*** Official Sager NP9877 / Clevo P870TM-G Owner's Lounge! - Phoenix 4 ***

It's not. If you check the several disassembles that are online the heat-sink look beefy.
If it's better than other DTRs, reviews will tell.

 

Compared to the vapor chamber in the p870 series it wont stand a chance, that much is obvious.

 

I keep forgetting the P870 model
I wonder if anyone could custom build a vapor chamber cooler for the P7xx series

 

Interesting comment in a thread about that exact model.

 

The vapor chamber is the metallic grid at the end of the fan?

Enviado desde mi G8441 mediante Tapatalk

 

It's this, the big flat odd shaped black painted structure. A vapour chamber is like a non-pipe shaped heatpipe - using the enthalpy of the vaporisation of water to shift heat away from a hot surface (evaporator) to a cold surface such as one attached to a fan forced radiator (condensor, the grid thing you mentioned)



If you want to see the same principle in action, look up der8auer and his Novec fluid cooling experiments on YouTube

 

The heatpipes you see guide the vapor, they are not doing the heavy lifting.

 

Interesting Indeed! hahahahaha. ! "keep forgetting"

 

The vapour chamber properly setup can dissipate 500-600w effectively.

 

I meant the vapour chamber cooler for this model. MB.

 

Wan't help much if the last bucket in the thermal chain (fans and grillls) can't do the job remove the heat coming from VP. See... Out from the chassis. Remove 2x300w gpu heat in a notebook is a heavy task.

 

Talking about personal experience with it here and total system power consumption in the 650w range.

 

Vapor chamber is only for GPU? CPU tend to run always 10c more than GPU

Enviado desde mi G8441 mediante Tapatalk

 

Two different thing. I talk about 2x mobile 1080 "if running 300w each" aka 600w heat. Not the total power draw as the Vapor Chamber is only for the gpu's (Tm models).

 

Most of the draw would be the GPUs in my case.

 

Hey my laptop people, I know I should know this, but can someone remind me where the third M2 NVMe slot is? two under the keyboard...and?
Remarkably difficult to google it...unless its just me...

 

All three are under the KB

 

There you go then. I could have looked with my eyes...

 

Or your manual

 

There was a manual! No idea where that is...
Wow I’m on fire today.

 

Our website offers a download for instance. A concise version would be in the box.

 

Yep I just didn’t think. Too many things in my head at once. Thanks though...

I’m considering getting an early rtx upgrade via an m2 slot and egpu. I don’t know why exactly...I think I may have an obsession with just doing things to this laptop. I’ve got extra heat sinks over the cpu, and I’ve printed a thing to draw extra air through it via the spare gpu fan, and I’ve cut a new skin for over the keyboard panel out of textured rubber, and now I’m obsessing over rtx...

I have a problem.

 

I'm pretty sure the M2 port has not enough bandwidth for the RTX, plus you won't be able to use the laptop's built in screen.

On the side note, notebookcheck has the single RTX 2080 (laptop) ranked higher than SLI GTX 1080. Do you think that's accurate?

Enviado desde mi G8441 mediante Tapatalk

 

Yeah I was wondering that. Tbh I’m just bored and have cash. Not enough to justify however many thousands an mxm rtx will be though.

Really I’m on my way to building a mini itx I think...

 

Highly, highly doubt that. Single RTX 2080 is at best 35% faster than a single GTX 1080.

SLI GTX 1080's with average scaling should be 10-20% faster than a RTX 2080. Games that scale well, should result in performance >50% faster.

 

Generally frame times and driver issues are less so in terms of gaming a single RTX 2080 will give a better experience.

 

What filament did you use to print the shroud ?

 

Stuff called EX200, but keep in mind my shroud works from the OUTSIDE of the case only. So air is pulled from the spare gpu fan across the out side of the case through the holes directly above the cpu itself, through a third party copper heat sink, through small fin bank that is attached to the standard cpu heat sink, from the back (yes the hot wash side). So shroud itself isn’t by the hot stuff in the case. It’s roundabout, but the simplest solution that yields 5-6degree drop. Could be more if I tune the fan curves so they’re louder.
Initially I just cut up a beer carton and taped it on...
Worked better than a u3 cooler...

Edit: I should add that’s 5-6 deg (Celsius for the Americans among you) in gaming. Synthetic tests are less pronounced, like 3-4 deg...

 

Gotcha gotcha. I have two different designs , one mounts in place of the one on the CPU fan and directs air over the whole heatsink. The other is kind of like your idea, but I printed off a spacer kit I guess you could say to bring the bottom cover further out. But still being closed if that makes sense.

 

Bottom cover further out... from the gpu fan? So my idea but inside the case?

In an ideal world I’d trial a second set of heat pipes on top of cpu going left to radiator in from of 2nd gpu fan (from my Clevo p370em Radeon gpu’s) but who has time/motivation? Plus I’m dubious about effectiveness. I bought a 2nd cpu heat sink to trial it though. It lives in my cupboard.

