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

The folks at Dell/Alienware have very graciously offered to host a conference call with Louis (lead engineer in Alienware mobile graphics) as they did for M18x owners with NVIDIA 580M SLI to discuss throttling and performance issues.

This thread has been opened for discussion topics relating specifically to Alienware M18x systems equipped with AMD video cards.

In preparation for our conference call, please share both the good and not-so-good aspects of your ownership experience, questions, concerns, comments or words of praise relating to the AMD 6970M or 6990M video cards. Those who have experienced serious or frustrating issues are invited to be open and frank, but please be mindful of the forum rules when posting.

Your posts will help with feedback and I will be asking several members of the NBR Community to assist me with compiling a list of agenda items for discussion with Louis and his team.

In my communications with Dell-Bill_B, we both agree that a conference call is one way to develop the level of customer engagement and feedback that Dell values. As owners of their flagship product line, this adds even greater value to our ownership experience.

We have an opportunity to raise issues, concerns, etc., so let's use it!

New Information and Updates

Update #1 - 11/06/2011: "Team Red" Panel

Spoiler

The "Team Red" Panel includes the following individuals, in no particular order of importance:

• Mr. Fox
• djskorch
• Aikimox
• Nicholaus.rossi
• steviejones133
• skygunner27
• mharidas
• johnksss
• nuroo​

This group, as their schedule permits, will attend the conference call as representatives of all affected forum members. (Note: the call has not been scheduled yet. Information about the call will be provided as an update at a later date.) Note that johnksss and skygunner27 were selected as panel members as they both owned one or more M18x equipped with 6990M CrossFire until very recently. Their technical expertise and experience will add valuable perspective even though their current M18x systems are equipped with 580M SLi.

Feel free to PM any of us (and post on the thread) any questions or suggestions concerning the topics for discussion or questions about the conference call. Remember, our primary focus will be AMD 6900M Series issues. However, we hope to have more to talk about than just the overheating problem that is receiving the most attention at the moment. It can include anything AMD-specific that would be of interest to everyone.

Update #2 - 11/06/2011: Current Issues and Affected Systems

Spoiler

Although speculative at this point, there seems to be a correlation between motherboard revision numbers and thermal shutdown due to AMD 6990M cards overheating and fans not functioning properly. It is noted that there are quite a few revision numbers on machines that were shipped within a very small window of time. (Issues began to surface for users on this forum in the last week of September and continued surfacing through November). To locate your own motherboard revision number, look for a tag under the bottom cover, situated between the RAM slots and battery interface connection. An image is available here for visual reference.

On the affected units, it has been discovered that the problem is severe when running from a cold boot; but, rebooting causes most of the affected systems to behave correctly, but only temporarily. Additionally, entering the BIOS from a cold boot and resetting system defaults produces good results, but likewise, this only corrects the problem until the next system shutdown. It must be repeated at the next cold boot. Below is a list of motherboard revision numbers and system behavior reported by the respective owners. (Post your information in the thread and send a PM to Mr. Fox if you would like to have your information added to the list.)

Systems Currently Having Issues with Fans and Thermal Shutdown:

iPhantomhives - Rev. F2 1129
gderreck - Rev. F2 1131 (Original M18x)
torgars - Rev. F2 1132
Nicholaus.rossi - Rev. F2 1136
un1b4ll - Rev. F2 1136
steviejones133 - Rev. F2 1136
nuroo - Rev. F2 1139
mert361 - Rev. F2 1139
djskorch - Rev. F2 1139
FredFlint - Rev. F2 1139
vanhausse - Rev. F2 1139
rpg1201 - Rev. F2 1139
scottydosntknow - F2 1139
5150Joker - Rev. G2 1142
CrysDark - Rev. G2 1142
eyric101 - Rev. G2 1142
gderreck - Rev. G2 1147 (Replacement)​

Systems NOT Currently Having Issues with Fans and Thermal Shutdown:

