Hey yomama, I have read alot of threads in this forum regarding the M1530 and I noticed that you have the lowest temps ever posted. You also have insane overclocks. Please note that at the end of this post I have a request for you.
I notice alot of confusion and conjecture about so called 'faulty' GPU's. I just got my M1530 in November of this month. It shipped with bios A11. I thought that I had gotten one of the 'faulty' GPU's as when I tested it with 3dmark05 I saw the obvious overheat/downclock occuring. When 3dmark05 got to firefly forrest there were points where I knew that if anything was wrong thermally, the FPS would hit the dirt so I could anticipate the exact moments of downclock. I measured temps and like clockwork the GPU would downclock at anything over 85c and there were times when I saw max GPU temps of 90c and higher!
When the GPU would downclock the FPS would drop to 5 FPS, then when temps got back to below 85c the FPS would kick back up. I think that anyone else would have just concluded that they got a 'bad chip' and requested hardware exchange from Dell. But I thought otherwise.
Here's how I fixed it:
1. Flashed the bios back to A09. Result - fans kicked on 5c earlier and max temps never passed 90c.
2. Removed heatsink and cleaned off the crud that Dell refers to as heatsink grease and replaced it with Dynex silver paste. Result - max temps never passed 83c.
3. Used RMclock to undervolt the CPU by 150 milivolts at the highest multiplier; my settings(6x-1.05, 7x-1.062, 8x-1.075, 9x-1.087, 10x-1.100).
Result - max GPU temps never passed 76c, and most of the time max at no more than 74c even while playing Crysis.
Total result - a completely stable GPU that never downclocks anymore. I tried overclocking (previously even a modest overclock like 575/750 would bring the GPU to it's knees) and now I can clock 600/800 and it runs like a champ. This GPU that most would have blamed for being faulty is now flying through Crysis at med/low settings and never stutters, not even once. This thing also flies through Area 51 at default settings and 1440x900 res. And with overclocking the GPU max never exceeds 79c, most of the time staying at a max of 77.
Here are the benchmarks with a properly functioning GPU:
(All taken with driver 180.84 in Windows XP)
3dmark 2005 1024x768 no OC = 8780
3dmark 2005 1024x768 600/800 = 9842
3dmark 2006 1024x768 no OC = 5371
3dmark 2006 1024x768 600/800 = 6171
I am almost finished with this conquest, however one thing remains. I have the 'bad' revison of the heatsink (one with aluminum over the northbridge and GPU). I believe that you have the good one (with copper over each chip).
Could you do me a big favor and take a picture of that thing and post it here?
I can get the one with all copper on it and I think that your heatsink revison is the key as to why you have supreme temperatures. BTW - I appreciate all of the info that you have provided in your posts.
Thanks
-
-
You might try sending a personal message (PM) to him so he knows that you want to talk to him specifically. To do that, click on the link "User CP" on the top left to go to the Control Panel, and then click the link "Send New Message" on the left.
It might be a little hard to find these links because there are lots of link menus on these forums. If you can't find them, use the "find" feature from your browser. They are spelled exactly as I wrote them. -
you need 15 posts to PM
-
Thanks for the tips guys. I'm trying to build up 15 posts. I already replied (and hopefully helped someone) to one thread. Only 14 more to go!
-
Hey slowdown117, here's a pic I found of the copper heatsink, however like I said in my thread, I thought it was the old one? (even though copper is meant to be a better conductor)
Oh and I found this thread where someone comapares the heasink from their old and new XPS M1530.
http://forum.notebookreview.com/showthread.php?t=308359&highlight=m1530+heatsink
When you compare the pics it looks like there's 3 different revisions now!
I'm pretty certain the one in the attached pic is the very first because that pic was from a review from a site just after the XPS M1530 came out. I think the 2nd one is the aluminum one and I think the 3rd one is the copper one that guy got as a replacement. So looks like Dell went from copper>aluminum>back to copper? -
Alright bboy1, that's tight work! I like how you deduced the timeframe of the three different revisions because that tells us quite a bit. Provided your time line is accurate it seems very strange indeed that Dell would first use what looks like a pretty good design - then go to aluminum, which should not perform as well as copper - then revert back to copper, but this time they got cheap and used smaller plates over the chips. The aluminum one should be the worst because for 2 reasons that I can think of off hand. One is a bit obvious, and that is the thermal conductivity : aluminum = 237 W/m-K and copper = 400 W/m-K (give or take a few). But the second reason that I think of is the bond between the plates and the heat pipe. The bon in the case of copper is going to be best because they are like metals.
I also would deduct by your data, that the first revison could not have caused problems, because why would dell bring in the third revison, based on the first revision, if the first revision was not mechanically sound?
I am fairly confident that we together have figured this out (I will try to verify the timeline that you have come up with just to be sure). It looks like Dell's first version was actually a good design. Then they tried to cut cost by bringing in the aluminum revision. That one created heat problems so they went back to copper while still trying to be cheap skates.
It seems to fit together like a puzzle. To add to this, my aluminum revision heatsink was paired with an 8600m gt manufactured on June 9, 2007. We should conclude that in June of 2007, Dell was using revision 2 during that time and that seems to match the emergence of the accusations that Nvidia had faulty GPU's. I maintain my stance that the only thing faulty are these vendor's cooling soultions. I will dig further.
Thanks for your work
Edit: I think that yomamasfavourite can help us ice this case. He has supreme temps. I am betting that he has revison 1 heatsink. -
I'm going to buy the copper heatsink and help this case. It will be very helpful to see the temperature results on the same machine, same heatsink grease, and a comparison between the aluminum plate heatsink and the copper plate heatsink which we believe to be revision one.
Edit: I have purchased the all copper heatsink. When it arrives I will conduct testing and post the results. This should solve the mystery of which heatsink is best. -
-
anyways, i have applied AS5 about 3 times now, because it seems i didnt get the right amount on the chip, it has lowered my temps to max 79c on my GPU, i hope i can see those temps fall when the house gets less HOT
EDIT: yoma's temperatures are a mystery... i know, ive seen his temps... and i really dont have any idea of how he gets those low temperatures, maybe placing your computer over snow ? using a custom AC outlet from home that fits into the air intake on your fan ? (lol i did this once hahahaha, still an unfinished project, since tunrning the AC on winter is like inviting the penguins for a party ) OR! simply living in a cold place and having a well assembled computer
Hey yomamasfavourite, I have a request
Discussion in 'Dell XPS and Studio XPS' started by slowdown117, Dec 29, 2008.