M17X R3 + GTX 580M - Random Lag's - Help

I can test in more detail when I get my new machine, but my house is generally 20-24c Ambient and I commonly hit the 77c throttle. Obviously, with a higher ambient temp I hit the throttle faster and more frequently.

 

To demonstrate: A08 on the left, A09 on the right. (Crucial 256GB M4)

 

Hello guys!

I'm a french owner of the M17x since christmas and as everybody here, I have experienced the same problem at 63 and 77 degrees with the GTX580M. At 63, core frequency is 475Mhz instead of 620Mhz originally (so it's a big drop of fps in game) and at 77, i'm around 5fps.

I'm disappointed and also glad to see i'm not the only one with this problem. Before, i couldn't understand why my own benchmarks have nothing in common with the benchmarks I have found on the Internet before purchase.

I hope Dell will find an issue as soon as possible.

 

You can fix the 63c one with the 580m vBios throttling fix. It is pretty easy and fast. Hopefully they will fix the 77c throttle and then it will be perfect.

 

Thanks for the answer. Is it the fix you re talking about :

http://forum.notebookreview.com/ali...throtting-fix-no-need-modify-system-bios.html

 

Yes, just follow it and you should be set.

 

I'm facing the same issue of 77-78 C throttling on my M17xR3 bought in India. Is there a tentative date from Dell's side for fixing this issue?

 

Hey guys, I've been running the m17 since early december with the same throttling issue, still hitting the 77c mark in a freezing canadian basement. Glad to see it's being dealt with!

Thanks for that VBIOS fix btw, much appreciated

 

No pics? This will slow things down. We have to rely on the system capture process just to have a peek at the heat sinks.

 

Sorry, I just hadn't gotten around to posting the ones I took. Hopefully a few other people can also post so you have more samples. This is right out of the box from the factory, I did nothing but take off the back and snap some pictures.

Photo Album - Imgur

 

No worries. I sent a note and added some key folks to it to increase visibility. I think things will heat up on our end pretty soon.

 

Here's a picture of the heatsink from mine. I received the laptop last week (brand new, not from the outlet) and it will throttle at 77c

 

Almost all 580m users will have that heatsink. And a lot of people will still get the 77C throttle.

 

Woo-Hoo!!! I hope you're right, I'm on my second one and having two complete rebuilds on this second one I'm too scared to tyy any of the 'fixes' in case I break it completly again (Alienware support last time told me I broke it and try to get me to use the pay-as-you-go service - only had it 6 weeks!).
Thanks Bill - this would get you hero status in my book

 

There is one other piece to this throttling issue i've seen in other posts and I think I am experiencing. That is a throttle at a specific power drain.
I use MSI Afterburner for the on-screen display of what is going on with my GPU (no OC'ing) and sometimes i will see a drop to 75Mhz when a game loads and return to 650 when (eventually!) the game finishes loading. Usually with Rage but also with Assasins Creed and Batman Arkham City. Temps are low (58-60) and it's not consistent but the second the HD stops accessing clock speeds return...

 

Why does the 580M run so much hotter than the 6990M? The hottest I have seen my 6990M hit is about 157F or 70C

 

The hottest my 580m has gotten is 65°C It very much varies from system to system and the ambient temperatures.

Btw, according to my logs from hwinfo64, my 580m doesn't throttle at 63-65°C in games, but it does throttle at 63° in furmark??? Is hwinfo64 wrong, or is there something else going on? I almost always play with fraps on and never noticed weird fps drops.

 

Your games don't throttle your card because it won't stress the voltage as much as Furmark. As we all know the 63C throttle is a power throttle so what you are seeing seems natural.

