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

This is during CPU intensive application. If you are using it for normal work/office type stuff it's great. Slightly warm to the touch and great battery life (compared to what I'm used to). Gaming it needs to not be on your lap unless you don't want kids

 

hah. thanks. the AW15 is great for what it is.. i'm getting about 6hrs web surfing, youtube, etc. and it's quiet. but yea it's not that portable. i have light bleed on mine, so i have a replacement coming in, but if the display quality is the same.. i will return it and just take a shot at the 1080p RB.

 

6 hours on a 15"? That's great! I am really excited for Broadwell. Battery life should extend significantly. Maybe we'll see a Razer Blade Late-2015 model.

 

yea i was quite surprised by it too, since the only review that is out ( http://www.notebookcheck.net/Alienware-15-Notebook-Review.136667.0.html) says about 5hrs w medium brightness. i usually have mine set at about 75% bright and i am getting 6hrs avg so far.. i'm thinking the 1080p Blade will be on par with this.. hopefully.

 

This is true of EVERY gaming laptop.

 

@LVNeptune would you recommend getting the 2015 RB or would it be better to get something to last until Razer release a Broadwell version of the Blade?

 

Razer hasn't done 2 refreshes in the same year before so you may be waiting a full year if you wait.

 

I can game on my lap and it doesn't overheat, as long as I don't put my leg under a main vent.
Yes I've done it before, both on the P370SM3 and on my old D900F.

 

I'm OK with waiting a year, I just don't really want to buy the RB 2015 now, only to have the 2016 be much better as I can't really afford to sell a 2015 just to buy a 2016.
Generally speaking how much of an improvement will Broadwell make to something like the Blade? or will Razer jump straight to Skylake with the 2016 model?

edit: sorry to sort of derail the thread btw, I just can't seem to send a PM

 

When you buy the 2016, the 2017 will end up being better. It's an endless cycle.

 

From what I have read skylake should be a definite improvement but with odds of delays I would not expect things at face value.
Besides you have to jump in at some point. Just like hfm said it is an endless cycle. I always upgrade my phone with the next gen at every stage. Sure it costs some money but if you always want the best you have to pay for it.

If I were you I would see if you can get a blade 2015 from the microsoft store with education discount and extra warranty. Then that way if you want the 2016 blade, the 10% discount will have helped ease the upgrade cost and plus you will have an extra year of warranty left.

 

No odds of delays. It was already confirmed delayed because Broadwell hasn't been fully available yet. They already have Skylake R&D done and ready to release they just have to wait until Broadwell sells lol.

 

Neptune when you were experiencing the throttling did you ever try throttlestop? I know you probably wanted an actual fix though. I have to use throttlestop on my surface pro 3 or it is gimped.

 

No, what does that do exactly? Haven't used it in years.

 

has anyone taken temperatures of the casing's top and bottom after at least an hour of gaming? would really like to know how hot the aluminum gets. thank you

 

Testing with AS Ceramique 2 now. Will post back after 4 hours of stress testing a few times to break in the compound.

 

It is a program that essentially does as the name implies....stops thermal throttling. I use it on my surface pro 3 because once my surface pro 3 hits some temperature in the 80's it will throttle down to until it cools to 60 then goes back up. Running throttlestop prevents the throttling down and when I was using it would run constant at 88-89 degrees without throttling issues.

 

Doesn't seem to do anything on the RB.

 

Question for you Neptune. I cant stand the ultra high fan speed kicking on. Waaayyy too loud imo. Frist 3 notches are ok. If I want to undervolt is that the best way to do it? and that is done in power management settings for the processor and changing the min and max right?

 

Undervolt is possible however I would not recommend it. I had a bad experience in the past with doing that on the RB and the machine hard locked and would not shut off holding the power button for 2+ minutes. I literally *HAD* to open it and unplug the battery to get it to shut off.

I would recommend launching XTU and lowering the core multipliers. Dropping them by 2-3 really lowers thermals.

 

Okay will try that. Not liking the 97 Celsius temps either and all I am playing is dota 2 lol. You got yours down to 85ish with a repaste and then a copper shim correct? or was it just a repaste?

 

That was repaste w/ globs of diamond. I am testing with Ceramique 2 now.

 

What are the preliminary numbers looking like? Clobs of ic7 you said was 89 when flat and 85 elevated? How does the new compound look? I am waiting on your results to experiment on mine.

 

Ceramique 2 was bad. I'm talking consistently 98-99c @ 50% throttle bad. I tried applying it multiple times and consistently had the same results. Switched back to IC7 and now I'm looking at max 90c. It's definitely all in the application.

Shim worked when I coated both sides with IC7 but when I tried epoxying it to the heatsink the temps were above what I was hoping for so I nixed that idea.

 

Thanks for the news. I will pick up some IC7 tomorrow and generously apply it. You used the 3 globs and then sat it?

 

I don't want to take it back off but I coated the crevice between the pipe and the mount on 3 sides and put it on the copper and on the mount off to the side to make sure it got everywhere on the die

 

While I can't open my Razer Blade, I performed a test for you on my old NP8130 with an i7-2630QM GTX560m to help out. My recollection was that I had seen better results with Prolimatech PK-1 over IC7. This NP8130 was the one system I had never switched since it was working well enough with IC7. So off came the diamond paste.

