Thermal Pad Reviews?

First, I did search and Google so please don't crucify me if i missed the info.

I ordered some NT-H1 to repaste my 6990m and CPU. With BF3 and Skyrim coming out I will probably overclock, plus, as a techie, it is just fun. I was watching the how-to for the 6970m and saw the thermal pads replaced. This got me interested and fujipoly seems the most popular. However, I can't get any by tomorrow for a reasonable price (there are no Amazon prime shippers and everyone else is like $40 for overnight).

So i started googling and searching the forums. Thing is, there are NO benchmarks for thermal pads that I can find! Do they exist?? I imagine the lack of support for memory temps recording is the root of the problem. However, I am a very objective person and I have found no evidence to support lower temps, longer life, higher overclocks from using high quality vs low quality thermal pads.

Does anyone have any evidence at all? Otherwise I will simply repaste and be done with it.

 

i repasted my OC'ed 6990m reusing the stock pads and my temps dropped 8-10c. i don't know what kind of improvement you will get with new thermal pads.

That said...i've since ordered 12w thermal pads. I'll let you know if it makes a difference over the stock stuff

 

I used fujipoly .5mm pads. I stacked several layers where I needed the thickness, although you can also get some thicker pads as well.

I don't have a comparison on my machine using a repaste with stock pads, so I can't tell you how much difference the pads alone make. My GPU temp under full furmark load is 72C though.

 

Well, I repasted CPU and GPU with NT-H1 about 2hrs ago. Using OCCT and HWmonitor, it looks like max and average temps for both dropped 10 degrees. I went from CPU 92max/88avg to 82max/77avg and GPU 94max/82avg to 79max/76avg. If you don't mind, I would love to know what you get with pads.

 

Running Prime95 with 8 threads at 100% CPU, and Furmark at the same time pegging the GPU:

Once the fans are going and everything, the CPU is stable at 78C and the GPU is stable at 71C. The highest I got (while the fans were spinning up at the start) was 82C on the CPU and 74C on the GPU. It looks like our paste is about equal since our CPU temps are similar. I think the pads give me another 4-5C on the GPU though. The ones they have on there are really bad.

 

Interesting, I just posted a longer post in the 6990m temps thread. Short version is: with GPU-Z I get 70.6 (+/-0.2) on 3 sensors and 76 on the 4th sensor. I wonder if the temp monitoring software reads that much differently. Also NT-H1 says put a dot then twist the heatsink to spread. I couldn't do that so i just put a dot and slapped the heatsink back on and then wiggled it the very very little it can wiggle. Looking at the heat pad placement it seems to really only touch the memory but who knows. Maybe i will repaste based on the recommendations of this forum or maybe i will just order more pads and try again in a week. ( I would hate to eat into my BF3 beta time because I will be going out of town on business for a while in October)

 

The high temp is probably the memory or other non-core components on the card. That is the stuff which is cooled by the thermal pads. You will definitely see that drop if you repad. That said, your temps are fine and aren't gonna hurt anything, even if you overclock.

 

here is my temps with icd and fujipoly. 30 mins extreme burn.

 

Wow I will have to try that when i get home. I bet my max temps are noticeably higher. Maybe i will just order a pad. It can't hurt. Are you guys ordering 1/4 sheet 0.5mm thickness?

 

what room temp u have??thxx!!

 

Inap, why are your clocks at 250/900? Those are crazy low Furmark temps.

 

man can u run a furmark test and tell me what temps u have with ur 6970m???i run one furmark for 10 mins dynamic backround disabled,full hd 8x AA and burn-in with extreme burn-in enabled with room temp 26 and my temps was 77.5/78.5/84.5 with gpuz!!is this good??

 

Looks like the screenshot was ~20 sec or so after he turned off Furmark. His GPU had already clocked down and the temps started to drop, as you can see on the graphs. Also, the GPU temp (the one that says 63C) is the current temp, not the max. The max component temps below lead me to believe his core was probably around 70C.

 

Wow, $13 for 1/4 sheet and $8 s&h? I know it is a drop in the bucket compared to the m17x but still. All that for a fee degrees on the ram? Has anyone found fujipoly for sale cheaper? I can't seem to find anywhere but frozencpu.

 

I am weary about buying thermal pads because I'm afraid they won't be soft enough to compress. I tried some 12W/Mk versions which were 1mm thick, they actually didn't compress very well and prevented the rest of my heatsink to fully contact their corresponding components, the die especially.