Aikimox - Rev. E2 1116
5150Joker(2) - Rev. E2 1116
Shaden - Rev. F2 1123
vzachari - has Rev. F2 1127
bumbo2 - Rev. F2 1131
reborn2003 - Rev. F2 1131
eoinmfoley - Rev. F2 1136
Zero989 - Rev F2 1136​

Systems that Will Not POST:

Mr. Fox - Rev. F2 1133 (beeps when powered on - will POST with 6990M modules removed)
Morgotth - Rev. ?? ???? (beeps when powered on - mobo info unavailable - system at Depot)​

Systems with 580M SLI (Reference Only):

skygunner27 - Rev. F2 1136 (system #1) - no issues to report
skygunner27 - Rev. F2 1139 (system #2) - no issues to report
johnksss - Rev. F2 1139 - black screen/no video resuming from sleep/hibernation; SATA Port 0 drop-out​

NOTE: During our conference call we learned that the revision numbers that we captured from motherboards are only an indication of differing production dates. They should not be interpreted as a change in design, as there was no change in design with any of the above hardware revision numbers. Review the meeting minutes for more information.

Update #3 - 11/26/2011: Fan Malfunction/Overheating - Quick Help

Spoiler

M18x AMD 6990M CrossFire Overheating: Quick Guide for Manual Fan Control

Download the latest HWiNFO64 (Latest Beta ZIP version is v3.90-1445) here: LINK

Unzip to a folder where you want it to be placed.

For a quick and easy way to manually control fan speeds, follow the pictorial steps in the spoiler below...

Spoiler

You may experience hangs during shutdown running HWiNFO64.
If you do, disable HWiNFO64 before shutdown using these steps...

If you decide you do not wish to use HWiNFO64 any longer, uninstall the
driver using these easy steps...

Using HWiNFO64, it is possible to set up custom fan profiles that operate based on temperature thresholds. I have always used manual controls. At this point there appears to be an issue with some of the sensors on systems affected with the overheating issue and using manual controls may be a safer alternative. I would not recommend that anyone with an affected system rely on a user-generated fan profile. (See paragraph #1 from meeting minutes.)

There is additional useful information about using the program on the HWiNFO/32/64 Forum - General Discussion

As you may have discovered, there is additional information on these threads as well: http://forum.notebookreview.com/alienware-m18x/622102-need-help-hwinfo64-gpu-cpu-setup.html -and- Alienware Fan Control

In addition to HWiNFO64, Almico SpeedFan is a great tool. Some prefer it over HWiNFO64. Shaden has a thread on using SpeedFan. You will find it here: http://forum.notebookreview.com/alienware-m18x/625606-new-fan-control-tool.html See post #12 by Aikimox.

Update #4 - 12/08/2011: FIXED: Fan Malfunction/Overheating

Spoiler

The new Dell driver for AMD 6970/6990 is now up on the support site. See details below:

Spoiler

Also for AMD systems, BIOS A05 is also posted. This update is intended to improve AMD graphics fan algorithm to fix system shutdown issue.

Spoiler

Please visit these threads for discussion on the updates:

http://forum.notebookreview.com/alienware-m18x/629157-m18x-bios-a05-released-discuss-here.html

***AMD Driver Release (Dell Recommended Update)***

Alienware M18x - Team Red Conference Call Agenda​

November 15, 2011 – 2:00 PM to 3:00 PM Central Standard Time​

Teleconference Hosts: Bill B., Dell Product Group Marketing and Louis B., Lead Engineer, Alienware Mobility Graphics
Team Red Panel: Mr. Fox, johnksss, Aikimox, skygunner27, steviejones133, mharidas, Nicholaus.rossi, djskorch, nuroo​

I. Current Issues and Defects with AMD HD 6990M

a. Overheating and thermal shutdown​

i. Fans not functional from cold boot
ii. Fans do not spin up until temps are too high​

b. Inconsistent CrossFire function (randomly disabled, clock speeds not in synch)
c. Display corruption and flickering, unexpected intermittent dimming and yellow display hue
d. Intermittent failure to wake from S3, reboot required to restore performance
e. What is the tentative resolution plan and timeline?
f. Total GPU failure (failure to POST, BIOS beeps)​

