Clevo Overclocker's Lounge

There are no components under the SSD in P775 actually at least not if you install it in the HDD bay along with the other large HDDs.

It looks like I found a thermal pad somewhere, but I don't understand if it is supposed to isolate the SSD from the rest of the laptop or make the SSD cooler

This is the only product I found in Romania, costs around 8$

http://www.shop4pc.ro/set-paduri-termice-silverstone-tp01m2-pentru-racire-pasiva-ssd-m2-p-46737.html

 

Speaking about M.2 only; the ones that need such things.

 

Maybe @iunlock can give you a good advice what pads/heatsink you could use for cooling your M.2 ssd's. He had same machine as yours. Or ask in the appropriate thread for your machine.

 

I ordered a pair of those, for any event possible.

I'm experiencing something iffy, just installed windows. Everything is swimmngly fast, but HWiNFO detects two temperatures on the SSD, one that is around 70C in idle and reaches 80 max, and one that is around 80 in idle and reaches around 97 max, any idea what is that?

 

I mean newest Hwinfo have two sensors for NVMe drives now

• Extended reporting of NVMe drive temperatures and added thresholds to sensors.

(don't remember who of these read current temp). You could either download Samsung Magician, older version Hwinfo, Hwmonitor, Aida64 or speccy for control check of temp for your new NVMe Samsung ssd.

 

I really have no idea about the double temps. My temps usually are all low.

 

Something something portable air conditioner arrived and now idle temps are humane - 50C on the main sensor with 70C on the secondary sensor for the SSD.

I have a question or two though - I'm getting poor write speeds (AROUND 300 mbps) on it and I am also still slightly concerned about the temps from earlier.

@D2 Ultima , how are you getting low temps on SSDs? Is there a magic trick that would help with this?

 

P870DM3 is airflow central

P775DM3 is not

Sent from my OnePlus 1 using a coconut

 

You could run ATTO Disk Benchmark a couple of times. Post results and screenshots of temp. And ssd brand.

 

Samsung 960 Pro 512 GB

Oh... You're right...

 

Has anyone tried this thermal compound?
I were thinking of buying it for my GPU's on the next teardown.

Phobya NanoGrease Extreme 3.5g 16W/mk

 

@Mr. Fox has it, we might see something this weekend.

 

Hi, I have that grease.
W/MK is misleading.
Thermal Grizzly Kryonaut cools better than it.
With nanogrease, at 4.5 ghz with 8 thread chess engine, temp difference between core 2-core 3 is 5-6C, sometimes 7C.

With Kryonaut, temp difference between core 2-3 is 3-4C (at worst, 5C, but haven't had that lately).. Overall temps are a few C lower with Kryonaut.

I think Nanogrease Ex is relabeled Gelid GC or something, although someone said Phobya's other previous traditional thermal compound was GC relabeled, so I really don't know.

It's not bad stuff. If you cant get Kryonaut for a good price per gram, go ahead and buy it.

 

The extra fan you added to the inside of your P870DM1; where did you connect it to? or does it have its own external power source?

 

Ran them.

 

Yes, you will. @Papusan is waiting for it also.

 

Second reading looks like it might be controller temp. You're gonna need to use the thermal pad and if that fails a plain heatsink should work.

@Mr. Fox can you check the un-modded 960 Pro in your DM3 in a similar fashion and report the temps via HWiNFO like this here? I'm interested.

 

Dang, those crazy high SSD temps are "Alienware-Grade" overheating insanity, bro.

Here are the BPX and 960 Pro temps in the DM3. I'm guessing Temperature 2 must be the controller since it is warmer. This sensor does not work on the BPX. It always reads 88°C 24/7 (even when machine is started cold) so I have it disabled and hidden on the BPX.

 

Was this taken with auto or max fans?

 

Auto fans. In fact, I never heard the internal fans even turn on.

 

Can you run the 960 Pro one again with max?

 

Attached. Ran 960 Pro and BPX again.

 

I connected them through the rear USB port, they are 3/4-pin pwm fans connected with usb adapter.
The link here should show you a better image : http://forum.notebookreview.com/thr...0b-alive-thread.795048/page-152#post-10548944

 

Interesting, not much difference in temperature. In my case I can get a decent few degrees reduction in normal operating temp going from auto to max. So that's interesting.

 

F$£K yeah. That next level

 

Similar results to mine. Mine with idle fans in the above.

 

That's interesting. You guys have anything placed on the SSDs? Like shims or anything like a cooler? I might try to return the SSD if this keeps on like this.

 

There seems to be throttling on my setup and I moved the SSD module to the large components bay, from the HDD/SSD bay. Now the temperature has dropped by a few degrees, first temp idles at 60C while the second at 80C, but those temps are stil way too high and will throttle as soon as I run a benchmark. Looking for some advice on how to reduce the throttling and what else I can do. Is there a chance that the SSD is a DOA, or those temps are normal on 960Pro if it doesn't get enough air and if it doesn't have anything to help disperse the heat?

Ambient temperature is about 20C...

Edit::: Temps right now

 

See my suggestions for cooling in one of my previous posts.
And please, ask about what NVMe ssd temp other get in the proper thread for your machine. Post the ATTO benchmark link and ask other if they can try same ssd bench with Hwinfo for temp monitoring.

 

I'm not curious about how it should react in P775, but curious how it should react in general, if it is out of the driver's specs then I might as well return it.

