Razer Blade Pro 17" GTX 1080 GPU/7820K CPU 2017 Owner's Lounge

I’ve apparently good silicon lottery, almost reach 1080mq
https://www.3dmark.com/3dm/27184677
Btw default timespy is not 1080p I believe it’s 1440p dx12. Firestrike is 1080p dx11.

Looking for a new toy , i wanted to give 1300 but unfortunately I’ve to move to other country so it’ll be too much things I’ve to bring

 

Lol sorry guys, I’m leaning more towards keeping it. Not gonna lie though, this machine sure does run hot during regular use (I haven’t even played any games on it yet)...I have about a 9-10 degree variance between core 0 and core 3 in throttlestop benchmark. I’m assuming I’ll definitely need to repaste. Other than liquid metal (not sure if it’s recommended on this machine) what’s the next best thing?


Sent from my iPhone using Tapatalk

 

Go LM. The machine body will be hot no matter what repaste job you choose. Might as well do the best.


Sent from my iPhone using Tapatalk

 

Is the battery full charged (like 90ish% not charging markt). I don’t have Pro model but while charging Blade cnc aluminum unibody usually hotter because it transfers heat faster than other cassis material, after windows show no charging mark, usually it’s less hot. I tried though both your 4K model in gamescom here, it’s basically all days usage for visitors there, I play a bit Overwatch and it was indeed warm but not too hot to touch, maybe it was too short, but considering there was line before and after me it should be used quite long at that day.

 

Charging/discharging will certainly cause heat to build up around the battery. But this will pale in comparison to the heat from the vapor chamber area above the keyboard. The unibody chassis will conduct that heat rapidly to the rest of the machine.

Before a repaste, my keyboard was uncomfortably hot under load. Afterwards, warm but not bad for your hand resting on a chunk of aluminum.


Sent from my iPhone using Tapatalk

 

The machine wasn't even plugged in, I was just having it on my lap while browsing some streams and reddit threads. I went ahead and ordered Thermal Grizzly Conductonaut last night, I should have it today. I'll report on my machines results when I finish. Is it important I find some foam barriers/electrical tape to stop the liquid metal from leaking onto the motherboard?

 

Idk much about LM but yes they add it for avoiding the leak, you can ask @Raidriar further he did that on Pro FHD. Yes I think playing on battery on Pro 4K may make the machine hot since it always use dgpu unlike other Blade with optimus technology, like playing games on your phone the cassis would be hot, the temperature difference may because uneven thermal paste, I would try first repasting with better compound though before go to LM personally.

 

Problem/question:

I was interested in undervolting my CPU this morning and dropped voltage down by 50mV. Not sure exactly when it happened, I suspect it was during an OCCT stress test, but the processor is now locked to 800Mhz. Cannot get it to budge. I've downloaded Throttlestop and tried to disable BD PROCHOT to no avail. Reinstalled Synapse 2 just now and same thing. Looking for some next step advice...

 

Hmm weird, Have you done a full power cycle on the machine yet? (shutdown-disconnect battery - press and hold power button for 15 sec) You might need to reset/disconnect cmos I believe.

 

Problem resolved. I created a new thread with this question yesterday and forgot to close out both that and this here.

Turns out, I managed to reduce the Core IccMax to 1A (down from 128A) in XTU. Not sure how, really. It's not something I've ever messed around with but I XTU does behave oddly with the RZperformance integration. Regardless, CPU now throttles back up to whatever clock speed I want instead of 800MHz and pulls more than 3-4Watts max. I was convinced that I had blown out some array of sensors on the motherboard and was going to have to endure another 2-3 months of not having the laptop. Yikes.

 

I would look into unlocking the BIOS and setting your desired values there, I always prefer BIOS level tuning so that it always works independently if software.

 

Just an update before I repaste with liquid metal, was playing Anthem @3440x1440p maxed out without AA with the below settings. My question is, if I could oc the gpu to at least 1800mhz while on a fixed undervolt, shouldn’t that mean that I’m at least matching a stock desktop 1080? Ive noticed that I’m definitely not power throttling on either components on the laptop.