II. Annoyances Involving AMD HD 6990M

a. Error message from CCC when booting on Intel IGP (AMD devices not found)
b. Erroneous (or absent) readings on fan speeds and voltage
c. Lack of manual fan speed control using CCC

i. Does using a third-party app (i.e. HWiNFO64 of SpeedFan) to control fans cause any harm we should be aware of?
ii. Is there a reason this functionality has been disabled?​

III. Issues with Replacement Parts Availability (AMD and NVIDIA)

a. What are the issues affecting replacement parts availability?
b. Is there a plan for avoidance of availability problems, or consideration to offer substitution options versus of indefinite waiting periods? ​

IV. On-the-Fly Graphics Switching (AMD and NVIDIA)

a. Why is the M18x missing this feature? Technical or cost factors?
b. Are there plans to provide this functionality in the future?
​

V. Driver Support

a. Infrequent Dell updates to support latest games (AMD and NVIDIA)
b. Can we expect regularly scheduled updates from Dell?
c. What are the hurdles for Dell/Alienware to provide enhanced driver support? Cost, resources, AMD/NVIDIA copyright issues?
d. Hardware ID does not distinguish a difference between 6970M and 6990M graphic modules
e. AMD Catalyst Download Manager does not recognize support and denies automated installation (is the “blacklist rumor” valid)
f. Reference Drivers lack OEM features/functionality (e.g. HDMI Audio driver support often missing from Catalyst package)
g. Inconsistent brightness control or total lack thereof
h. Latest Dell AMD Graphics Driver Status Update (delayed release)
​

Conference Call Minutes​

Below are the notes from our teleconference. Special thanks to all of the "Team Red" panel members that helped, and to DR650SE for helping to capture these discussion points. ​

• With respect to the fan issues, overheating and thermal shutdown, a couple of potential causes are under investigation. It is suspected that an interrupt in the CTF (thermal temperature sensor) signal between the video cards and motherboard is causing the issues with malfunctioning fans and resultant overheating. Distribution of systems equipped with affected hardware has been temporarily held up to allow time for the problem to be isolated and workable solution implemented. It does not appear to be a motherboard problem at this time. (The list of revision numbers gathered from NBR forum members is not useful because these are merely date stamps. There are no physical differences between them. To date, only one motherboard design has been produced for the M18x.
  
  .​
• The issue of one GPU not being recognized during POST (which can be a cause of CrossFire malfunction) is also being investigated. It is important to note that CrossFire functionality and synchronization of clock speeds between both video cards is not an indication of a problem. CrossFire and clock speeds are dependent on a number of variables. It is not uncommon for CrossFire to turn off and for one GPU to run faster than the other when load does not demand 100% of the graphic subsystem resources.
  
  .​
• The other issues noted (flickering, lines, display brightness and hue, and failure to POST with BIOS beeps) may related to the same root cause. All of this is being investigated.
  
  .​
• The timeline is simple - as soon as possible and not any longer than necessary to fix this appropriately for Alienware owners. The fix is important to Dell as well as its customers, and it is being taken seriously. The problem has been replicated in the lab on 1 out of the 45 systems tested.
  
  .​
• The issue resuming from S3 is rooted in AMD PowerXpress technology's dependence on Intel's IGP. These issues are not present on NVIDIA based systems because NIVIDA does not rely on Intel's technology in the manner AMD does.
  
  .​
• The issue with display brightness controls not functioning properly (or not at all) is caused because code is missing from the AMD reference drivers. This can also cause issues when resuming from S3 or waking from hibernation. It can also prevent the communication of fan speed and temperature sensor data to third-party utilities such as GPU-Z. There are many lines of code and a number of registry entries in Dell's video drivers to handle brightness control and sleep which are not present in AMD reference drivers. Conversely, NVIDIA drivers are complete and not lacking any of the necessary code. Therefore, these issues are not experienced on NVIDIA-powered systems.
  
