[Conference Call] Nvidia GeForce GTX 580M 78C° Throttling [SEE POST #191 and later]

Please click here

7/31/12
The original post was written when all customers were still inside of warranty. For out of warranty customers, you will need to contact Dell Spare Parts for you region and purchase the part below -
RFYN9 N12E-GTX2 GeForce GTX580m Kit (Video card 3MF8R + Thermal Assemblies V1K2G)

You will need to do the removal and installation yourself. Refer to the online Service Manual for instructions:
Alienware: Online Documentation

4/4/12 Updating with information from the con call -
Why does this occur?
We have to look at the entire system when dealing with temperatures, not just one component like the MXM slot. So with the heatsink installed properly, the M17x-R3 system should not hit 78C° in a normal gaming environment. The 78C° throttling has nothing to do with the MXM slot. It is entirely the GPU and skin temperature requirements, thus, the thermal solution and ambient temperature are the only factors. If the system is not hitting 78C° within a few minutes (like two) with any GPU stress application, the system is very likely working as it was designed. Hitting throttle points for graphics is a normal occurrence with the Nvidia GeForce GTX580m. If you remember, there were two different types of throttling, power (wattage) and temperature (GPU). For power throttling, the system will use P1 and for temperature it uses P8. What you are seeing is not an M17x-R3 MXM slot defect, but rather a result of the system design thermal requirements.

If your M17x-R3 is throttling within 2 minutes when running a GPU stress application, then the heatsink and thermal paste were not configured properly.

* For USA and Canada customers, contact: [email protected]
* For all others, contact Dell Technical Support for your country

* Provide the following information:
Service Tag =
Name =
Shipping Address =
Phone Number =
Best Time To Reach You =
Email Address =
Issue = M17x-R3 GTX580m throttling at 78C°
Need = Onsite Service or Parts Only Service
Parts Needed =
3MF8R GTX580m Video Card
V1K2G GTX580m MXM Thermal Module Heat Sink

 

you have got to be kidding me..

 

Strange how Dell says that it is implemented by video card manufacturer and yet all other laptops from other brands with the same video card don't throttle at 78C...

 

neat trick, "clean the vents"

 

so is this the solution that everyone's been waiting for?

 

Especially considering the fact that Dell fixed the 580m throttling issue on the m18x!

 

All,

There is an email listed in there that you can use to voice your dissatisfaction with the article. But please be courteous.

 

No need to flame. We're just trying to help. "Retarded" references have no place in this discussion.

 

Just to be clear, this exact same issue was resolved on the M18x with a BIOS update, correct? I just want to have all my ducks in a row for when I send in the email complaint. I will be courteous, as always, just really need to make a point on this that it's not us. We've re-pasted, re-padded, cleaned, and elevated. Point of the matter is that it throttles at 78 degrees, and it shouldn't.

 

I'm guessing that Round Rock handed this issue off to the Indian telephone support techs, who simply implemented the standard policy of denial, delay, disinformation and indifference, coupled in this instance with astounding incompetence. I found this particularly incisive and responsive:

"When Video Cards starts reaching the throttling temperatures, it means the computer is experiencing some internal and external conditions which make temperatures going up. Software and Hardware can make changes on the system temperature."

This Memo and response is an unpredecdented level of lame, even for Dell.

@BillB and ChrisM -- Thank you for your efforts in connection with this issue, but you -- and we -- are playing against a stacked deck. I feel like a complete idiot for ever believing or suggesting that Dell would rectify this obvious defect once brought to its attention.

 

Guys, being disrespectful to anyone, including our highly valued members that are Dell Representatives, will not be tolerated. It is perfectly acceptable to express your dissatisfaction with the throttling in a civilized manner, but smarting comments and the like will not only be a disincentive for Dell/Alienware to help, but it may result in infractions or banning depending on the content.

I think some apologies are in order, and I will be the first to apologize (on behalf of others) for things that were posted that are inappropriate.

Please keep the comments civil so we don't have to lock the thread or take adverse action. I invite some of you to edit your posts. A number have already been deleted.

 

Seems the M17x owners with the 580m got the short end. Consumers buy this notebook for gaming and after an hour or so of gameplay when it throttles, it's not serving its purpose for the user. With this acknowledgment from Dell now, I wish I would have gotten the amd radeon instead.

 

Why was this fixed so quickly for the M18x users so quickly and not here?

 

I own an early R3 with a 6970m so do not have the throttling issue. However, as an Alienware fan and satisfied customer I have been watching this 580m situation with interest.