 

Fujipoly is very soft. Like a soft putty. Once you take the backing layers off, the .5mm stuff is very stretchy and feels almost like a very thin piece of play-doh or something. Once I had it on I could easily verify contact as I was fitting it by seeing the indentations in the pads.

 

where can i buy these thermal pads??

 

Frozencpu.com sells them. Just ordered another sheet of the fujiopoly thermal pads this morning.

 

thank you...what thickness and wattage do you need?

is it this stuff?

http://www.frozencpu.com/products/11631/thr-84/Fujipoly_Premium_System_Builder_Thermal_Pad_-_14_Sheet_-_150_x_100_x_05_-_Thermal_Conductivity_60_WmK.html

 

Yep that's the stuff I use. Have had really good results with it. I've used both the 1mm and the .5 mm thick stuff, just really depends on what you feel most comfortable (and gives the best thermal results). I know some folks will buy the .5mm and you can double it up if you need to.

 

It was about 21-22c.

 

what did you finally end up using, the 1mm or .5mm ?...cuz i was thinking of using the 1 mm since it would be easier to do (work wise) on the 580m

 

Currently I'm using 1mm on my 6970s, which has been working pretty well. I ordered the .5mm this morning just to see if I can get any lower temps with the thinner pads. I believe (folks can correct me if I'm wrong) that the standard dell size is 1mm, so should work well. Should have my .5mm within a week or so, so can provide an update on the heat differences between the two.

 

on my 6990, when you look at the stock pads over the ram, it looks almost 1mm...but not quite. I think0.5mm is too thin and 1mm is slightly too thick, so if fuijipoly is slightly compressible, 1mm should work fine

EDIT: you can find 1mm fujipoly on ebay cheap!!! Way cheaper than on frozencpu and no ridiculous shipping costs

 

anyone tried the phobya ones? Those are in different sizes and 7w/Mk so slightly more efficient than the fujipoly ones.

 

i use the Phobya WLP Ultra Thermal Pad 120 x 20 x 1mm and they are very good!!!

 

i buyed Phobya Thermalpad Ultra 5W/mk 50x50x1mm, and tried to use it on 5870m RAM in my R2 . but these ones are hard like batmans plastic armor (well maybe you can compress them by 1%) and i ended up on my heatsink to not touch GPU die properly or something like that becouse my GPU just wont work with it (or i did something wrong).
Anyway i saw nothing else but black screen (but everything else was working ok) and raged a little bit about it...........so i just reaplied the old-stock-crappy-soft thermal pads and hoped my system will boot properly..it did......UFFF relief...but these ones seem to be nice and sturdy..like i said....maybe i did something wrong...

10'C temp drop achieved just by repaste MX-4 on GPU core, stock thermal pads, and retention clip mod ....i found nice washers in old desktop PSU...much more better/safer than clips that can fell off or bend.

 

Wow i'm really late to the party. i understand if i'm trouble for grave digging. However i just pulled the trigger on fujipoly 1mm because i wasn't sure how big my gap is on vram. Also this is them ost recent thread i found about quality thermal pads via google.

 

I'm pretty sure you need 1mm because 0.5mm is just too thin... That's how the thermal pads is like on my 7970M.

 

You do want to keep it the same as the original pads (in this case 1mm), I always found that the 7w/mk pads liked to move around a lot at least with my MSI machine (which did admittedly involve lateral movement to get the heatsink in place) but got similarly good performance from their 5w/mk pads which were easier to use.

 

well i have an acer but the thickest part of my old/used thermal pad, the part that isn't compressed is about/is 1mm. they are ancient and rather dry, think it will be a decent upgrade.

 

Yes if they are getting crusty then any fresh set of pads should do the job.

I'll be moving to my 5935G temporarily as a side note, got an FX1800M to try in it which I got cheap.

 

has any else noticed sometime you only need 4 thermal pads even though there are 8 Vram chips? I know the non heatsinked side has to run hotter but, i did think/notice something. the un-heatsinked vram chips are directly above the the heatsinked side, just they are on the other side of the PCB. My thought is the heat might just be conducted through the pcb and BGA.

 

I just got mine in today and the fujipoly is closer to a dry/semi dry clay. it still compresses but, it's not as soft as my stock thermal pads (foam rubber... i think. also is 5 years old) it was also smashed in the mail and had few holes. the fact it was still intact makes me wonder if it was old and or expired, dryed out stock.

my temps jumped 10-20c all around but, i also REinstall my bottom cover after in stalling the new thermal pads so i'm testing again with out that cover.

EDIT: my temps went UP even with the cover off, my overclock is now unstable i'm not sure whats going on.

EDIT: I got the 11m/wk fujipoly