 

hehe ok thx, I wasn't aware it was a power issue

 

Yeah the 63C throttle is a power throttle. Doesn't happen to everyone but happens to everyone with a overvolted vbios that hasn't been throttle fixed so are assuming that it is a power throttle. All 580m's are different so some unlucky users get one with a power hungry card that run into the throttle. You seem to be in the middle. You are good to go

 

The solution is disabling Command center launcher System Configuration (OSD will launch it anyways) then under Services disable Alienware Fusion service. Check under Advanced power settings that nothing changed
I tested it with Stone Giant hwinfo64 3.93 GPU-Z 0.5.9 low manual fan settings, it went up to 82C and Voltage stayed 0.87 and and GPU core clock remained 620 MHz

Command Center version 2.6.10.0 For Alienware M18X
Power settings high performance Both Windows and Fusion (Then disable it)
Bios 08 for Sata 3 performance
VBios 70.24.2B.00.06


M17X R3 580M I7 2760 Crucial M4 128

 

Really?!?! This needs to be verified but looks interesting indeed!

Edit: Just to be clearer, the best way to do this is to type msconfig in the start menu, navigate to services, uncheck Alienware Fusion Service and restart. Please check this out. We may have found the culprit.

 

I've never installed the Command Center, yet I still have the throttle? I may have some new things to try out.

 

I am going to laugh if it turns out to be that service.

 

msconfig startup for Command center Launcher but for Fusion only under services properties disable

With Premiere CS5.5 after adding the 580M for MPE it reached 85C (15 minutes) 620 MHz 0.87 (MPE eats up all the resources available)

When i ran FurMark 1.9.2 it goes to 570 MHz and 0.85 even before reaching 60C






Alienware M17X 3D GTX 580M HyperX 16GB 1600 MHz Crucial M4 128 + 500

 

Clean install W7 SP1 w/ABR (not the THE SP1 from Dell)
Drivers Intel latest Chipset -overall command
Management Engine Interface driver from the M18x

 

didn't work for me, Had to work to hit the throttle(by covering vents and such), but it still down-clocked my card. Not a big deal for me personally as I have never hit the throttle doing anything, but it would be nice to prevent it from happening at all for the 580m users.

 

I had already raised an issue with Dell Tech Support before coming across this thread. Treating this as a generic overheating issue, Dell technicians have replaced both CPU & GPU heatsinks & fans, along with the GPU itself (2 separate visits, 'cause they brought the CPU heatsink and the GPU fan the first time around ). Since the issue keeps creeping up after the GPU hits 78°C, they are going to replace the motherboard next. I am tempted to cancel the service call as the issue isn't going to disappear on changing the parts. Any suggestions?

There was talk that new heatsinks had been introduced post-Nov 2011, that alleviated this issue. I got my laptop around the start of Nov 2011. While I could not get a snap of the new GPU heatsink that they put in today, can somebody confirm if this heatsink that they removed from the laptop, is the newer or the older version?





Also, @Dell-Bill_B, why can't Dell simply change the single-fin backplate design to a double-fin, one covering both sets of memory chips on the GPU? Is it that difficult to change the part, especially since it is already a part of the M18x GPU design (correct me, if I'm mistaken)? Isn't that also an economical choice for Dell, considering that the alternative is successively replacing perfectly-normal laptop parts, amounting to Rs. 45,282 so far (roughly $1000, I am using Dell's product replacement receipts to compute this value)?

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

 

That is the correct 580m heatsink. No point in changing mobos. The problem lies within the system bios.

 

Alright, I think I'll cancel the call then.

 

I got my M17x R3 about 2 weeks ago and have repasted the CPU and GPU with M-4. It did definitely decrease my temps but i still get the throttle at 78 degrees, i have never hit that in a game only when using furmark. I did do the A08 bios flash for the fix and so i could get my 6Gb/s speed. When i had the A09 my Samsung PM830 SSD would stay at 3Gb/s. I am worried if they do fix the throttling the drive speed will still be at the 3Gb/s once again. I saw widzu post something about this also. my ambient temperatures are pretty cool considering i live in north Dakota and its ridiculously cold.

 

The issue has been brought up with the rep, but we are not really sure what will happen. There is also the fact the some people are getting 830's, while others are getting 810's, and for thus of us that got the 830 the 810 is still listed on our warranties as the replacement drive.