Undervolt cpu/cpu cache @125mv offset
Undervolt gpu via after burner @ 775mv locked at 1600mhz




Sent from my iPhone using Tapatalk

 

Try to measure it with benchmarks and FPS achievement here to check if you get desktop 1080 performance but I doubt it because ac adapter (250watt) is already not capable to deliver full Gtx 1080 laptop performance, and since anthem new it’s still not listed on notebookcheck game FPS for that card, how much FPS did you get on Anthem?

 

I can easily get 1800MHz out of the GPU. Pulling 145ish Watts.


Sent from my iPad using Tapatalk

 

I get around 45-60 FPS in anthem @3440x1440p. So I have finally repasted with liquid metal, and idk about my results. I see a difference on the cpu side (and can actually oc to 4.1ghz stable), but the gpu I’m not seeing much of a difference. I’m now able to set a voltage curve @ 900mv with a core boost clock at 1848mhz. But on the below pic I’m getting the same temps on my ( low power curve @1600mhz/775mv). Maybe I need to realign the dumb cpu/gpu tri-posts?




Sent from my iPhone using Tapatalk

 

According to this video your Anthem FPS achievement isn’t bad at all it’s desktop 1080 same resolution on ultra setting. Though it’s still demo, maybe not optimized yet. And if it’s gaming temperature, it’s great I like that temperatures.

 

I’m playing the full game right now via EApremium, I’m sure the actual day one patch will boost FPS so I’ll stay closer to 60fps more often.

**Edit
Just watched the video, I'm sure they are running with AA maxed out. I turned it off since I didn't see much of a difference.

Sent from my iPhone using Tapatalk

 

Here is mine. I max out with 1836 MHz @ 900 mV right now. Seems very stable and temps never see 80C.

 

Man....my 1850mhz “ish” curve puts my gpu @ 90c quickly but doesn’t go any higher. Sounds like your heatsink has way better contact with the liquid metal or I’m not using enough
I’m sure yours would be very quiet with 775@1600mhz

Edit:

I believe the sweet spot is to keep the machine below 75 c on both cpu and gpu (which is possible) and we will achieve the the “maxq” setting in fan noise but with faster boost performance. Base clock on the desktop 1080 is 1633mhz. Can anyone send me the link to iunlock’s review on the Razer Blade Pro 2017 v2? I read this whole thread and most of it was people arguing with some razer fanboy who wouldn’t send proof for ? Reason lol

Sent from my iPhone using Tapatalk

 

I had HID Evolution do my LM repaste so I imagine it is about as good as it gets.

 

Ran some PUBG and Rocket League yesterday for about 3 hours straight. Average CPU and GPU temps were 79C, with the GPU pulling an average of 125 watts and the CPU pulling 25 watts. The GPU clock speed was pretty much pegged at MHz, with some random 50-100 MHz drops in PUBG.

The PCH was averaging around 90C unfortunately. I know the Tj,max for that device is around 110C, but 90C still seems high to me.

 

Jesus Christ, that’s way too high. Add some thermal pads or something, that is not safe operating temperature. I added thermal pads on my pch and stacked them until they contacted the bottom base so that it acts like a sort of heatsink.

 

Don’t they come with them from the factory? Mine has a huge thermal pad sitting on the pch and sticks to the bottom plate.

Also, I'm abit confused on the misinformation on this machine and its performance (when tweaked) It seems we are able to get full desktop gtx 1080 performance if you find the right voltage curve for the laptop. If needed, I'm able to have this machine run very quietly if i control the curve on the lower end and lock it at 1470 mhz which is a 1080 max q pretty much.



