After the GPU Downclocking... Do we have a CPU one? A call to M15x owners.

Thank you very much about that dhoang. That's extremely helpful.
Well... I was told that they "control" everything you post. Appart from that... I can't really use it as I'm in Europe (AW knows we are more demanding... ).

Keep us posted regarding that thread then.
Thanks for your help.

 

Will do Milcs. I'll occasionally drop by the forums and ask for an update. And keep you guys updated.

 

thanks guys for all your testing. I really wish AW would do something about this but NO ONE WILL ANSWER US!!!! I think after dealing with this issue I'll never buy AW again!!! by the way we have a new record for time it takes to downclock!

 

Yea I replaced my paste, takes a good 6 minutes almost. Oh well!

 

This is still not fixed by AW. Holy S$@T that is appalling. AW needs a good class action suite and public humiliation. I would just return the system if I were you guys. You have more then enough evidence to prove your system are defective, and demand a full refund. I think the time has long passed for AW to offer a reasonable repair time frame for you guys. I wonder if anyone has even tried contacting Intel about this. I have ahd to deal with them before in the past, and they where amazing to deal with. And our work issues were resolved quickly and professionally.

 

Why not give AW a chance to fix up the system by sending it in and let them honor the warranty that you paid for?

 

Dhoang, regarding the extreme CPUs... this issue has been "exposed" by newfie, Taco and so many others. This has been what? 3 months ago. Has anything been done? New laptops being shipped with the extreme processor still have the exact same problem.

Sort to say... they simply can't solve it. But they don't even stop giving the extreme processor as an option.

 

Capitalism at its finest.

 

They haven't sent their m15x's in for AW for a chance to try to take a look at the problem yet.

 

well I think the problem is AW doesn't want us to send the system in for the simple fact that they have been able to reproduce to problem are trying out ways to fix it which they haven't yet. They are saying to themselves why spend our money sending the system in when we can't even fix it ourselves.

 

Sure about this? Send it in and let them see for themselves. I must admit AW is pretty hard-headed at times.

 

well I would but like I said they have been able to reproduce the issue themselves so if they can't fix on their on machines I highly doubt they'll be able to fix it on mine. I keep in touch pretty regularly with AW and they have never given me the option to send it.

EDIT: Just got off the phone with them and they still haven't got a fix for it. and i quote "We still have not gotten a fix for that yet but our mobile group is working on it, please send an e-mail to [email protected] with your info" So there we go.

 

Man, if everyone on this forum with the extreme processor has this same downclocking problem and can reproduce the issue easily how hard can it be for AW. So, AW is working on this issue right?

 

well they been saying that for around 2 months now.

EDIT: Ok so I've e-mailed a corp guy at Intel that I know through work. I'll see what he has to say about all this maybe we can at least stop AW from selling the x9000 on the m15x

 

Has anybody been able to reproduce the downclocking issue with real world applications? Meaning, can anyone do it without actually using a stress test program designed specifically to bring a cpu to it's knees? I was going to read back through the thread and see for myself until I noticed we're on page 42... I suddenly got very tired at the thought.

 

Maybe that's why AW is taking so long to test because they are testing it through real world apps and trying to find downclocking through them instead of these stress tests.

 

Several people experienced this downclocking during WoW, some during video compressing/editing, etc etc...
AW hasn't done a thing because they don't know how to solve it. They have no solution for it and they keep selling computer with the extreme processor even if it overheats and downclocks.

Isn't this unacceptable? Why don't they simply stop selling the extreme processor if they are aware there is a problem with it? You know why? Because that would be assuming in front of the world they do have a problem with their cooling and they do not want to do that.

Marketing and corporativism... Anyway.........

 

yes mine will downclock while converting videos. It will happen after a few min of the cpu being at 100% and Lessa and New had downclocking while playing WoW.

 

Taco, has your downvolting fix solved the problem?

 

well yes it does work thats what I do when I do work like that, but at the same time I shouldn't have to do that. It should work without a work around. Thats why I really pushing for AW to fix this plus its not cool for new people that buy that cpu.

