blaydes99 repaste results with AS5 and ICD7 (G73)

Well, I'd like to chime in with my results on my GPU repaste as well.

Compounds tested:
AS5 (I had this laying around since last year)
ICD7 (just bought a tube for $6.99 this week)

When I first took my G73 apart, I watched the youtube disassembly video before and during to make sure I did everything in the right order and didn't break anything. I can now take it apart by memory and have avoided any damage to my laptop.

I first replaced the default Asus compound with some Arctic Silver 5. I didn't save my GPU-Z logs previous to this, but here are some temps from the post AS5 application:

72 C idle (warmed up and after running a few games)
98 C peak during Metro 2033 (DX10 mode)

I just wasn't happy with the temps, since they were only a few degress lower than the stock thermal paste from Asus. Tonight I reapplied with the IC Diamond 7 compound, applying a small amount: about 1/2 a kernel of corn (like Chastity). I tried to twist the heatsink back and forth a little before screwing it back down to help spread it. Here are my ICD7 GPU-Z temps:

61 C idle (warmed up after some gaming)
83 C peak during Metro 2033 after playing for 15-20 mins (DX10 mode)

Another note: was 78 C max during a 3dMark06 run.

So, I must say that I was beginning to lose hope that I would never get decent temps, but I can say that I wouldn't believe it unless I saw the results that I got with ICD7. If you really need to get your temps down, give is a shot. $6.99 shipped isn't a bad deal at all.

 

grats on that..i too am waiting for the day that i get the guts to pry my laptop open..

 

blaydes99 those two TIMs wouldnt give you such a difference in temperatures - at best around 2C after AS5 has burned in as you can find in many reviews and tests like THIS.
You must have applied the AS5 or secured the heatsink that time more poorly. You cant just post something like that when theres so much variables and when all the reviews and other tests contradict your verdict (some members here also got the same temperatures with AS5 as users who did a good ICD application like THIS or THIS)...
When it comes to high end pastes you wont get more than 1 or 2C difference and the most important thing is application not the paste itself than. Any of the top ones should be just as fine as the other as long as you apply them properly.

 

Now that HellCry is done with the all-too-common disclaimer on TIM performance, I'd like to welcome you to Team IC Diamond!

Congrats on your new temps, and enjoy the weekend

 

Just ordered my ICD7 last night ... I've read predominently good things about it. Can't wait.

 

For a noob.. is it better to pay 35 $ for ken to apply it or should i go for it myself?

 

This is the kind of job where if you want it done right, do it yourself (as long as it doesn't void your warranty).

 

I did a pretty thorough AS5 application and immediately saw 52-53 idle and 85-88 load

I also remember that because your not putting directly on the die (due to how a GPU is situated versus a CPU) you need to actually cover the entire area. (thinly of course)

 

Ken @ Gentech is quite adept at it, if you want him to do it. Well worth the $35 if you have high temps.

 

Thanks everyone for commenting - I'm very happy with my IC Diamond results.

You may be right, I may have put AS5 on improperly. I actually disassembled my laptop twice and did two different applications of AS5 with similar results. In reality, I'm not trying to slam or say bad things about AS5, but instead praise ICD7 for the easy application and great results from the get-go.

I would be happy to see other users get great results with AS5, but it may be helpful to find out once and for all which application is best on the G73 with AS5, as I tried both the "thin spread method" and the "glob in the middle" method and got about the same temps.

 

Good to know i have some AS5 sitting hear on my desk from my last desktop build. I might give it a go next week when I have a chance.

 

Just one thing to note with AS5 - put too much on and it can ooze off the die and shorten something near the core since this paste is electrically conductive unlike ICD or MX-3.

 

Arctic Silver 5 is not electrically conductive. Arctic Silver says this on their web site and I've personally stuck a multimeter's probes into the stuff and seen infinite resistance. So it is an insulator.

They say that it is slightly electrically capacitive if under pressure but that does not mean it will fry anything.

I even once tried spreading it all over the top of an Athlon CPU. Those have electrical contact points on top for multiplier and other configuration. It worked perfectly fine as far as I could tell.

 

That's good to know. Judging by the fact that it is shiny and metallic in color, I would have thought it was conductive, but I never put my Fluke probes in it. Now I know.

 

AS5 has a different cure time than ICD7 which might have skewed his results. Also, we don't know how well the surface was cleaned, by the sounds of it he applied AS5 then ICD7 right after.

Personally, ICD7 is in a league of it's own. It's conductive properties are way greater than AS5, however, it's hard to compare the two with a laptop setup since both transfer heat away from the CPU but still bottlenecked at the heatsink. If you can find a way to transfer heat off the heatsink faster, ICD7 will perform far superior to AS5.

Eitherway, we're not going to notice a 10 degree difference just through thermal pastes, I'd say max 3-5 degrees and that's pushing it. But, none-the-less, a few degrees is a few degrees and ICD7 is well worth the 7 bucks.

 

It all depends on who is testing and what the setup is. I've seen tests where ICD7 comes out on top and others where mx-2 is 2 degrees better on a watercooled setup. On water the best temps were always CL liquid pro because that stuff basically fuses the heatsink to the die so that stuff is probably best utilized on those setups, air not so much.

 

How much will the temps really drop using AS5 after the cure time has passed? I really can't see it dropping more than 5 C, and then it is still not hitting the temps I'm getting with ICD7.

Maybe my application was bad (air bubbles/too thick)?

 

good thing i've got ICD 24 applied.. it should do well... I always though AS5 was crap and today its been proven.

 

I have thermaltake TG-1 sitting on my desk. I just keep looking at it then looking at my G73. So hard to decide...

 

I had the same dilemma. I had a tube of Arctic Silver II on my desk. A friend had some AS5. In the end, I decided to order some better stuff and wait a few days for it to arrive.

 

That's not true. As it was said all high performance TIMs(i.e ICD, AS5, AC MX2, Tuniq) have pretty close performence 1C. What is really matter is applying method. Just look here http://forum.notebookreview.com/asus-gaming-notebook-forum/507854-thermal-paste-g73jh-gpu-7.html#post6625608. And it is still 1.5 week to go of curing time. Manufacturer states that temps will drop 1-5C during curing period. (Hopefully 5C)

 

Comments like this just make me laugh.
You know better than all the much more tech savvy folk doing objective and much more accurate tests and reviews on the web.
Thermal paste Roundups:
HardwareLogic
XBitLabs
Do your homework before posting such dimwitted comments...

 

What's the shelf life of AS5? I thought one of the main advantages of IC7 is that once applied, it will last much longer than other compounds, so that you don't have to replace it every 6 months.

 

I couldnt find any info on how long ICD is supposed to hold, all i could find is "durable" and that can be interpreted really widely but i know that MX-3 was advertised to hold "a computers lifetime unlike AS5" and on their website they give you a 8 year warranty for it.