Also, I can't seem to find the proper thread, do you happen to have a link?

I am curious because I'm trying to get OC levels of SSD speed but failing, I should search for something something cooler thingies in Romania. How do you guys glue them on the SSD?

 

Okay, seems that I found something, now trying to figure out how to order it

http://www.extreme-computers.ro/Detalii/Xilence-COO-XPRAM--VGA-Ram-heatsink-e8/

 

Those wouldn't work. Too thick I think. You need to find how much space (headroom) you have above the ssd's for your machine. You would probably not be able to close the lid if the heatsink is too thick.

 

SHOWDOWN: Thermal Grizzly Kryonaut versus Phobya NanoGrease Extreme

OK, @Papusan (and others). Here are the results you have patiently waited for.

Very minor improvement with Phobya. However, it is somewhat thicker and heavier than Kryonaut and might prove to be more durable and less susceptible to pump-out. I say NanoGrease Extreme has a tiny edge over Kryonaut, but so small as to be almost irrelevant. I like the heavier viscosity better than Kryonaut and it might work better than Kryonaut where sloppy fitting heat sinks are an issue. It was applied the same way as Kryonaut, spread in a thin layer using the provided spatula. Both were fresh applications of thermal paste (done immediately prior to testing each product).

I have no explanation for the slave GPU being a little bit cooler with Kryonaut during extended (and more brutal) 3DMark 11 test, but the results are close enough with both pastes across all three benchmarks to be within a small margin of error. I could speculate, but that would be worthless.

Spoiler: Benchmark Screen Shots

Notice watts and percentage of TDP using 3DMark 11 versus Unigine gamer benchmarks.

 

Thanks Fox. A interesting option for people who don't use Liquid metal/lost in the Clevo HS lottery. + A good option for use on graphics.

 

You should try this test on a CPU with multiple cores, testing the difference in temps between the hottest and coldest cores, which is probably a better test than a single core GPU, however if you have liquid metal on the CPU, this may be unfeasible.

What I found in my own tests, which i have done consistently (I did two tests) is that on a BGA core, using the X method (not recommended, spreading is better), the nanogrease extreme had up to a 7C temp difference between core 2 and core 3, while Kryonaut had a 5C difference. The coolest core (2) seemed to be about the same on both, but the hottest cores (3, then 1) had too much of a difference.

When I switched to spatula spreading, Nanogrease dropped to a 5-6C difference, while Kryonaut was a 3-4C difference. So for me, Kryonaut was better.
I tried three attempts total with Nanogrease (two spreads, one X) and I was never able to drop the core difference down to 4C, no matter what I did.

I did have a problem with early applications of Kryonaut 'degrading' after a week or so, but that hasn't happened with the last application. Of course, summer 100F temps throw spanners into everything, but I did some things differently on the last application, like applying a little extra to be sure, and trying to apply the 4 heatsink screws more evenly in an cross fashion without full turning. (edit) of course I did full turning at the end, but in an even fashion. The nanogrease had the same "issue" as well but again it's really difficult to test with a weak AC and gigantic ambient temps. But for me, I'm satisfied with Kryonaut.

 

Amazing results over the kryonaut...impressive!

 

So i will try this paste or maybe LM together with Prema magic (i hope waiting time will soon be over) for my next attempt (aiming for second place though)
http://hwbot.org/benchmark/cinebenc...Id=processor_4390&cores=4#start=0#interval=20

 

OK, as promised... here is a comparison of Kryonaut versus Phobya NanoGrease Extreme on El Cucuy (Tornado F5/EVOC 16L-G-1080)

 

Thanks for the explanation and the tests and all those material, they re really important for us, thank you so much! Great job!

 

Hard voltage mod is easy. I'm surprised they aren't done as often as they used to be. Just mod the voltage feedback or reference voltage circuit like mods a decade ago.

From what you told me about the programmer reading a short at the BIOS chip though, getting it to work again may be difficult.

 

The chip has more ways of measuring its activity so you can't just volt mod it like that.

 

I should had shipped you the card to do voltage mod before it's dead, lol. Fingers crossed my gpu core is still alive!

Hard voltage mod on Pascal came out awhile ago: https://xdevs.com/guide/pascal_oc/
I remember @bloodhawk wanted to do it too in the past.

 

Almost did, kind of, but then the other card (untouched) died, so i kinda put a lid on it. But its super simple to do as stated above by @Khenglish .

 

Super simple to have trim pots in your laptop?

 

Yes.

 

Wow so is the recommended traditional paste for overclocking Grizzly Kryo? I was under the impression it did not last long though.


Sent from my iPhone using Tapatalk

 

It takes away from it being a notebook at that point though, that's really more desktop then as it will never look tidy or be sturdy and easy to use.

 

It all depends on how well you can hide and stick the stuff. Super easy using Hot Glue / Kapton Tape / Electrical tape.
End of the day the DM3 isnt really a Notebook. Its a portable system that stays on the desk until someone wants to relocate it.

 

Alienware grade overheating


With Dell SSD plate + full system load (CPU turbo disabled)






Without dell SSD plate + ssd load only (system failed to complete SSD run before BSOD due to SSD safety cutoff)






I had to argue with support for them to send me the plate, even though they acknowledge that the SSD was overheating

 

Depends on your use case, maybe for you but I move it around a fair bit.

I always try and keep the original spirit of the machine but certainly each to their own.