  .​
• Built-in manual fan speed control within CCC is not possible. As much as this would be valued by some power users, it cannot be provided because MXM cards used in mobile systems have no fan control devices. On a laptop they are essentially "system fan" and controlled much differently that they are on a desktop GPU.
  
  .​
• The act of using a third-party app such as HWiNFO64 or SpeedFan does not technically cause any harm. Users that find doing so beneficial are not damaging their system by increasing fan speeds and improving cooling. The danger in doing so would be a result of human error. If the fans are turned off or set to run too slowly for proper cooling, this certainly can cause harm. It is not the manual control that is harmful, but what the end user does with that ability that causes harm. As soon as a third-party app overrides the fans, all safety features Dell included by design are being overridden. Do so at your own risk. If the end user does not have a thorough understanding of how to do this, the risk may be great. This practice is not supported by Dell.
  
  .​
• "On the fly" switchable graphics is not possible on dual-GPU systems because AMD and NVIDIA have not recognized a need for it. Dell has asked them more than once to develop the technology, but it has not received support from either vendor. This could change in time, but until they support it and develop the technology, it will not become a reality. One of the issues that prevents it is the lack of control over placement of the memory frame buffer. At any point in time, the data stored in the frame buffer can be resident in either card, spread across both cards and/or cached somewhere other than the GPU. Implementing real-time graphics switching would entail a massive joint effort between AMD, NVIDIA and Intel. The reality is, it may never happen on multi-GPU systems if AMD and NVIDIA don't have an appetite for it, at least not at this time. Offering this on an M18x with a single-GPU configurations is not feasible due testing and development constraints. It would take two entirely different platforms and the cost would be prohibitive.
  
  .​
• Complaints that Dell uses poor quality "paste" and "thermal pads" are unfounded. Notwithstanding human errors that may occur in the assembly process, the materials used are of a higher grade than products available through aftermarket channels. The TIM (thermal interface material) product Dell uses will remain effective for many years. Thermal paste will dry out and deteriorate over time.
  
  .​
• The issues with AMD reference drivers not installing properly is also due to AMDs driver design for the M17x. The M17x includes the Intel driver with the installation. Dell and AMD are working to get the M18x to work with the retail driver for future releases. Currently, for the HD6970 & HD6990, the device ID information is not present in code used by the online driver or installation manager, so the cards are not recognized by the AMD system scanning tool or by manual installation. This is not an issue with NVIDIA reference drivers because they do contain the information. Dell and AMD are working on a better cadence for updating drivers.
  
  .​
• An updated graphics driver from Dell is forthcoming. Dell is working out some bugs and the driver update will be released very soon.
  
  .​

​

Additional Resolution Venues

The first step in securing assistance with unresolved technical issues should always be contacting Alienware Technical Support. Their reps are generally friendly and eager to assist. You may reach them by calling 1-800-ALIENWARE. This will be effective and you will receive the assistance you need in most cases. When a new issue emerges that affects a broad ownership population, sometimes the front line support can be inadequate due to lack of information and awareness. Patience may be required. Leading your Support Rep to this thread in a desktop sharing session may also be helpful.

For assistance sorting out issues where you have not been able to secure satisfactory help from Dell/Alienware Tech Support or Customer Service, a couple of effective venues for escalation that Dell Representative Bill B. recently shared with us:

(1) on Twitter send a message to @DellCares, or

(2) use the Support link on the left side of the Dell Facebook page.​

 

Ok the good:
I received to m18x, 2 weeks ago. I was impressed with the build quality and the leather bound manual. In all honesty I liked the hat too. The spec's I ordered were the spec's I received. I luv the colors schemes and the fact you can make a usb recovery. I luv the backgrounds even, set windows to cycle them every 1/2 hour.