What I have observed:
1. Dell knows about the throttling issue in the R3
2. Dell fixed the throttling issue in the m18x
3. Dell blames the R3 throttling on dirty vents and bad paste jobs
4. Dell continues to sell the 580m (at a price premium!!)

I have been showing off my R3 and recommending Alienware to anyone interested in PC gaming for almost a year now. I think that is going to stop. Very disappointing.

 

A disappointing response, and hopefully not the final resolution.

On the bright side I guess we can now email resolution experts and try to get things resolved. Sadly, this is not a heat sink issue, a tiny bit of thermal pad won't fix it, and I fear contacting the resolution experts will just prove a colossal waste of time.

This is a bios issue, specifically it is the Compal EC which causes this behavior. First, the EC employs extremely lenient fans, so temps are allowed to rise too high. Then, the EC won't kick the system into high cooling mode before we hit 77c, and then when we hit 77c we are bunted down to P8.

What Dell needs to do is go look at the M18x and find this issue. Then send all that documentation to the M17x engineers and tell them that it is the same issue, and they should just do the same fix. The M18x had the issue resolved with a bios update and that is what will be required to fix this issue on the M17x.

 

Now I guess I have to decide if I'm willing to put up with this or return the system for a refund. 78C is a very low temp at which to start throttling. I would imagine it won't be long before your average DX11 game is pushing it hard enough to keep it in the mid seventies all the time. It just doesn't make any sense. With proper fan thermal profiles this card can hum right along in the eighties all day long.

 

I was just not expecting this level of answers from Dell reps, SO unprofessional (we are customers, we are not supposed to be professional, that is a courtesy already given to you because of your good service and because we are overall satisfied customers), first time hearing a response like this in any business relation and from a Dell rep, still cannot believe it...

anyway, people already pointed out to obvious, why it is impossible to do the same what Dell did for m18x, no clue, hope they will change their attitude towards this matter, and their manners contacting their customers... disappointed.

 

I am planning to write a guide soon, maybe even tonight, that may prove to be a fix for these issues. It was something another member brought up a long time ago, and I have recently been investigating it since I found that The Witcher 2 would consistently hit the 78c throttle, so far it seems promising.

Hopefully Dell will take another look at this, since this response feels like it was thrown together in a couple of minutes by people that did not really investigate the issue, the community has done far more research than it feels like Dell did to write their response. This is why there were so many angry people in this thread, a lot of people have REALLY struggled with this issue, and to get "Clean the dust, add a half inch of thermal pad, and all will be fixed" as a response is pretty infuriating. We have people that have replaced all thermal pads and paste with top grade stuff and still have this problem, it is not a heat sink issue, it is a bios/Compal EC problem just like with the M18x.

 

I have The Witcher 2. Is there a benchmark in it where you can just let it run through some screens to test? I haven't actually played it yet. It's in my backlog. I bought it during the Steam holiday sale. I will play it and see if it runs me up against the throttle temp. To be honest, the games I have been playing since I bought it get no warmer than 71C and with the new 295.73 drivers I stay below 70C all the time. At least, in the games I've been playing. Also, I did repaste my GPU and CPU with Arctic Cooling MX-4 already.

 

Who I'm sad for the most, is the average consumer. We geek out and do research on here, repaste, do all the tricks. But an average person that wants a gaming laptop and has stuttering and a slamming halt to their games while playing, would probably be even more upset than us, because they don't understand. even though I would repaste and such either way, people shouldn't have to.

 

You can just play the tutorial or something for a quick test, be sure to run on high settings and set vram size to Very Large. However, my issues actually came after extended gaming, the default fan profile never kicks up to max speed, and The Witcher 2 would gradually heat up the GPU until I hit 77c, I got my first throttle near the end of the Prologue chapter, which is maybe a few hours long.

Before The Witcher 2 I had not had any game take my GPU over 65c on my new system (my first M17x had serious throttle problems), from Modern Warfare 3 to Assassins Creed 2, Syndicate, Darkness 2, and many more. All those ran at around 65c with the default fan control.

I went ahead and made my guide if you do happen to hit the throttle, it has resolved my issues with The Witcher 2, both with a more aggressive fan profile, but also because it seems to disable the 77c throttle.

http://forum.notebookreview.com/ali...e-fixing-m17x-r3-580m-thermal-throttling.html

 

OK all. I no longer cover Alienware platforms, but I'm trying to get the appropriate response from the appropriate team. Having said that, the KB article linked to here is NOT the official response. As far as Dell reps responding unprofessionally, feel free to PM me and point out what was unprofessional. If you are referring to my negative response to a previous post in which the word "retarded" was in the caption of a picture that was posted, that is what you will get out of me every time. I don't like that word used in that sense for deeply personal reasons, and I would appreciate your understanding. If it is a different post you are referring to, graz 'zt, feel free to PM me and let me hear about it.

