band wagon with a Rev. F2 1139 and I DO have the problem.

 

E2 1116 here - NO thermal shut down problem. Mine is most likely one of the first revisions.