Repaste those Razer Blade 2015's at your own risk!

It does definitely run cooler than stock but with the amount of effort I put into it, it's not even worth it. I would just switch Max CPU usage in Power Options to 99% or drop your cores down a few multipliers and avoid it all together.

 

You think it is not worthy because you have too many failed attempts. However, it may still be worthy to other people if they can get straight to your key point.

So in the end by how many degree celcus? Roughly! And what exactly is your key modification, say filling "the gap between the 2 heat pipes" with thermal paste "Prolimatech PK-3 or Gelid Solutions GC Extreme" so that the entire die surface is covered by thermal pastes?

Or your improvement is insignificant for a new laptop? But only significant for an old laptop (after 2-3 years) with mostly solidified thermal pastes, tons of dust inside, and expired warranty, so that you want to open it to clean up the dust and at the same time replace the solidified thermal paste?

 

I think it isn't worth the time invested into it and the voiding of the warranty. Yes I did get it cooler however you could accomplish the same just dropping the core multipliers a bit.

The latest attempts at repasting were using the compound and lining up the copper pipes and the gap in the pipes with thermal compound. I would suggest leaving it as-is until it is needed a few years down the road.

 

I have an incoming already-opened Razer Blade 2015 that I have to fit back together since the seller took it for whatever parts he could get... (though its mostly soldered).

Got a few more reapplications of Liquid Ultra to go through, and I never had problems using it.

 

Did you notice any temperature drop with Liquid Ultra?

 

I can't tell you if there's any notable difference as my system came opened by its previous owner, so the paste would've been bad to begin with. So far, the temperatures seem roughly the same, if not, maybe better by 2-3C compared to a friend's Blade. Not a big enough difference.

 

To give an update. Temperatures are very bad again. had to drop multiplier down to 26x to make it fully usable. 26x with Intel XTU is at 91c with no throttling.

 

Hey, I just wanted to give my experience repasting my 2014 Razer Blade 14". My warranty just ended so I decided to give it a go. I found the old paste was pretty dry. I used arctic mx4. Putting paste in the gaps and lines on each heatpipe I got a decent temperature drop. In the game I tested it used to drop framerates if I didn't use my cooling pad so I'm assuming it got to the thermal limit. Now temps are 90C max for the cpu and 81C max for the gpu with no frame rate drop. So I'm pretty happy and would recommend it as long as you've got some experience with this type of thing.


Unfortunately, I didn't check baseline temps
*****|||||
*****|||||
*****|||||
*****|||||
*****|||||
-------------------
**--------------
****--------
******---

 

Do you have a baseline before?

 

What game are you playing at what settings when testing?

 

I'm embarrassed to say since it's such a good gaming laptop and I only play such an old game but it's left 4 dead 2. It's on full resolution at the maximum settings where I can get above 60fps which is somewhere around medium for most.

Ok, I ran the XTU cpu stress test to hopefully get a more standardized number for you guys. The cpu temperature got to a max of 87C on throttlestop. This was without a cooling pad with ambient temperature probably in the lower 70s F.

 

Loop 3DMark or something and see where that takes it.

 

That's an odd stress test... 98-100% fluctuations are common for XTU stress, but you're below 90% util on a couple cores in the second that screenshot was taken. That's not supposed to happen. Is that just a random quirk due to the screenshot program or does it fluctuate to the high 80% range (your c0%) often during XTU stress? You also have a HUGE range between your temps per core. 10 degrees between two different cores is a bad paste job. I've done enough repastes here to know that much >_>. I used to get annoyed for a difference of 5; 10 would drive me nuts haha.

 

HOPE NOBODY MINDS A SPOT OF THREAD NECROMANCY


Right, so I've got LVNeptune's Razer now, and from when I got it, one of the cores would hit 99C very quickly under turbo loads. Temps were pretty low and stable without turbo (79-83C) but I didn't like the spiking. I've been trying different pasting methods with MX-4 which I used to replace his IC7, but I haven't succeeded in improving temps beyond stock performance (I judge this by unlocking turbo and running 3dmark. Less throttling = higher score, more throttling = lower score. Scoring about 9120 with turbo and 8700 without).

