M17x R2 full internal cooling mod - huge improvement!

You can put the Shin-Etsu beside a lightbulb for a few minutes, it will make it easier to apply. The MX-3 is definitely a good choice. Between all three you probably won't notice much of a difference ~2C.

As for the CPU I think Joker on Tech|Inferno did a CPU retention mod, but I think it's riskier. Try buying some thermals pads off of frozenCPU for the CPU.

 

Would I be out of line if I asked if I could do this with the r3 and the 460? I'll probably just repast and get .5mm pads. I hope those are thick enough.

 

Repasting and pads, yes. Retention mod, no. Not sure what thickness you will need, I would say .5mm, and if you need it to be thicker you can just stack them. I'm basing this off of the R2, 4870M = 1mm, 5870M = .5mm

It's always a good idea to replace the stock paste/pads if you know what you are doing of course, you can drop temps significantly.

 

Thanks a ton for the tip, in that case I'll go with Shin-Etsu on the RAM and MX3 on the dies I guess, unless someone tells me I'm stupid to do so.

As for CPU retention mod that wasn't my plan to begin with (talk about a recipe for disaster), I was just thinking about doing a simple repaste and my question was:
If I do it, will the lil' bstard stop acting like a hairdryer? Or will the retention mod be enough to quiet it down?

 

Why would anyone put a piece of thermal pad on the CPU die if paste applied correctly does such a great job?

As for thermal paste brands, it all comes to preference these days all that matters is quantity(less is more ) and a proper 'job', screw your heatsink while monitoring your temps at the same time at a constant load.

Now for thermal pads I'll repeat myself saying that one should first measure/know the original thickness of the pads and buy the same before replacing them with whatever size if no modification was done to the heatsink/card or you might end up with some components not touching the heatsink resulting into a failed job.

Cheers

 

I used AS5 on the GPU and MX-3 on the RAM...seems to work fine.

I didnt bother with the CPU retention mod as I wasn't concerned with the temps...I think mine are between 60 and 70 under heavy load?

What kind of temp drop can be expected from the CPU retention mod?

 

If you replace the memory pads with paste you have to do the retention mod. I used just Fujipoly 1.0mm pads and some tx-4 paste and my temps are in the 70's rare spike to 80. No retention clip mod and I am running my 5870's OC'd at 825/1200. Honestly though people have cracked their GPU dies from doing the retention mod that is why I omitted it.

 

I have some AS5 in syringe but I've been told it's not on the same level as the MX3 and the Shin-Etsu.
Arctic Silver said that with time their paste can be electrically conductive and that is a deal beaker for me as I don't plan on doing this stuff every week end.
My only concern with the CPU is the noise, I have roughly the same temps as you and I don't want to try the retention mod on the CPU, it's way too risky.
I was wondering if a simple repaste would help with noise.

If I tear my laptop down I'll go all the way and do the retention clip mod, the results are better.
And yes I know that in order to replace the pads I have to do the retention clip mod...

 

Woah....thats a nice overclock, and your saying that your 5870's rarely go in to the 80's? Including memory temps?

Thats very impressive! Do you have any screenshots running furmark or anything like that?

Serious? That sucks.... I have used Arctic Silver for years and never had an issue, so i wonder what time frame they are talking about...

 

Cygoris, how did you go with the re-paste?

I am interested in your CPU temps with only the re-paste and not the retention clip mod.

So let us know when you have some results!!

 

"While much safer than electrically conductive silver and copper greases, Arctic Silver 5 should be kept away from electrical traces, pins, and leads. While it is not
electrically conductive, the compound is very slightly capacitive and could potentially cause problems if it bridges two close-proximity electrical paths"

Arctic Silver Incorporated - Arctic Silver 5

Meaning it's OK for the die but not so good for the RAM in our case.


As for the paste becoming electrically conductive I don't know the time frame but I've heard that the paste itself is not electrically conductive, but the oxyde it produces with time is.
Now how long before the paste oxydes? I have no clue, but I know I won't bet my M17x on it.

Right now I'm testing which paste is better on the GPUs and the results are weird, the Shin_Etsu is WAY better than the AS5 and MX3: 4°C less for a thermal paste is insane (it's hard to spread and hard to clean).
But, when I plug the Cooler Master Notepal U3 I just bought, the gap gets narrower and the MX3 pulls ahead for some reason... why?
Once I'm done with the GPUs I'll do the CPU, I will probably make a long post about it with a retention mod guide with pictures and some charts.
By the way guys, I have more thermal pastes that I can test, if you want to know how your compound performs just ask and, if I have it, I will put it to the test.