The bad:
1. (not sure if this is ATI 6990 related)
The Wireless HD, has a very poor quality picture, not clear, fuzzy in fact. Even when distances between my laptop transmitter and HD reciever are short like 5 feet. Screen re-sizes are very slow. Youtube videos often have 1080p content, if I switch to 1080p content mid stream the picture goes black for up to a minute or more or even loses connection. The stand alone transmitter has a better picture quality but still suffers the screen re-size problems. I actually would not have ordered this if I saw its quality first.

2.
Sometimes my ati 6990's overheats during game play, causing a thermal shutdown. They are not over clocked. This problem occurred on stock factory drivers. Im now running the ATI Catalyst 11.10 Pre 3 driver with CAP4 profiles. Intermittent overheating still occurs with both drivers. If I reboot the problem does not occur. Seems like cold booting causes the issue in my case.




.

 

Greetings

Pros
-excellent build quality
-screen is very good
-high quality materials
-excellent support from tech and customer care
-2 x hybrid drives not as fast as ssd but very quick
-video (AMD 69990M Crossfire) and cpu (2760QM) performance well above average, given stock video drivers
-this is not a review, so I won't get too detailed. Machine is certainly worth
the price considering build, materials and support

Con
Major issue with gpu overheating:
-first noticed when I attached machine to exterior monitor using HDMI. Attached to Sony Bravia KDL-32EX500 (1080p)
-suffered thermal shutdown about 10 minutes after attaching. No noticeable fan activity before shutdown. This occurred 3 times before I detached from the external monitor. I have not attached since.

-was able to recreate situation with tech support. My service tag was not in my name, so could not dispatch repair.

-problem seems to be with primary card, although secondary card acted up once.

-over next few days, worked with tech to get problem resolved. Tech offered to replace graphics chips and supporting hardware. Back ordered until November.

-opted to get replacement machine.

-there is no pattern to the problem.

-sometimes when I boot up, the temps are within normal (55 - 75 degrees depending on load) range. The fans throttle to maintain temp when necessary. Am able to load the machine without issue.

-other times temps rise to about 90 degrees and then fan(s) run at max to bring down temp. This happens with little or no load on the machine; and minimal gpu activity.

-20 minutes ago, with very little load on the machine, the temps rose to 100 degrees with no fan activity. I shut the machine down, restarted, went into BIOS and reset factory defaults. Exited BIOS and fans came on max. Temp was normal when I logged in.

-the issue(s) seems to occur after starting up the machine. After turning it on, if it's going to be ok, it stays that way. If it's going to max fans on/off, it also seems to be consistent. Otherwise it just shuts down with no fans.

-this is the pattern, or lack thereof, regarding this issue.

I am not going to speculate on the cause. I am not a system engineer. However, we are risking serious damage to these machines, operating them under these conditions. It would make some sense to get a quick fix; whether it be setting minimum fan speeds (with the support of Dell using software or a BIOS patch), or something else until this is corrected.

Heat notwithstanding, I have been quite happy with my first experience with Dell/Alienware. This is a relatively new design, but something is out of whack here. I have a replacement machine coming. My concern is that it will have the same issues as this one. I have no intention of abandoning Dell. I like the machine and Dell's approach to its Alienware customers. Furthermore, I am impressed with this community and the consideration it has for its members.

Moderator: This is a long post. I don't want to clog up the thread. If you wish to remove this post and set it aside for reference in preparation for your conference call, I certainly understand. Thanks for doing this.

 

I just wish Dell updated their drivers more often.

 

Overheating 6990m, I'm experiencing the same known issue as many others.

Fan shows 100% but it's barely on at all.



Edit: I need to add my entire experience.

Here are my complaints:
1. Fans do not trigger, card overheats, system shuts down or video is lost. This occurs sitting on the desktop over a long period of time, or very quickly in game.

2. AMD mobile drivers do not detect the 6990m as a graphics card that is compatible, I have to jump through a bunch of loopholes just to install drivers, this is a bit frustrating.


Here are my raves:
1. These cards are awesome! They scale SO WELL!

2. I haven't been able to notice microstuttering. Whatever work they've done on this, good job.

3. I'm extremely pleased with temperatures, when the cooling system actually works. 65c at full load on stock clocks is fantastic, I couldn't ask for more. It leaves plenty of overclocking headroom for us enthusiasts.