What I have done currently is use MX4 to fill in the gaps between the copper heat pipes, then spread it flat with a card. Then, I put a generous dollop on the CPU/GPU die and attached the sink.

What I am thinking of doing is using this photo as a guide to reapply the paste: http://forum.notebookreview.com/attachments/photo-feb-20-6-30-09-pm-jpg.121937/

I'm thinking that perhaps I don't want any material in between the copper pipes and I don't want to put any paste on the die directly, but instead only have the paste on the parts of the copper pipe that make contact with the CPU/GPU. Since the pipes are what carry the heat away from the dies, perhaps having paste on the chips which is not in direct contact with the pipes will harm the heat transfer instead of help. Thoughts? Has anyone else out there figured out the best way to apply paste on this design?

 

Well, I tried mimicking the distribution of the picture of the original pasting. I mixed both IC7 and MX-4 and the results were pretty good. Idle temps down to 37-40. 3dmark11 peaks at 78C GPU, 80CPU (x26). Turbo on, CPU peaks at 95C and 3dmark gets a score of 9200. So, I think I have successfully improved on stock temps.

 

Breakthrough: Tried Xi - 3 HDT -- a TIM specifically designed for HDT cooling solutions, as the blade has. The result: XTU stress tests are 5C lower (85C average) with full turbo. Excellent.

I simply applied it to the CPU using line method and to the GPU using 3 generous stripes (one strip lining up with each heat pipe).

 

That's promising. It looks like that TIM is specifically designed for this use case. (Heatpipe direct touch)

 

Here's a quote from someone at razer who got in touch with me via PM on razer insider:

"I can't comment specifically on the methodology you used (nor the mixture of TIM) as we do things a differently here - but based on your temperature readings I believe you have discovered an excellent solution."

 

I am doing a motherboard swap (my existing motherboard's keyboard and touchpad dont work because of the motherboard). I am having trouble finding the specific TIM for HDT (Xi-3), so whats the recommended paste and application method for me as I'm doing the replacement?

 

Nice. Thanks for the note. I might give this a shot, if it really does drop the temps by 5C.

If it matters, I attempted to repaste using a more "traditional" paste like Arctic Silver 5, and replicated the issue originally mentioned in this thread (temps went up). I find it really odd that Razer went with direct heatpipe cooling, instead of having some kind of copper plate.

 

Hey there!

Where did you find this Xigmatek XI-3 HDT? I've been looking for it but I cannot find it anywhere.

 

Went to Pantip Plaza and bought it from a random shop in the basement.


seems that amazon sells it: http://www.amazon.co.uk/Electronics-Xigmatek-Xi-3-HDT-Thermal-Paste/dp/B00EZCLTD6

 

Well, but I don't live in the UK Thanks, tho!

Sent from my Nexus 6P using Tapatalk

 

Me neither, but I searched for xi-3 HDT and I got this off amazon.com: http://www.amazon.com/Recombinant-A...=UTF8&qid=1454556363&sr=8-2&keywords=xi-3+hdt

 

From those pics, this is an absolutely joke of a design for a "gaming" laptop. I have not read through the whole thread but what about the gaps on the heatpipe, they also tend to bend at the pipes.

 

What a ******** design. What you need to do is solder a thin copper shim to the heatsink and close the ridiculous gap so you can make proper full contact to the die area. This is ridiculous.

 

My thoughts exactly. If they just made the laptop 2mm thicker is would have so much better cooling performance with a copper heatspreader. The temepratures are nuts with these machines on the long run. I dont see them survive for more than 3 years.

 

I ordered some Tuniq TX-4 thermal paste. I'll post a message after it shows up, to report the results that I get.

Sent from my XT1575 using Tapatalk

 

Ok. So I got this done, and posted it as a new thread. I got a 6C temperature drop between repasting and undervolting.
http://forum.notebookreview.com/threads/razer-blade-14-2015-repaste-log.788861/

Overall, it isn't worth repasting if you're looking for any kind of practical real-world benefit. But if you like to play with your computer gear because you find it fun, then it's a fun little project to jump into.