 

Ah ok, well i used MX-3 on the RAM and AS5 on the GPU...so here's hoping it is ok!

Interesting results........the cooler would only make the heatsinks more effective by providing more flow, and the effectiveness of the heatsink all comes down to the paste...so the Shin_Etsu should stay ahead with or without the cooler.

Not sure man :/

 

I'm done, but geez that was a nightmare.
The retention clip mod is blindingly difficult, it's like performing eye surgery with a baseball bat, on a train.
Even with the right tools it took forever and I had to try 7 thermal compounds.

Here are the results without cooler and after 2 hours playing Starcraft 2, a 3dMark2011 run and 6 minutes of Furmark:



This is with the Shin-Etsu X23-7783D on the die AND Ram, which is by far the best compound I've tried so far.

Here are the thermal pastes I've tested:

-Arctic Silver 5
-Arctic Cooling MX3
-Arctic Cooling MX4
-Arctic Ceramic
-Noctua NT-H1
-Zalman ZM-STG1.
-Shin-Etsu X23-7783D

Keep in mind that the tests were not done in a controlled environment and that my skills with thermal compound are average.
I own a CoolerMaster Notepal U3 but I decided to turn it off as the results would be biased otherwise.

-The Zalman and the Noctua were nowhere near the others, the Shin-Etsu outperformed both by 8°C on average, you might as well keep the pads.

-The MX3, MX4, AS5 are roughly on the same level, on average the Shin-Etsu is 4/5°C cooler.

-The big surprise came with the Arctic Ceramic, not only it was easy to spread, the results were impressive to be honest, only 3°C hotter.

Last but not least, the Shin-Etsu is the winner, in game the temps never went past 65°C and only Furmark took it to 74°C when the others were beyond 80°C.
But it is the most expensive compound here and the most difficult to spread.

If you're not obsessed with silence go for the MX3 or Arctic Ceramic but if you want silence at all cost, go for the Shin-Etsu and don't forget to preheat the paste a little.

I won't even touch the CPU, my temps are good enough in my opinion.

 

Nice post!

It is interesting to see the differences that thermal compound can make!

I might go ahead with the CPU re-paste when I get bored and post some temps. I just might have to purchase some Shin_Etsu for the occasion

Also, if I feel up to it, I may re-paste the 5870's and post some before and after temps. I will be interested in comparing both of our results!!

Thanks for all the hard work!

PS: have you noticed a big decrease in fan activity?

Another thing, off topic, but have you guys noticed your screen flickering when set to Extended or Clone since the overclock? I have googled it but there are a lot of different theories....something about the 2D clock??

 

I've just ordered an M17x-R3; anyone have any tips or advice for cooling on this unit?

I've been looking at laptop solutions for cooling. I currently have a hard plastic table type pillow for extended usage away from the desk but am concerned about heat for long periods of time (3-4hrs).

I've seen a few cooling stations but not much for the "on lap" style solutions. Anyone have suggestions, or experiences with these type items for cooling? Or is it even needed, could I still use the current hard surface pillow I use now?

 

Before even talking about a cooling solution check your temps and tell us what you see (with HWmonitor or any other software).
The R3 is usually pretty cool.

 

I hope someone will be able to answer my question:

I've attempted the retention mod on my R1 with 280m, changed pads and old paste with mx3 and gelid gc extreme. I didn't see much improvement.

So if anybody else tried, would it be possible to know your results (with furmark options, and ambiant temps).

Also, I couldn't make the sli work under furmark even with fullscreen enabled, what's the deal with that?

 

How did you apply the Shin-etsu? With results like this I am thinking about driving down to a local electronics store and buying this same paste, how much do I need?

Im about to add a 2nd HD5870 and do this cooling mod. I am doing the retention clip mod as well, the pictures in this write up dont really illustrate what needs to be done, Im hoping that once I am looking at it that it will dawn on me.

Any tips?

Did you have some problems on ebay? I noticed negative feedback from someone with a similar user name for a seller of the Shin Etsu X23 on there. Where did you get yours?

What application method did you use, drop, line, spread etc? How did you heat it up?

Thanks.