Overall, these crossfire 6990m cards are a 9/10. They'd be a 10/10, especially for the price, if I could just use my computer for anything other than world of warcraft on low running on the IGP.

 

I agree with this. If there were more release like AMD provide, that would be brilliant.

Also, and I know its not Dell's area really (maybe they could have a word with AMD) but what really annoys me is that 99% of the time, trying to download and install direct from AMD site results in error messages "haardware not recognised"... small complaint but nonetheless....

 

i remeber that, i couldnt install the WHQL 11.8 drivers

 

Seems to be a problem with some cards....mine on the other hand are fine so it seems at idle (hooked up to my U2410 atm too)....latency also seems not an issue for me (using bigfoot drivers, not dell's for the NIC)

EDIT: Should point out that I am experiencing random in game shutdowns.....the scenario above was simply at idle without any load. During gameplay I have had Crysis 2 and Black Ops simply just crash my system completely.

 

Could be the reason of overheating bad from-factory paste job? In my system I think it seems to be so

 

Is the consensus, bad bios/driver/vbios, or bad 6990's?. I can wait for new bios/driver but if its the card itself I'd be better off getting new system or return for refund.

 

i think it's hardware related, specifically the 6990m cards.

 

I believe it's not a bad paste job, and I also believe it's not a bad card. The reason is because I did a full fresh install of drivers and it worked perfectly for a day. When the fans DO work, the cards run about 65c under load. The problem is that the fans simply don't kick on.

 

I have my M18X for about 5 mont and haven't had any problem, I installed Ati drivers more than 50 time I think.

 

As a reseller we stress test every system before it is sent to the customers and so far the conclusion is - it would be very helpful if Dell could provide a system BIOS update with better fan thresholds. For example, the fans would often stay under 30% till the temps hit 80C+! This could be an easy fix. 75C -> 80-100% fan power. Or something like that.

Also, things would be further improved with copper radiators on GPU heatsinks. But let's be realistic, - not gonna happen.

Another observation, most systems we tested had poor paste job or poor HS to GPU contact and repasting always slashed the temps by 10-20C (under load).

just my 2 cents here.

 

some of the very points I have suggested as well Aiki

 

I have also experienced the over heating issue with my brand new machine (delivery date 10/26/11)

After reading on here while waiting for it to be delivered, the first thing I checked was if mine was also faulty.

As with other accounts, mine will overheat if it is shut down completly and then rebooted however if i do a simple restart, the issue is rarely present.
I have also noticed on occasion, that on a restart, only the fan on the primary (left side) will work.

If you would like more info on my tests I will gladly link to the other thread and post with detailed descriptions of steps taken, tempatures, displayed fan speeds and other information.

I currently am also on the waiting list for a new set of cards however I am a bit dissapointed that I will most likely be recieving refurbished cards on a system that is less than a week old.

Maybe after this issue is solved, all owners with this issue can recieve a BRAND NEW set of cards insted of getting refurbished cards??? It would be a great gesture on Dell/Alienwares part that even if a set of refurbished cards fix the issue that they offer to replace them with brand new cards.I am almost tempted at this point to hold off getting the cards replaced since I dont want to tear into my new system so soon and now it seems that we have some help in finding a solution.

On the plus side, the Dell/Alienware warranty has been WONDERFUL to me. I will continue to buy Alienware systems largely because of the warranty and will be purchasing an extended warranty for my M18x as soon as possible. I also love the look, feel, design, build quality, basically everything about my M18X. The M15X that this replaced was great but this new machine is more than I could have expected (aside from the videocard issues).

Just my thoughts. Will be following this thread closely.

 

yeah...apparently it is a copper heat sink, but it's aluminum plated for painting.

 

The suggestion bout a firmware upgrade sounds good. Are u talking a 6990 firmware update? Is it possible to change the thresholds our selves with ati tool?
Does repasting our 6990's void warranty?