Sent from my iPhone using Tapatalk

 

My RBP FHD 1060 model did not have a pch thermal pad from the factory, only for the hard drive and nvme ssd

 

This machine was made before maxq born, most probably that pads is their one of cooling solution for handling 1080 thermal. No you won’t get 1080 desktop performance there’s no enough power supply doing that, that’s why it’s locked lower, higher maybe you’ll get battery run down issue even on wall plug.

 

Anyone want to mark up an image to indicate the PCH location? I'm going to open her up and ensure a thermal pad is in proper contact.

 

Here ya go:

 

Fast response! Thanks!

 

Sure enough, my battery is growing again. I wanted to investigate the PCH thermal pad that HID put on after my last issue so I took the back cover off this morning. One of the central cells is clearly 25% thicker than it should be. Regarding the thermal pads, there were 4 of them stacked to create contact between the PCH and the back cover.

Right now, I've got the cover off (and the thermal pads), the machine is at a 45 degree angle, and I am running a Valley loop. Average CPU temp is 70, average GPU temp is 72, average PCH temp is 87 and very slowly rising. Any reason the PCH would be getting so hot?

 

I have no idea what you got going on, my PCH temp maxes out at 57 and averages 54. Something is wrong with your machine.

 

Hmmm that’s weird, we have the same machine, here’s my temps while playing anthem for 2 hours on my “1080 max-q” setting on the blade:




Now when I switch to my “faster” gpu voltage curve the temps do spike up abit:






These machines really can’t be pushed much further “longevity wise” I will be keeping mine on my 1600mhz/775mv curve to keep the pch at 71c as heat is the killer of batteries (swelling) and pc components.

Sent from my iPhone using Tapatalk

 

Perhaps there is a bad sensor or something. I've left the machine idling for 12 hours now. Average CPU: 56C, average GPU: 57C, average PCH: 64.4C

Why in the world would the PCH be ~ 8 degrees warmer than the CPU/GPU. The offset is somewhat calming though as the same offset exists while in game (8-10 deg)

 

Are you sure the thermal pads are making contact with the bottom case?

 

I just removed the battery this morning and I think the bulge had deflected the back panel just enough to lift the pads. In fact, the pads were diagonal so I'm guessing that they had slid down and away along the panel. After re-seating the pads and buttoning the machine back up sans-battery, the PCH temp is now slightly less than the GPU. Frustrating.

I think I'm going to skip putting a battery back in the unit. 5 minutes of interweb research shows that Li-ion batteries don't do so well when over 45-50C and clearly the design of the RBP is not able to hit the mark. I don't feel like I'm pushing the machine dramatically, either, as I'm using the balanced CPU mode and a custom voltage curve in afterburner. CPU and GPU temps are really quite good at this point with LM TIM. That said, the armrest is still what I would call "very warm" right now without the battery in place. Don't think it is worth it to continually soak a battery in that environment. The uni body chassis conducts so much heat that I think it is a losing battle.

 

I would never buy another razer laptop again, I’ve babied my RBP and temps have never exceeded 80 degrees, routinely run in the upper 60s full load, and my battery is acting up. I’m under warranty but I doubt anything razer does will be a permanent fix. Best to just sell and find something else that isn’t poorly designed. My goal is to offload a bunch of computers and hold cash until icelake arrives later this year and jump on an Alienware Area 51-M. Razer was a costly mistake that I will not be repeating again, no matter how attractive the chassis is.

 

This x1000.

I want to love this laptop, I really do. There is just no reason to think that Razer will ever design a good laptop given the abysmal track record.

I've also been looking at the A5-M. I would not be happy about the weight and the dual power supplies, but Jesus. Full power desktop CPU and enough cooling vents to actually evacuate heat. May be a winner.

 

Just reading that your having issues with your blade pro which runs ALOT cooler than ours is discouraging me from messing with mine any further, the warranty on mine is until April this year and I feel the more I use it even while I have tweaked it to the best of my abilities to keep the temps under control, I’m afraid of the battery swelling due to the thermal design. I recently picked up a lightly used Aorus X5 v7 with very similar specs and once I’m done testing it out I might just put my blade pro up for sale


Sent from my iPhone using Tapatalk

 

Single 330W PSU is enough to run that beast, I have quite a few leftover from my M18x systems. It also weighs about 3lbs less than my M18x, I really do think it’s going to be a winner

 

It has dual DC inputs around back. Will they have a single PSU with a split cable?