 

@vulcan, responded to your thread here-
http://forum.notebookreview.com/har...oling-mod-indigo-xtreme-vs-shin-etsu-x23.html

moved for a larger viewing audience. Not sure if that indigo will fit correctly (?)....

 

Here are my temps using TX-4 and Fujipoly pads without the retention clip mod. After gaming and running furmark





Uploaded with ImageShack.us

 

Nice temps! Specially seeing that you didn't do the retention mod!

Question, Why is your memory clock on GPU #1 at 3175Mhz???? And the second GPU's memory clock @ 1200Mhz?? Are these correct?!?!?

I am going to go a little off topic if you all don't mind, feel free to flame away

So i have been experimenting with overclocking my i7-740QM and I have been getting good results (nearly 3.4Ghz on one core, or 3.0Ghz on two cores with stock cooling. BCLK = 154Mhz). I am now VERY keen to do a re-paste on the processor.

I have been following: [PHOTOGUIDE] Cleaning/Replacing Thermal Compunds (GPU+CPU) on M17x-R2

I have noticed that there is a thermal pad which the guy replaced AS WELL as re-pasting the CPU.... but there is no description as to what it is for or how thick it is...

Anyone have some experience on this?

Whats this about someone dodgy selling on eBay?? I bought from a guy called kbthingamajigs....haven't received it yet....been a bit over week but I thought that's to be expected cause its coming from the US.......now you got me worried!!!

 

I OC my 5870's to 825/1200 each if they where stock temps would be even lower. The pad near the CPU I used .5MM Fujipoly Pads

 

Wow, nice overclock. What is your general room temperature?

Did you re-paste your CPU as well as the GPU's?

 

my ambient temp is in the range of 76-80' F, and yes I repasted my CPU as well and that is overclocked with a 75mv raise in voltage.

 

Ah cool, cause I noticed your CPU temps were a little high if they were stock.

What CPU do you have and what clock do you run?

 

I have the CPU in my sig a 720QM and my clocks are 1.81 ghz, 3.2 ghz 1 core

Stock BCLK is 132 I go to 166 if I put more juice 150mv i can go upto 167/8

 

Sorry, you do indeed

Not bad, not bad at all Your giving me even more of a reason to do the re-paste!!

Thanks heaps for your help with this one, I will post some results when done. Still waiting on my paste, and now gotta order some thermal pads.

 

Thanks Bat Boy, yeah its my understanding that the Indigo Xtreme is for CPU's only, in case anyone else is interested in it. Going with the Shin Etsu X23.

 

I wish such people as Author had Sony VAIO F so I would see same tips and mods for mine laptop...

 

Fujipoly High performance pads get those .5MM

 

Where do you get those thermal pads? And do you buy AS 5,7 from internet shops? I am not from US and thinking to buy AS grease but eBay looks to me not the right place

 

I purchase my pads from FrozenCPU.com | 1.877.243.8266 | World's Largest Selection of PC Modification Supplies, and I use Tuniq Tx-4 thermal paste as it has better performance than AS5 which I also purchase from there.

 

Just to be sure, are you referring to:

Fujipoly Premium System Builder Thermal Pad - 1/4 Sheet - 150 x 100 x 0.5 - Thermal Conductivity 6.0 W/mK - FrozenCPU.com

This is referred to Fujipoly Premium Pad's, not high performance. Just want to make sure.

 

Nice temps, but there's something wrong about them.
What I'm about to say is not kind so I apologize in advance.

Your min temps are very very low, either you live in the north pole or you just had time to boot and play a game.
Trust me on this, if you play a game and run furmark as you said you did, there is no way to have such low minimum temps.
We have a pretty good amount of experience on this forum (and others) about the M17x cooling and I am pretty sure that usually the pads, even if they are new, do not perform on the same level as a full retention clip mod.

Not to be rude or anything but this forum is to help people, I hope you understand that, this is not a competition.
There is no way to sugar coat this, I think your temps are odd at the very best.

If I'm mistaken then I apologize even more, but I think i'm not.

 

You caught me!

No you didn't actually, my system is always on except for when I goto sleep, I put it to sleep too never shut it down.When I took that was after running furmark and playing some League, Now This is not a competition like you said, but what you are seeing is good temps. I am not new to this forum and I have helped lots of people on these forums.