 

My only issue is also the lack of up to date drivers from Dell. They have never been good at offering them in a timely fashion, even when I had my XPS M1730 with 8800M GTX I always had to download modded drivers/.inf files from laptopvideo2go.com before nVidia started to provide notebook drivers. Is it a huge job for them to optimise the AMD WHQL releases for their machines? It's fine to say that 11.8 Dell drivers work fine and are stable, but customers who buy these machines play the latest games and need the latest drivers when they're released, not six months later.

 

All excellent feedback! Keep the info coming folks, as this is exactly the type of information that Louis and his team can use to gain a clear understand the issues we are encountering, and hopefully address them.

 

I wonder if its a Dell card thing. I have not had any fan issues or thermal shutdowns with my clevo cards. Maybe Im just lucky?

 

Is there a good stress test application that can identify if my PC has an issue with overheating?

 

Furmark or MSI Kombustor really stress out gpu's but if you are suffering from overheating, be careful to not let them get too high or you could toast your cards.

 

I soooo agree. I would also like regularly released up to date drivers from Dell for AMD & Nvidia.

 

I have the Dell cards and have not had any Thermal, Fan or Shutdown issues. Well, not yet......

 

same here, my system was built in mid-september

 

What system bios version do you have installed?

 

I'll keep my info short and sweet since most of the details have already been discussed in the last few pages.....

I got my system about a week ago and the first thing that happened when I used it was a thermal shutdown. Since then I have had a total of 12 thermal shutdowns. I then decided to switch the graphics off and run off of the Intel GPU. By taking this route I got 5 blue screens. The only fix I have found for all of this is to go into the bios when the computer starts up and hit f9 to set to defaults. Then the computer works great and I dont have any problems.


This proves to me at least that it is not a hardware problem but a serious software/bios/vbios problem and simply putting in new cards and heatsinks and fans will NOT fix the problem. Not to mention that the tech guys who will put in the new hardware will more than likely break something on the computer and things will only get worse AND it will most likely be a refurbished set of cards which in my book is an unacceptable option for a computer that was broken right out of the box.

If the problem is hardware related, I feel we all deserve to either have brand new system replacements, or have brand new hardware installed. I have had WAY too many headaches with refurbished parts and have had 2 system upgrades because of them.

XPS m1730 > M17x > M18x < all upgrades because of constant refurbished parts only making the systems fall apart more.

 

Got to agree with the general sentiment about replacing new parts in a system with refurb ones. Not only that but getting a system free of defects the first time around isn't very easy in my experience and the last thing I need is some random tech tearing into the system. Would much rather have a new factory replacement with fixes in place and/or a bios/vbios release that addresses the problem.

What I don't understand is why older systems from August/September don't have this issue. It could be a hardware fault with the new motherboards or something changed in the latest bios (mine came with A04).

 

apart from that I've used A03 and A04 with no issues.

 

Thanks, may give the vbios a shot.

Update: Tried system bios a02, a03 and now a04 with the vbios above and no dice. System still overheats on initial cold boot and the fans only work when a restart takes place. Could be a hardware problem since others don't have this issue or maybe compal or AMD created a newer revision of the motherboard/video cards that doesn't work well with the existing bios. Who knows, but Dell needs to figure this out asap and get back to us.

 

FYI, I have the same vBios that Zero posted in gpu-z (mine came with that version from the factory and is unchanged) and my latest rig shipped with A03 which I have since flashed to A04. No temp problems par se before, actually I think I have lower temps after flashing main bios to A04. It must have something to do with the cards themselves individually I would imagine....

 

When did you receive your last system? I got mine about 4-5 days ago.

 

I've had this one now about a month to six weeks. I've gotten into the habit of creating the recovery media on or around the day of receipt and it was done on 17th September....Just another FYI Joker, maybe of some use or not but "info-wise" - here are my temps after about 1 hour of Crysis 2 all maxed out ultra/dx11/high res/etc etc....played on my U2410 whilst M18x screen off & U3 cooler full blast...ambient room temp no more than 20c.....no cpu or gpu OC either....cpu with OC disabled.