 

I agree with Milcs for one thing. If AW can't fix this issue that happens during real life applications, then it's a better idea not to sell it. But, I think that they are still working on a plausible solution.

 

Yea how about a cooling system that doesn't suck my left testicle. I like that idea. Or having the fan actually ontop of the component might help.

 

I was wondering why my framerate is Wow was periodically plummeting. I loaded HWmonitor/RMClock and low and behold, my X7900 extreme M15x was steadily cycling between a 14x and 6x multiplier once the cpu temp got to 86-88C.

The cooling system just isnt capable, seems a blatant design flaw.

I tried the undervolting experiment, dropped the stock 1.2625v setting down to 1.15v (1.10v was not stable), and presto, my max CPU temps dropped to 71-72C and no cycling down to 6x multiplier.

SO far the fix seems rock steady, but its hardly the job of the users to find this fix. I have tried to give AW the benefit of the doubt here, but after all the bios probs, the smoking vid card temps, and now the downclocking CPU issues, the m15x fails at thermal design. It is a flawed system as shipped, perhaps there is a hardware fix that can increase the cooling to the CPU, but other than that, software kludges (even if this undervolting fix holds up) is just not a valid long term solution.

A redesigned cooling system is the only solution that would be acceptable.

For those willing to try, I heartily recommend the undervolting "fix" suggested here. Its just sad the fix comes from the user base and not Alienware....

 

so now we have someone else with downclocking! Thank you for testing and now we can add you to the list. I agree with you 100% that the cooling system is flawed and they lacked in testing. Also glad to hear the undervolting helped you out.

 

The more I read, the more it makes me sick... Added him to the list.

9 out of 10 users with extreme processors report downclock. Does anyone else needs further proof? Does Alienware needs more evidence?

How can AW justify their inability to solve this? How is it possible that they just ignore this horrific problem?

 

I think they have all the evidence they need. THEY NEED TO DO SOMETHING ABOUT IT!

 

hey guys want a tip on how to downclock your extreme CPU faster! Switch to integrated graphics and watch as the temps rise super fast!! I'm assuming this is because the integrated graphics card and the cpu are using the same heatsink. Don't worry guys if you undervolt even when using integrated graphics it still won't downclock assuming you did it right.

 

sorry taco ... but also this is not happening on my m15x

 

i'll trade you! LOL JK well you are one of the very few. Whats funny is that you upgraded yourself so it seems that we need you to repair our cpus and not AW LOL

 

Im not seeing it either taco.......im sure there is some kinda fix to this problem...there has to be some kinda of solution to the situation....I dont see why they would make some cpu's diffrently than they made others. Are other laptop makers with the same cpu have these problems??
-J.B.

 

Are the people "not seing it" actually running prime95 and RMclock to follow the core speeds (like suggested in the 1st post)? Because I am intrigued with this "not seeing it".
Could you post us screens of your CPU's being stressed with RMclock, Thermofan and HWmonitor so that we can understand what is happening?
That would be really helpful...

 

Guys, count me out of the non downclocking x9000 group. I originally didn't get downclocking when I used Prime95, but I decided to monitor it while ripping a DVD to HDD. Posted screenshot of RM Clock with settings on Mobile. The indicator at the bottom right of the taskbar fluctuates between 2.8 Ghz and 1.2 GHz, a trend I saw in CPU-z, but decided to check out in RMClock to see...

 

Interestingly enough, my CPU load and OS Load bars go sometimes go over 100%- pics to come

 

Hi evil.
I am a bit unsure whether that's true downclock...
Would you mind getting HWmonitor and Thermofan running to follow temperatures. If your temperatures are going anywhere above 92ºC, I think that you can consider you have downclocking. If not... I am not sure whether that's only a normal CPU behaviour.

I think that you really have to do these "tests" plugged. If prime95 does not do it for you (with CPU stress test- small FTTs), I don't think your CPU is downclocking.

I don't know... do some more tests and let us know how it goes.

Good luck,
M.