If you think my temps are odd that is fine, but fortunately for me they are not and they are what they are. You have a little better temps with the retention clip mod and that is fine. I personally have always been against that mod and if you actually read the entire thread I have been active in it since the start. Since I have had my 4870's where I showed I could get get good temps with this combo.(but as scook pointed out this is a mute point)The retention mod IMO can be dangerous as a couple people have chipped their GPU dies doing this. My temps when I am just surfing the web and all that is in the mid to low 40's and my max temps are in the mid 70's. I am sorry that you feel that I am trying to pull the wool over your and others eyes, but that is simply not the case.

So basically, All I have ever said was buy some good high performance pads and paste. That the retention clip mod is too risky for just a few degrees.


The Fujipoly Premimum system builder pads are high performance as they have a thermal conductivity of 6.0 w/MK which is pretty high for Pads which are normally 3.0 w/MK.

"We have a pretty good amount of experience on this forum (and others) about the M17x cooling and I am pretty sure that usually the pads, even if they are new, do not perform on the same level as a full retention clip mod."

I also liked how you excluded me from the group of Alienware Owners/ forum members and of the people Knowledgeable about the M17x Cooling, as if I have no Idea of that. I also never said that they will perform on the same exact level I said that I can achieve SIMILAR results without doing the retention clip mod and not worrying about cracking my GPU dies.

I won't sugar coat this also, next time that I have somehow made you feel the need to be condescending please don't, I hope you understand that.

 

Just to throw it out there....make sure you are not comparing 5870 temps with 4870 temps.....they are not comparable at all

The retention mod does NOTHING for 4870s, it is purely needed for the 5870s. I verified this personally with my 4870s in my R1 (same card as the R2)

 

I am not, was just saying that I previously did make that comparison. I am talking about my 5870's now.

 

I was questioning your results because I know a lot of people follow this thread looking for informations on this modification and I honestly think your results are misleading.

In the real world your method doesn't produce nearly the same result as a full retention clip mod, the reason is simple: a pad will never be as good as a good thermal paste, however awesome/expensive the pad is and that is not a debate if you know anything about this subject.

I didn't want to hurt your feelings but looking at your aggressive reaction comforts me in the idea that your results are, let's say "miraculous" so you don't get mad again.

My only aim is to let people know that, excluding your "miracle", a full retention clip mod produces far better results, and the die won't crack even if your hand coordination level is on par with a hippo.

And please, refrain from posting any kind of insult, I'm not young enough to participate in this kind of virtual feud.

 

Simply using thinner pads makes the retention mod a lot less necessary for the 5870s....also cracking the die is pretty EASY to do regardless of how coordinated your hands are. I cracked one of mine on my 4870s when trying out the retention mod - poop happens

 

Lol Thanks for the laugh, I am not angry in anyway. You where being condescending and I asked you to not be in a sarcastic and equally condescending manner.

If my reaction to your assumption about my "Miraculous" results has made one thing more clear it's that you're speculating about my results.(show me your comparison using the same brand pads and paste, show me your real world results.) I am using .5MM pads which are thinner than the stock dell pads, therefore, making the gap smaller which is what the retention mod does just with thinner pads not as much, therefore my temps are slightly higher than yours, right? I have said that I personally IMO(that's In My Opinion) did not feel that the retention clip mod was worth the RISK of cracking my GPU die's. When I achieved decent results with the method that I used. What is more helpful than a statement like that to people that may want to know the risk and an alternative which might not yield the same results, but are not that far off.

Besides the fact that your omnipotence is a farse, as you where not here to observe me in what testing methods I have used. Your once again speculating that my results are "Miraculous" instead of asking me to show further proof of this you decided for some reason to take it personal and be condescending in your response. Let us now put this petty argument to rest and move on with our lives.

Sincerely,

Mkelliny

Congrats on your M18x btw Scook! You going to put that in the freezer also ?

 

Not sure yet. We will see if that will be necessary or not

 

I jumped right to the end of your rant without reading anything but this I couldn't agree more.

4870s are not 5870s, you can screw the heatsink of a 5870 to the max with the clip mod and it won't break.

 

Screwing the gpu screw blindly down can cause damage to your GPU. Use caution when doing the retention clip mod.