It seems there is some confusion as to which drive Dell wants to use and at what speed. Maybe Bill will respond on this issue at some point, as it is a bit of a mystery at the moment.

Anyway, as to your problem, just uninstall Furmark. It is a completely useless application, and calling it a "power virus" (as Dell does) is not really incorrect. It breaks assumptions that graphics card manufacturers make in an effort to produce ridiculous levels of heat and power draw. It has no bearing on the real world, and can actually kill your 580m. If you are fine while gaming, then you are fine. If you want to test temps things like the Heaven benchmark are pretty good, since it will run at 100% GPU usage and show you your max temp under real world conditions.

I would also just flash back to A08, A09 adds nothing of worth.

 

Thanks YodaGoneMad, i have been reading a lot and have decided to uninstall furmark because of other issues i have read. i will try heaven benchmark out. I have been running A08 ever since mainly because of the 6Gb/s, might as well take advantage of the speed while i can and hope i never have a problem with it because of the 810 on the warranty. is there any other benefit of the modded A04?

 

I am in process of transitioning to a new role that unfortunately leaves me no time to deal with Alienware platforms. I have tapped Federico from the Alienware L3 team to take over for me. He's a good guy who's enthusiastically learning the ropes here on NBR. I'll let him know about this thread and see if he can respond.

 

Sad to see you go Bill, thanks for your help so far.

 

Sorry to see you go, I really appreciate all of the help and information you've given us. Best of luck in your new role.

 

Is Federico watching this thread now? If so, any update on the throttle fix?

 

Hi blink_c,

I' aware about this issue and I'm trying to find out the best way we can help you and any customer affected. I will keep you all posted soon with the last update. Please feel free to send me a message with any question or inquiry you may have, I will be more than glad to assist you.

Regards,

Fede

 

Awesome, thanks for the quick response! I'm hope is for Dell to put out an updated BIOS that fixes both the 78 degree throttle and the SATA speed issues. I sure others will be happy to hear it's still being worked on to

 

thank you for the quick reply , we all are waiting for the fix

 

any news? I have a 560m but I have the same problems and I hope this'll be fixed.

I have tested it with Tribes:Ascend(free to play shooter but graphics are good). When I play with max. settings FPS are around 30-40 and drop to 10 and on low settings 55-70fps and drop to 15.
The temperature at that moment is 73°C (if I can trust this GPU meter: http://addgadget.com/gpu_meter/). If I'm just browsing the internet or do similar things the temperature of the card is around 58°C.
I don't want to open my laptop because I don't want to lose the warranty. I bought it in october 2011 and my BIOS should be A08 (I never updated or done anything)

It seems the limit for the 560m is lower than for the 580m and that sucks because my room temperature is around 20°C in winter and around 25°C in summer. I put my laptop in approx. 1cm high wood bricks and that helped a lot. Without them I can't even play UT3 without FPS drops.

Would it help or damage if I cleaned the fans with a compressor?

 

I've been running on a clean installation for a little over a month and haven't had noticeable recurrence of the lag -- then again, I have rarely run DPC Latency Checker much at all since.

When I re-installed the OS, I skipped several components, whatever that wasn't absolutely necessary or if alternatives to those from Dell Support existed. For example, I didn't install Intel Rapid Storage Technology and instead of the Bigfoot Networks drivers I'm running generic Atheros drivers. Also using Nvidia 295.73 WHQL drivers. Can't say which contributes most to the lag if at all, but I'm thinking the graphics drivers.

Now I have other problems to contend with, but I think (hope?) it has nothing to do with my laptop but my router/WIFI network (very crowded in my neighborhood) - massive latencies, frequent pauses and connection resets. Can't switch to 5GHz band or I'll break the connection to my iPhone 4 but I guess I should just try it for S&Gs. The problems seem to mostly crop up when a house mate is on his computer connected to the router with a wired connection.

 

Why do i get ramdom Stutter when running furmark even playing Rift my fps drops and it my screen freezes for .5sec i just ramdom lag (i currently have a gtx 580m and used the throlling fix nvflash with the 580m.rom)