• Immediately set fans to 100% before running tests (something we can't unfortunately do to easily normalize repasting tests on the Razer)
• Same ambient temp/location on desk/etc.
• Furmark & XTU Stress were run for 15 mins. XTU Graph is final 10 mins.
• I cleaned my fans of dust before testing either paste.
• On graphs, the light gray lines you'll see the CPU Temp between are [62C-67C]

IC7: CPU Max 68C, but the temp spent much of its time @ 66-68C:


PK-1: CPU Max 67C, and the temp was consistently lower @ 64-66C:


Final note: I would call this is a win for PK-1, which requires no break-in time per the manufacturer. I'll stress test again after a week or so of use anyway. Also in general, I have found with the higher the normal temperature (like w/ the Razer Blade) the benefit you'll get from a better paste becomes more apparent. With that in mind, I would encourage people who are attempting a repaste to go with the improved Prolimatech PK-3 or Gelid Solutions GC Extreme per this comparison I found online: http://www.tomshardware.com/reviews/thermal-paste-performance-benchmark,3616-20.html

... Or if you find any other pastes that are newer & better yet, to have at least been reviewed in comparison to those I have mentioned today.

 

I don't want to discount your effort, but without proper controls and repeated tests across a period of time, such tests can't be extrapolated across even a population of the same device.

The old watercooling forums I visited used to be a source of hilarious discourse on thermal compounds and application methodologies which bordered on witchcraft from otherwise scientifically minded people.

 

Many variables, heatsink differences, copper thickness differences. I'm aware it won't be 100% reproducible.

 

Yeah no worries. I just repasted something because I had to after reading this thread I guess But, that's why I referenced the toms hardware article too. I found other older reviews after posting this showing Prolimatech outperforming IC Diamond, so hopefully someone tries one of these pastes and can help the Razer chill a bit more --

 

Does anyone have a profile to share or settings I can copy to use in XTU? My 2015 is showing 98 degrees top temp in XTU benchmarking and it termal throttles. My temps go up and down constantly from 81-98.

Im not looking for miracles here, just a steady temperature which doesnt throttle.

 

Drop the core multipliers down about 4 on each core and it should stop it.

 

Thanks for the quick reply.

I've had a bit of a play around and I've found that to completely stop thermal throttleing in XTU I have to set the core multipliers to 29 each. Any higher and I get some throttling, more as I increase the cores. Everything else is stock settings.

If I lower it to 27 then it also stays much quieter but my benchmark score is going from 730 stock to something like 630 with 27 selected.

Any advice? I'm very new to XTU and don't know what Im doing with it that much.

 

BTW, good effort on trying to improve the thermal efficiency. II read through this thread and I see you've gotten some stick for it from other sites. Bit of a shame as you've taken a chance which most people won't have the bottle to do yet they'll happily take advantage of any reward you discover when doing so. Ill be following this thread to see the outcome.

 

Crap. Was hoping I was one of the few and that maybe I had a defective unit...or one that hadn't been pasted very well or something. Check my thread here. My XTU profile is at the bottom. Undervolting a little with those multipliers did the job. You'll notice it brings the temps down just enough to "settle" the damn system down. The up/down/up airplane on a runway effect was tough to live with during games every 3 mins.

http://forum.notebookreview.com/thr...2015-fan-noise-up-down-possible-fixes.772128/

 

I should also note that the CPU is crazy fast even with the multipliers dropped down. The XTU benchmark I never put much faith in myself. For me it's always wanting the fastest I can get but with thermal throttling on this it's just not happening. As far as "real world" I doubt anyone is going to notice the difference with the multipliers clocked down some. One of the very few games that would be impacted, which really isn't that impacted, is SC2.

 

Or lovely World of tank - single core game

 

Ill keep the cores down for now then and check back once in a while to see if you find the perfect repaste method are you continuing or staying where u r now?

 

I'm pretty much done at this point. My only other test would be to try that other paste recommended but with the amount of work with break-in and such, I'm good...

 

Are you running you CPU at full speed without throttling now then? Or did you reduce the CPU to some extent? Sorry, I read the pages in a mixed up order.

 

If the laptop is elevated some I can max out the multipliers at 40/40/40/39 if I recall correctly without throttling. once it lays flat though I get throttling. I get no throttling at normal multipliers.

 

Pretty good then.

 

Overall after all this would you advise against repasting?

 

It's really YMMV. I probably wouldn't recommend it until your factory warranty is up. I would also suggest if you bought from MS store pay the $149 for the 2 year accidental, if that overrides the factory warranty repaste it. Alternatively you should purchase a warranty from Squaretrade, the accidental warranty I believe overrides the factory warranty as well. Need to call to clarify though.

 

hmm.. i never got asked if i wanted the accidental warranty from the MS online store. can i still get it later? like Apple gives you a certain time frame post purchase to get their apple care

 

Should be able to. Call their 800 number. It's a separate cart item you have to add.

 

Yeah... tell me Neptune. Front to back. Has your "story" changed at all? Your daily use? Your gaming use? What's the experience like with the fans? (Edit: removing throttling, you answered that and I was behind on my alerts from NBR).

IE: Now that all is said and done, it's about the same...? it's a little worse...? You can play a game (examples) and don't hear the fans *at max* anymore... since thermals must have improved!?

 

Definitely at max during gaming due to the CPU/GPU heatsink being one piece so it heats up the CPU more when GPU is running. So far nothing has changed. Using CPU only though the CPU fan doesn't max.

 

I'm in the UK so none of those warranty options for me. I paid for the HID global warranty though. I doubt they'll cover a repaste and I wouldn't expect them to.

 

So what is your final conclusion after all? Have you managed to get cooler than the stock paste? How to repaste RB2015 so that it can get 5C cooler?
Maybe it is time for you to give a full summary for the sake of both the company (so that they know how to improve thermal dissipation in future versions) and the customer (so that they can repaste their RB2015 after their warranty expire or the stock pastes solidifies after 2-3 years). I am also quite curious about this!-