 

The problem is screwing down one at a time instead of constantly rotating to keep the heatplate somewhat level. Doing this can chip off a corner of the GPU die which in the case of my 4870, rendered it non functional, this would be easier to do on a 5870 as it does not have the metal band around the die like the 4870s do so a more extreme angle can be achieved

 

Let's make it short, show me ONE person who broke the die of a 5870 (not the 4870) from every forum you want, not just NBR, everywhere, and your point will make sense.
You seem to think this mod is very risky, please show me why.
And if you can't, why do you still think it is?

I'm a member of many forums and I've never heard about anything like it.

The method Mkelliny used is best suited for those who are not familiar with laptop components (not saying he is, just in case he goes off again), but for those who feel like they can do it the retention clip mod usually produces better results.

You wrote this:

 

I know what I wrote, the GPU worked fine putting the retention mod in place, it was removing it that chipped the die on one of my cards.

And as noted, it made no difference in temps for my 4870s. I was not paying attention to a lot of the 5870 posts regarding breaking a GPU so am not a good source for that but I do remember a warning due to the possibility of sustaining damage.

 

Since we want to pull up quotes the maker of the mod says: This should sum up why we think that you can crack the GPU die.

 

@scook9

Well just so you know, I did the research for you, not a single 5870's die has been broken on this thread or anywhere else for that matter, not one, and something tells me that not everyone who tested this mod on 5870s were as talented as you are or as your friend Mkelliny.

That's why I think the information you and your friend are giving can be misleading, I truly didn't want to hurt anyone's feelings but you are talking about a risk that does not exist (if there were no car crashes at all, none, would you say that driving is dangerous?) and Mkelliny says a pad with no retention clip mod is comparable to thermal paste + retention clip, and that simply isn't true for an overwhelming majority of users.

Edit: Mkelliny, you realize you quoted the first post of this thread from last year when nobody knew about retention clip mod and the TS didn't want to have some problems?
I understand you don't like me and you want to prove me wrong, but you are throwing logic and common sense out just to achieve this goal, this is ridiculous.
But please make my day and find someone who broke a 5870 die.

 

Dude I was trying to avoid saying this, but you are acting like a DBAG! No one said anything about your decision to do the mod at all.Why is it such a big deal that I got good results a different way? I chose not to do it because I have 3 years of warranty left and don't want to risk cracking my gpu die. You talking S*** is not helping your cause it just makes you seem like a child. God forbid that someone has achieved results using something other than this mod End of the world you say. Then you attack me and scook for defending our points. If anyones feelings have been hurt it's you, and stop referring to me as "your Mkelliny" Just refer to me as noob owner of the M17x R2 oh wise one!


Also to insult scook or me for accidentally breaking his GPU/ getting good results without doing the clip mod by calling us "Talented" is not really an insult like I think you were going for.

EDIT: nice try editing your OP. Lol

Mkelliny, you realize you quoted the first post of this thread from last year when nobody knew about retention clip mod and the TS didn't want to have some problems?
I understand you don't like me and you want to prove me wrong, but you are throwing logic and common sense out just to achieve this goal, this is ridiculous.
But please make my day and find someone who broke a 5870 die.

This is asinine I quoted the OP who made this MOD! A year ago everyone knew about the RCM it was not a secret and many tried. Look kid it's nothing personal you attacked me first for achieving something a different way. Look for yourself if you found a cracked 5870 great if not....great! NO ONE CARES BUT YOU! let it go now like I tried to do last page. If you want the last word feel free. I am done!

 

Pretty sure this is not about who likes who or who is smarter than who. Being a jerk will get nothing accomplished, I promise. I can assure I have no special talents over anyone else, I simply shared my experience which is something others may have opted not to do.

We are not trying to mislead anyone, we are simply advising against a risk. Just because no one posted about breaking a GPU does not mean it did not happen. Additionally....this is one of the worst examples I have ever heard:
"if there were no car crashes at all, none, would you say that driving is dangerous?"

I absolutely would say that driving is dangerous, only a fool would say otherwise. You put too much trust in other people it would seem. Occur to you that people may not have mentioned if they broke a GPU simply out of shame? That is a simple human emotion....

Back to the root of this entire argument though, why is this so personal to you that you have to prove mkelliny wrong on him stating the temps he is experiencing? Whether he is right or wrong he is trying to help others by posting here (because like you said, that is the point of forums - something I am well aware of as I have over 10,000 posts across 5 forums myself) and a crusade against him is not beneficial to readers, you, him, or anyone else.

I apologize if you did not realize as good of an improvement in temps as he did, but can we stop arguing now?