If no one posts a response by mid day tomorrow, I guarantee I'll come back and shoot straight with you. I will give you the answer we get, and if you don't like it, I'll be happy to advocate on your behalf as hard and long as I have been doing since I joined this site. I am on your side, and I'll tell you when we've hit the wall (meaning we have a final answer none of us like). So far, the issue has only gotten out of chapter one.

You guys want to do a conference call for this one?

I'll return tomorrow. Right now, it's time to get my unruly 3 yr old to go to sleep.

 

Thank you, Bill. You've always been a straight-shooter and I don't expect that to change. Looking forward to hearing more on this subject as soon as more can be shared.

A conference call may not be necessary, but we can do that if the Community feels there is a need.

 

oh I see, didn't see the other post with this picture including "retarded" on it. That's my bad then and I take that (unprofessionality) back (I was so shocked when I saw it, but it makes sense now, some posts are deleted I guess?). Anyway as disturbing as a problem this is, I am sure the consensus is that alienware builds one hell of a system and pretty much everybody loves their laptops, even with throttling and so on, I am utterly satisfied with alienware so far. Thank you for explaining the previous situation, good luck with the kid!

 

Bill you truly are an awesome company representative. Dell is a better company simply because of you and those like you. I just wanted to say that I really appreciate your help so far and your continued assistance, especially now that this is really not your problem anymore. I also have not seen any unprofessional conduct by a Dell representative, you guys have always been professional and helpful with me.

It is good to hear that the article was not the final response. I think that a lot of the backlash in this thread is that people interpreted that article as Dells idea of a resolution for this issue. I am still hopeful that the correct bios changes can be made to fix this issue.

I will be happy to assist in any way I can, if something is needed then feel free to contact me. I have done quite a bit of research into this issue, especially since this last week I was playing The Witcher 2, which is the first game that has caused throttling on my new replacement M17x. My tests and the "fix" I have discovered suggests that the Compal EC is the culprit here, and that it should be updated in a bios update. That could involve removing the 77c throttle, or making the fan profile much more aggressive to avoid the GPU ever hitting 77c. Again, the bottom line that I would tell the engineers is just look at the M18x and replicate the fix. The two systems are extremely similar and our problem is identical to the M18x's old problem, the fix applied to the M18x should work just as well here.

Thanks again for the reply, especially as it is clearly after business hours. We really do appreciate your efforts, and we all hold out hope for a satisfactory resolution.

 

Thank you sincerely, Bill for addressing this problem with us. It is awesome that you go above and beyond the call of duty to work with us. I apologize for the attacks. Folks in NBR forums are typically very knowledgable about laptops and very serious (too serious sometimes) about the performance of their machines. Sometimes the group gets a little emotionally overheated on issues. On the flip side of that, we are "that dude" to our friends and families that everyone goes to for advice on what computer to buy and how to fix things. When we tell them Alienware is awesome, they listen.

I sent a respectful email referencing the research work that Yoda did to show the card runs stable and safe at temps above 78C. I hope they read and consider it. The problem is with the fan ramps with respect to the thermal profile. It's very fixable and thank you again for bringing this to the attention of people who can possibly fix it for us.

 

There's a lot to this situation. I will have to huddle up with the team to talk this through before any of us can can talk specifics, but trust me, there is sound reasoning behind the number- and it's 78, not 77, BTW. That's not to say there isn't sound reasoning from you guys, though. It's a classic communication issue. There are tons of reasons for not responding to some of the specifics you guys are discussing. I'll see which of those reasons we can overcome and communicate a little better as a company. Not to worry. We'll get it sorted one way or the other.

 

I have a feeling the 78C temperature reported by the sensor is incorrect. With the,M18x, if the temperature reaches 75ish the upper left side of the computer feels quite hot.

Sent from my GT-N7000 using Tapatalk

 

From what I'm told, the pertinent line of code in the BIOS has a value of 78. I'm not the most technical guy in the world, but my guess is either the software indicating 75ish or the sensor it's reading is off a bit.

 

Using the stock vbios, my R3 throttles at 78C exactly as reported by HWiNFO64, GPU-Z, Aida Extreme, etc. I don't know whether those readings are precisely accurate, but they are at least consistent.

 

Just a guess, but often the temperature reported by the GPU and by the Compal EC are slightly out of sync. In my experience it throttles based on the EC, so if we are using GPU-Z to monitor temps it may show a max of 77c as reported by the GPU, when the EC saw 78 and bunted us down to P8.

