i have an alienware m17x r3 with nvidia gtx 580m,updated the graphic drivers,running on high performance battery mode.when i try to play any game the game runs smoothly for five minutes and lags for another five minutes and it continues like this.dont know what the problem is.i need your help in this.also the ventilation is fine and there is no overheating
I am having the same issue with the same machine and graphics card. I hear the fan kick in and games become a slide show. This is a new issue and was never like this before with the same games. This is not a heat issue, as it will happen within 5 min of turning my machine on for the first time that day. It can operate fine for 10 min to an hour and then I'll have another hour of periodic freezes. Sometimes it will act up for the first 20 min and then be fine for 30 min. this is not a question of loading a new area or anything of the kind. It will just simply become a slideshow at intermittant times and I can not seem to discern any definitive pattern.
What you are both experiencing definitely seems to be a part of the throttling on the 580m. Due to skin temps and i/o port safety concerns, the 580m has been given two throttling points.
Download nvidia inspector and you can see the specs on the different P states of the card.
P8 is typically used for web browsing and other minor activites, p12 is used mostly at low idle
When you play a game, the card goes to P0. This is the highest performance level.
When you hit 63C is drops to P1, which is about a 20% decrease in performance
When you hit 78C it drops to P8, which is the same performance speed when web browsing, or when unplugged.
There are a few ways to overcome this. First off, download the vbios off of notebookreview.com
I would also download gpu temp and use it to monitor your gpu temps. Keep in mind that the temps recorded are only core temps, and there are other temp concerns on the card.
....and also, download FRAPS and have it running in the background. This will allow you to see on your screen, when playing games or videos, when FPS drops to a low amounts. Then you can exit the game and check temps. This will help you to track whether it is really the temperature or something else. HWiNFO is also great for temps, but I find GPU temp to be easier for most.
Thank you for the post. I'm not sure if I'm ready to void my warranty. But as it turns out I think you're right. I didn't think it was a heat issue as it was happening sometimes within 5 min of starting the machine. I am using a cryo cooling pad now and that seems to be controling the heat for me so I'm no longer having this issue.
I appreciate your response and will keep this handy in the event that I do need to deal with this later and the cooling pad doesnt keep things manageable.
Glad to help. Keep in mind that nothing I recommended doing will void your warranty.
It is important to keep tabs on the heats and other issues, as there have been issues with the cards being improperly pasted, and if it is something like the causing heat then it will need to be addressed by a tech.
I am experiencing the same problem as well - I have an M17xR3, GTX 580m - got the latest Nvidia drivers and the latest System and Graphics BIOS too. The machine is only 4 months old and I noticed this problem right off the bat, but it would not occur too often earlier - I wasn't playing too many games and when I did, they were not that intensive. Over the last month, with Mass Effect 3 and say, Reckoning, the problem is much more frequent.
I use a dual monitor setup so I can keep Nvidia Inspector open while I play a game on my primary monitor. After playing for around 7 minutes, the temperature rises to 78 (I don't get to see 78 because I believe the throttling cuts in immediately and the temperature starts dropping - it is at 77 for a minute) and the game is un-playably slow. I think it's likely around 5 frames per second. Once the card cools to 62, everything works fine again, but the temperature starts to rise immediately and it reaches the 78 threshold much faster. I can see the change in the P State from P0 to P8 and also the clock speed drop from 620 Mhz to 72 Mhz when the throttling is in effect.
This post http://forum.notebookreview.com/8488465-post283.html within the thread you pointed to earlier indicates that Dell is now aware of the problem and proactively fixing systems with a heatsink replacement kit. I hadn't yet received this email from them, so I contacted Dell myself. Dell India (where I am based) appeared to be totally unaware of this problem and they were unable to get further information on the problem either - I thought they could contact Dell International (or US) and get the official solution to this problem - but that hasn't happened.
Dell did send a tech over with a new heatsink, thermal paste, graphics card fan and even a graphics card too. The tech first replaced the heatsink after reapplying the paste on the existing card. He replaced the fan as well. This did not fix the problem. He then tried with the new graphics card (which was identical to the existing 580M which I already had). Unfortunately, this did not fix the problem either. So I now have a new fan, heatsink and card, but the problem remains.
They then tired uninstalling and reinstalling the Nvidia drivers and said I should reinstall the OS. After four hours of troubleshooting I was in no mood to do this, nor do I believe it would fix the problem. So I have a call scheduled with them later - I will have to go through this process to reach the next level of support, quite a waste of time. I believe the next chained solution is to replace the motherboard too. Again, I wish they would contact the right team at Dell and get the official solution.
I do not want to use the original solution mentioned in the notebookreview thread which would practically keep the GPU from throttling when it reaches 78 C; if it is not throttled, then I believe the temp could continue to rise to dangerous levels and actually damage the card?
This appears to be a design problem, extremely disappointing that I cannot play a game for more than 3 minutes at a stretch on a machine that cost me nearly $4K. What is Dell's official stance on this? Who in Dell International / US can I contact (or have Dell India contact) for an official response?
Appreciate your support and any suggestions you may have.
brian, I have yet to hear back from Dell on the email I sent in to support concerning a heatsink replacement/vid card replacement. But I have found a few work arounds that have helped. I use my laptop for both work and home, at home I bought a Coolermaster U3 laptop cooler. That has helped the issue there, at most the GPU will get to 70-72C, so I don't experience the extreme throttling. At work I simply put an old removable cd drive from a laptop long since dead under my system. I positioned it exactly in the middle of the system so both fan vents in the back are clear. This has helped me at work when I need to...um...test my graphics. 😉
I did mess around with the HWINFO program and tweaked my GPU fan speeds for awhile, and it helped immensely. But I've set them back to the default settings and seem to be doing ok as long as the laptop is not sitting flush on a tabletop.
Really, I think the easiest fix Dell could put out is a quick utility we could run that adjusts the GPU fan's speeds to higher speeds. Obviously any heatsinks that just aren't on properly would still need to be replaced, but at least this would be an easy band-aid until they can roll out a better solution. HWInfo is not very friendly to use for most users and its stability leaves a bit to be desired.
Thanks for responding, keeping my laptop elevated was what I was going to try next - but I didn't have much hope until I read your reply. This has practically fixed the problem! So, I'm just using a CD spindle cover that's about two inches high and my temperature has not crossed 70 after even an hour of playing. I don't think I'll be trying further troubleshooting with Dell as I strongly doubt their current solution (replace all parts with brand new parts which are exactly the same) is going to work. And I'm certainly not going to format and reinstall my OS just to follow procedure.
I had a similar heating problem with my Inspiron 8800 about 6-7 years ago. Dell came in with a *different* heatsink and that fixed my problem immediately. The old heatsink design / build was faulty. This does not seem to be happening here and I'm still disappointed (albeit not so much) that this expensive piece of hardware doesn't seem to have been designed correctly.
HWINFO did seem complex and I will keep that solution in mind for a rainy day, thanks.
I would still like to get an official response from Dell though - if they do reply to you, please keep me posted.
Never play while you're on battery. Plug in your laptop. That will prevent lags while you're in a game. Our laptop will save as much power as it can while on battery. Make sense?