 

Not necessary, just hook in one 330W to one jack and in the back and it works just fine. I’m sure when somebody cracks the firmware we’ll be able to run 2 x 330W PSU to extract every last bit of power from those chips

 

We could bring further power to RBP if we can do something about heat dissipation . I am thinking of putting copper shin on both CPU and GPU plus attaching some extra DIY flat copper tube on the heat sink assembly? Anyone has done this? I came across a certain iunlock whos selling RBP 2016 with modded heat sink but no pictures so no clue if this is really doable.

 

Liquid metal + undervolting is more than enough to keep things under control. Don’t get stupid with copper shims, you’re more likely to damage and crush the core rendering your machine useless than improve any thermals. CPU already has a vapor chamber cooling it, best you can do is apply liquid metal properly and optimize voltage on CPU and GPU

 

You're right and make sense though was it safe using LM in longterm? I been watching some YouTube vids about the longterm effect of LM in copper and while they don't damage it they do left a permanent stain.

What about adding copper tube on the existing heat sink assembly?

 

There is no room to add anything in this computer, is already super tight space in there. I’ve used LM for years on several computers I’ve never had a problem. My RBP has had LM on it since August and it runs amazing

 

Thanks! Are you using LM for both or just CPU?

 

Both

 

Thanks appreciated. Actually just repasted my RBP 2 weeks ago with arctic 5 though there is a significant reduce in temp its still wayy high when playing Assassins Creed Odysseys @1080p with setting set on ultra most setting except shadow+volumetric fog in high to medium setting i am able to achieve 55-ish to 64-ish FPS but CPU is 96C high and GPU @ 90C . Im using Intel XTU plus CAM monitoring software disabliabling pzperformance. FPS quite high and satisfactory not until CPu start to throttle therw I would experience lagging amidst high FPS count. I owned once Acer Triton 700 witg Gtx 1080max q and I could say RBP is a little way above it performance wise just the heats killing it.

 

Liquid Metal application was great. I achieved -20C less in overall temp playing cs go @ 70c ish on both cpu and gpu while intensive games like assassins creed Odysseys @ 80c ish on both CPU GPU and barely hits 85C FPS on afterburner around 65 to 70ish FPS on 1080p resolution maxed out.

Having achieved a good temperature from 90 to 95c before i noticed that the fan kind of over exaggerated in a sense that when thermal start to reach somewhere 75C in will boost all the way up like it was about to take off like there is only low and high setting no medium setting at all. On gaming im fine with it but browsing sometimes when it feel like its getting hot instead of cranking up fan a little it will go full blast. Any idea whats wrong?

 

How high was idle temperature when doing that full blast? There’s no fans curve setting on Blade Pro, but I saw it on my bios setting on Blade 15 after i mode cpu vrm to be accessible, didn’t try it yet but it shows at which temperature the Fans should kick in, if you’re familiar with bios moding maybe you want to try that.

 

It working normally now. I opened it up again rescrew and inspect if something i missed out but none and when i put it back on fans quite now. Idle temp at around 53C to 60C playing on both CPU and GPU. Before once the temp hit past 60C it will go full blast then at around 55C it will suddenly go on low like it was choked out of electricity from 100% down to 10%.

Although problem solved i noticed that fans will only full blast on 90C plus which could make use of bios mod since demanding games like ACO with LM applied im hitting 80C ish on CPU and 75Cish on GPU before LM it was 95c and 90C and a lot of thermal throttling occurs. Although temps been lowered internal component suffers from heat radiation .

Anyway i can mod the bios or better stay safe with what i achieved so far with LM?