Maybe we can take up a collection to incentivize your engineers to "accidentally" transpose that 78 into an 87?

If there are reasons behind the 78c throttle that means it won't be removed or raised, then maybe an adjustment to the fan profile, I have recently been using HWINFO64 to override the fans and that has proved a reasonably effective solution, except that the software isn't perfectly stable and only runs the fans as a set rather than individually.

Anyway, bottom line something still needs to be done, as this does actually effect normal gaming at stock settings. I didn't have any issues with my new unit (in fact it stayed around 65c) until I started playing The Witcher 2. Now without using HWINFO64 to override the fans I will always throttle after an hour or two of Witcher 2 game play. Basically using the normal fan profile my GPU slowly heats up and eventually throttles when playing, however, less intense games like Modern Warfare 3 or Starcraft 2 are stable at 65c.

 

Yes, very much this. The 580m is rated to withstand up to around 100 degrees. The 78 degree ceiling that is implemented by the BIOS is WAY to low. People should be able to enjoy the full extent of the product they purchased.

 

I think you may be onto something there.

 

Hopefully you will be allowed to provide some more details, but I do wonder why the engineers feel there are reasons that the 580m should always stay below 78c, but they let the 6990m go as high as it wants (up to 100c).

If it was an issue of the heat harming other components then I would think the 6990m would have the same thermal restrictions, yet it doesn't. In terms of design Nvidia says the 580m is good up to 100c just like the 6990m, so what is the difference between these cards such that one is limited to 78c and one is allowed to go to 100c?

Anyway, a fan profile update would probably solve the problem in most cases, just please be sure they also fix the SATA speed issue introduced in A09, it will be a bummer to get a bios that fixes the 580m but includes the A09 stuff which cuts my SSD speed in half.

 

Keep connecting dots, Yoda. The more points like the last 2 you bring up, the longer it's going to take me to get a response together. (A problem we want right about now.)

I have the response all written up, but I'm having to get 2 of Yoda's points reviewed before I can post. Hang in there.

 

You're awesome. I wish every company had employees as helpful and dedicated as you.

 

@Dell-Bill_B -- Dude you are awesome!

It is great to see an employee who actually wants to help out the people. A few of us are from hot climates (I am from India) where the room temps hit 50'C (yes, fifty degree Celsius) so we hit the 77C throttle regularly.

Please do make sure that--
i. (most imp) we get the throttled removed from the card or atleast the throttle temp increased to the high 80s.
ii. proper sata 6gbps speeds in ssd. even if only in one port.


Currently waiting for my 2nd laptop replacement in 2 months. This heat issue is probably what killed my laptop this time.

 

OK. We've had some good initial internal discussions. Good in that information is starting to flow. I want to first set expectations. Then I'll expand a bit. Firstly, whatever we ultimately do will not happen overnight. We're talking in terms of a few weeks. Most of this time will be spent on discussing alternatives and outcomes. We're going to need some time. Second, whenever we do arrive at a course of action, things will move very quickly, and this topic will close out quickly.

We have pulled together engineers from 2 different groups to look at this. One is the quality team, whose focus is on customer experience and warranty service. The other is the video design architecture team.

Some of the engineering resources who will be looking at this are in the middle of some projects for future products, and some are finishing up projects. So, their bandwidth will be limited at first. The best way we, as a community here, can work with those limitations is to do some information gathering and documentation.

To that end, I'm thinking we need to separate into 3 groups of end users:
1. Those who want to crack the case, try some hardware reseating, and tweaking of settings.
2. Those who want to have us replace parts to see how symptoms are affected.
3. Those who want to do nothing and wait for a fix.

Decide which group you want to be in. Chris will be posting some steps to establish a baseline for those in group 2. I will look at the data already posted by everyone in group 1. If I have any clarifying questions, I'll pop in and post them. We'll use this thread to avoid fragmenting the conversation further.

Feedback? Hang in there folks. We'll get through this best by working together. Chris and I are here to help.

 

Thanks for the update, Bill.

 

This is a great positive step Bill. I'm a newer member (been around since last year, but now finally getting my name into the community). It's great to know we have people on the inside that we are able to interact with.

Currently I'll be in group 3 until I get my system (hopefully early next week). Once I do get it, I'll step up to group 1. Once I get settled in England (which will be next month) I wouldn't mind joining group 2 if needed, and the resources would be available to me out in the UK. Looking forward to lending any help I can to the cause.

 

Well im gonna be in group 3

 

Thanks Bill!!! We really appreciate your time!