EDIT: Seems that sometimes, I can get away with gaming for a bit but other times, the system completely locks up or shuts down. This is happening on any game I try atm.

 

ok...so it would seem to me that more people with out an extreme cpu are having far more fan problems than people with them....and since the other cpus dont really get hot...it may be why the first gpu fan will take longer to spin up. and keep the second gpu fan off.......speculation of course... needs to be further looked at.

 

very nice temps, did you repaste the gpus?

 

Just reading thru the post so far, and throwing this out there

People who have to problem ---> purchased recently in US
People who don't have problem ---> non US, and purchased more than a month ago

 

my order shipped from TX if that matters.

 

when did u purchase, are u having thermal shutdown issues?

edit:
purchased in September, shipped from TX.

I'm in Brooklyn, first fedex scan was Indianapolis, IN

 

Nah mate...no repaste for me. Dell factory paste/pads etc. Someone at the assembly line must have had a good day!

 

Order Date: 9/16/2011

no issues.

 

My order was 9th Sept. 2011 .... UK order obviously for me....

 

It's not a cooling issue. The fan simply does not come on regardless of the temperature. When the system is rebooted it works fine and temps are great.

 

august 2nd for my amd setup

it does come on. it's just on so low you think it's off most of the time.
if it didn't come one...your laptop will over heat about 40 seconds after just being on in windows.

turn the laptop side ways so you can actually see the fans. take the bottom cover off to see what is spinning. then start the game. and just let it sit.

 

When more people start reading this thread hopefully they can post when, where they ordered and if they have issue. Also their config if not in sig.

(It could be unrelated, and circumstantial, just noticed problem with recent US purchaser's)

 

and after reading through a few threads...it's pretty much everyone without an extreme cpu installed. not sure if shaden had this issue with his 2630 though

 

Wonder what the relation of extreme cpu and 6990's shutting down is?

Do the extreme's have better cooling? Is there a bios setting that helps them some how, making their fans kick in on cold boot? maybe the extra ambient heat effects 6990's causing fans to kick in sooner (pure speculation)

 

not sure why...but the fans ramp on more with the extremes. i asked about getting user controlled fans and the answer was simply put "no" because how the bios controls them. and from my testing...the xm chips seem to keep the gpu 1 fan on at normal speeds for most of the time. it's when you start using hwinfo..that you get stuttering in games and then all the fans will cut off in the middle of the game and guess what...laptop over heats and shuts off. you have to do a few things when using that method.... some can get lucky with just manual max fans, but most will get the fans dropping out. (turning off completely - and take about 15 seconds to turn on gpu 1 fan and maybe your cpu fan will come back on.

side note:
never paid attention to the restart after a cold boot though...that sounds interesting.

 

My system is still at the depot, but my issues were similar to all of yours. My second 6990's fan was not activating, and in addition in hwinfo64 and gpuz sometimes the second card system specifications were not detected past the model of the card like this:



or the clocks would remain extremely low even though I was running a game that utilized crossfire and should have been using both cards, as seen in this picture where hwinfo64 actually revealed more than just the model name of the card, this was after I toggled crossfirex on and off and I believe restarted the system some of the times as well:



Initially I had been trying to test a game again on my system and it completely froze. It did this once again later on after all the tweaking and monitoring I mentioned earlier, and then when I went to turn it back on, the system powered on, but the screen did not turn on and it beeped four times in a row every few seconds. After talking to a tech support representative on the phone, a tech was sent out and he looked at it and replaced all four of my memory chips. This did not solve the problem, so he tried each one of the new sticks one at a time in all four of the slots, which did nothing. So I had to send it to the depot as I mentioned earlier, so we will see what happens when I get it back.

I ordered my system on September 29th, and it arrived on October 17th. My system specs are as follows:

2960xm
8gb total consisting of 2gb x 4
Dual 6990m's
320gb 7200rpm HDD

 

my second gpu fan stayed off more than it stayed on.