 

Pics from RMClock below

 

/From page 20
//just wanted to make sure that I was counted

 

There you go... Nobody wants to leave you out!

 

Here are my results with HWMonitor, ThermoFan, and RMClock. One of the pics shows only Core 0 in RM Clock at a certain time, the other 2 are taken at nearly the same time, with Core 0 in RMClock in one pic, and Core 1 in RMClock in the other. Is what I have downclocking??

 

No... that's not downclocking. That is a normal feature from the processor which is called speedstepping.
Do the following, get Prime95 and Run "CPU stress test- Small FTTs". This will drive your CPU to be always at max clocks and will put the CPU at working constantly at max speed.
Use the same "trio" of programs to follow temperatures and clocks. Run Prime95 for 30 minutes. If everything is fine, you will see that the Clocks will be always on the top and never come down. If after sometime you start seeing it jumping up and down, then you do have downclock (by this point, the temperature of your CPU (according to Thermofan) should be around 92-93ºC).

Hope that all makes sense...
Let us know how it goes.

 

THANKS man!

I feel spechal now!

 

So I'm guessing this has been a problem mainly for the extreme processors? From what I can tell the majority of the T9300's seem stable, which is refreshing.

 

I'm sure that you will be perfectly fine. Given you are not going for the Extreme CPU and you are going for the 8700, you will probably won't notice any of these "overheating" problems.

Good luck

 

evil the way I test it is I'll run prime95 but also run thermofan and rmclock. Wait to see if your pc hits 93c if does and then drops down really quick then you have downclocking. From what it looks like to me is that it may be speed step.

 

taco,

I think I do have CPU Downclocking on my X9000. I have attached 4 pics with RM Clock, HWMonitor ThermoFan and Prime 95 in background. Two of the pics are without the CPU being stressed by Prime 95, one showing CPU 0 and one showing CPU 1 in RM Clock. The other two are after the CPU has been stressed for about 25 minutes with Prime95 (small FFT) with one showing CPU 0 in RM Clock and one showing CPU 1 in RM Clock. The conclusion of this test seems pretty obvious....

 

There you go... that's is definately (NO DOUBT) the CPU downclock we are all talking about.

I would go a step further and say that EVERYONE has this issue. Just that some people do not use the exact same programs to monitor it and, therefore, report a "non-downclocking" CPU. Like evilwill32 for example. In the beggining he reported no downclock... He just didn't look in the right direction.

So... Here is another solid and big statement:


Nearly 100% of the M15x with extreme processors have CPU overheating and downclocking!

 

Here we go ... is this OK now @milcs

 

Now it is very OK!

As I already said, I am not going to put you up on the list as what you've done actually voids the warranty.
But maybe AW should learn from you and the other guy (which name I do not remember).

Anyway... You still seem to be running pretty close to the "critical" temperature, as your ACPI is at 89ºC. Would you mind running the Prime95 for at least one hour (or a couple)? Just to make sure your colling can keep that temperature for above 30 minutes.

 

but the temp stoped moving, as you can see between the first screen (15min) and the second screen (25min) ... but sure, I´ll run the test tonight when I go to bed

And anyway, I didn´t lose my warranty, I just have no warranty on the X9000!

My Geforce 8800M was faulty, and the AW support knew the change from T7100 to X9000, and I still got a new 8800M ... It´s the same with Dell, you can do with you laptop/desktop whatever you want, as long as it´s not YOUR fault if something breaks down!

 

Thanks Markus.
Regarding the "stability", with my CPU, I had it running under 90C for almost 2 hours. Close to the 2 hours, it broke 90C and stabilized at 92ºC. I run the test for a total of 4 hours.

That's the reason I am asking you that...

 

hmm .... listen, just to be realistic, which game need the FULL CPU power over 4 hours?

I mean, I´m sure my m15x will pass this test as well, but still, thats just looking for some faults ... I´m sure, that with many other high spec laptops the same thing happens, if you just look for some faults!

Still .... I really like to say it again ....

I LOVE MY m15x .... ^^