I will be going along with group 3 as well, as when I prop the back of laptop my temperatures are fairly low, but I would appreciate the fix as we need some overhead for overclocking

 

Thanks, Bill, for everything so far! By 2., do you also mean to include those customers who have already had replacements for all the concerned parts, for this particular issue? If so, here is the list of parts Dell India replaced during the service call for my 5-months old M17xR3:

1. V1K2G: ASSY, HTSNK, VGA, N12E, GTX2, M17X


2. F1RJC: PAD, THRM, PSX, CPU, AW


3. GVHX3: ASSY, FAN, CPU, M17XR3


4. 650RY: ASSY, HTSNK, CPU, M17XR3


5. 4K1MM: ASSY, FAN, MXM-CARD, M17XR3


6. 3MF8R: CRD, GRPHC, N12E-GTX2


7. 5VYM9: ASSY, PWA, PLN, ANW, M17XR3

Needless to say, the replacements did not sort the issue.

M17xR3 i7-2760QM 2.4GHz, 8GB DDR3, 1080p 120Hz 3D, 2GB GTX 580M, 750GB HDD

 

as much as I would like to be indignant and angry about this, I really can't. I've used my M17x for a few months now, and I have not hit the thermal throttle once while playing games. I've played Skyrim for hours on end, and every other game I own. No throttling whatsoever.

The only time I was able to get the laptop to throttle was with Furmark... So I dunno what Dell did, maybe they just did an awesome paste-job out of the factory but I'm not having throttling issues at all.

I have also used my laptop on a chair, on my lap, on a laptop cooler stand, and even on my desk... doesn't matter where it is, whether it's propped up or not, or what.... plays games for hours on end, and never throttles.

 

I am glad that you guys are looking at this more in-depth. I would probably be in group 1, I am willing to help, but I won't be willing to reseat my card, as I feel like it is perfectly pasted at the moment (I get 65c max temp in every game except for Witcher 2 and BF3).

I am sure others will confirm, but replacing hardware won't do a thing, many members have already replaced every related part with no change in behavior, so I would not be up for any of that, as it will just be a pain with no gain.

I would really encourage your engineers to consider the question of why the 580m is throttled and the 6990m is not, while also realizing that no other gaming laptop throttles the 580m, and they are not having issues. If they conclude absolutely that the throttle will not be raised then they should look at the fan profile. The issue is that in some instances (which vary by how good your heatsink is seated and pasted) the card will continue to heat up even at the highest cooling setting.

A quick fix would be to add a fan trigger at around 73C which would run max fan until the system cools to 68C. This would break that slow increase in temps and effectively reset the cooling system. I have been running a similar profile with HWINFO64 and it proved completely effective in eliminating my Witcher 2 throttling issues. In case you guys wish to test it, max heat generation seems to be when in conversations, especially those with a few dynamic light sources in the background, and keep in mind that with a perfect heatsink job it is a slow but constant heat-up, I can't just fire up the game and throttle, I have to play for an hour or two before the card hits that critical throttle level. After it hits it once it will continue to hit it constantly.

 

Shame that the 580m throttlling is not yet sorted out in the M17xR3, as having a Dell 580m, I can tell it is running extremely well in a mere 15.6 notebook so I understand how disappointed some people are.

 

Perfection. Thank you very much.

 

Been watching this thread closely with me having a new R3 and suffering the problem and I'm beginning to think there is much more going on than dell are admitting to. I'm very grateful that we have dell reps here and working the problem but with what I see in other threads about the 580 being removed from the build and not even a 'out-of-stock' notice, the MX18 getting a fix easily, other laptops with this card having no issues, I start to think what are they (Dell) not saying. Tell me I'm way off-course but the only thing I can think of is that the MX17 R3's power supply is not up to the job? I see the P1 throttle at very low temps, around 65 and always when GPU utilisation is over 95%. If the throttle is to protect a poor power supply then it could cost Dell dearly. I hope i'm just being paranoid but having fought with this problem from day-one of purchase I'm not seeing any logic for dragging out what appears to be a simple BIOS fix...

 

If you have the low throttle, why not apply my throttle fix? It works for that.

 

Thanks, I did, I'm now hoping for the one you are looking for, the P8. i get that after an hour or so of gaming and right now all I do is switch out of the game for a few mins for it to cool down before returning to the game.
My first thoughts were that there was a component nearby that didn't like the heat but the ppl with AMD's are happily running in the high 80's so that theory went out the window.
I haven't seen a good reason as to why all the focus has been on reducing temps when the chip is rated far higher and indeed does run OK higher in other machines