 

Yep yep .It's nothing super big in regards to adding to the full depth of the laptop. But more space is greatly appreciated by the internals lol. I can't remember how much I actually raised the bottom cover. Somewhere around 3/4 of an inch maybe. But with printing off the cover extension you can add your own slope if you wanted to. I have mine a bit thicker at the back than at the front. Have even pondered printing off a gasket out of Flexi so when it sits on a cooling pad, majority of the air is pushed up under the laptop vs it just wastes or vents out from under the sides of the laptop.
You would be surprised at the amount of trial I have put into this laptop cooling shenanigan lol. And the extra heatpipe to Gpu is actually the rabbit hole I am currently going down now LOL. Trying to find another CPU heatsink at the moment

 

Surprised I am not. It’s easy to get obsessed. Yes I tried gaskets off a 120mm fan off the u3 but it only made a difference with the 3rdparty heatsink ducting out through that smaller cpu heatsink. That’s where this idea started. Ill post pics at some point...
Heat for me is more of an issue as my cpu chip is not delidded....however I’ll keep my warranty for now...

 

rtx2080 in clevo 17" : timespy : 9522 ( for reference: 8086K+gtx1080 timespy scores:7800 & 9900K+gtx1080sli : 13400)

and

 

How do we get our hands on a Rtx 2080 for a tm1-g?????
Welp . Answered my own question lol
https://www.notebookcheck.net/Lapto...aphics-coming-soon-from-Eurocom.402301.0.html

 

The SLI part is pretty misleading, but then again not totally unexpected considering the source...

 

You could run multi gpu in ashes... and that's about it, still not sli though.

 

Casting aside the puffery about being "the only company" to "manufacture" and "produce" the laptops it sounds like they are talking about a dual gpu workstation type system WITHOUT SLI/nvlink for pro apps that don't need it:

Yeah, hasn't explicit mgpu in games gone *absolutely* nowhere

 

Yes, because it's more difficult to code for.

 

I don't need but one LOL.
Anyone know if the 1080 dual vapor chamber heatsink will work on the 2080?

 

I don't think the Eurocom Workstation / mobile super computer or whatever they're calling it is targeting the gaming community. There' are other applications (both in the general and software sense) of multiple GPUs - http://forum.notebookreview.com/thr...1m-owners-lounge.826831/page-70#post-10857001

 

Though it's not SLI so that's them not understanding.

 

Maybe I'm misunderstanding your statement. SLI is for offloading graphics processing to multiple GPUs in order to generate a single image. Now, I am not an expert, not having done any CUDA programming myself, but am pretty sure CUDA over multiple GPUs doesn't have anything to do with SLI, as CUDA just allows stuff to be run on those extra GPU cores.

 

Non sli multi gpu: app is in control of telling each gpu what to do, completed work is sent back to main system memory. Gpu workloads are largely independent and equal and require no inter gpu communication therefore no need for a bridge (SLI/nvlink).

Compare that with dx11/sli/Xfire multi GPU for games, where the app (game) sees only one "Render device" driver's in control of the work distribution, and the output from the 2nd ("slave") must be copied to the first ("master") which handles the frame buffer and has the displays attached. This high speed interconnect used by Nvidia for this is the sli/nvlink bridge

More on sli/xfire: https://www.wepc.com/tips/difference-between-nvidia-sli-and-amd-crossfire/

Eurocom is doing the former while throwing in mentions of SLI, and gamers......

 

If you check the news release linked in the notebookcheck ad, it's more specific about being dual gpu with no SLI as a professional tool.

http://www.eurocom.com/ec/release(390)ec

Not really sure on the practical use, but an interesting way to promote the P870M following EOL.

Either way, the notebookcheck version using the term "RTX 2080 SLI" is offside.

 

Exactly. For a gene sequencing app, AI solution, big data number cruncher like a scientific weather model or crypto/hash engine having access to extra computational cores within the GPU to process data is a big plus. And having even more core resources thru additional GPUs means more data can be processed in the same period of time. Note that none of this has anything to do with rendering an image and spitting it out to a i/o, but rather using the GPUs' computational power and send results back to the main CPU.

 

There are a couple of niche cases.

 

Agreed. Although one could argue any CUDA based application is in fact a niche. That's where I would even place crypto-miners.

 

Not sure if this has been done previously. Haven't read through all 300 pages in this thread. . . But I'd figure I'd share. I spec'ed out and installed a fan for the empty position on the cpu cooler. It is from a lenovo laptop. It is a Delta P/N BSB05505HP-SM. It draws 5V and 0.4 Amps. Airflow is about 3 CFM. 8mm tall by 56mm wide and deep. It fits almost perfectly. I built a few brackets out of scrap plastic to interface with the existing mounting holes on the cooler. The connector on the fan needed to be rewired to match the pinout of the 5V fan header on the motherboard. I have it wired to be at 100% speed all of the time because it is actually pretty quiet and having it running keeps the idle temps way down. I was able to decrease my undervolt to 75mv and idle cpu temps are 29-31C. Previously with a 125mv undervolt temps were 33-35C at idle. (Idle temps are with 12V fans also running maxed as well)

I was also able to achieve a more aggressive OC and gained about 40 points in Cinebench - not a whole lot, but the extra cooling absolutely made a difference, was able to feed the 9900k a few more amps.

 

+rep. If you run same volt and clocks... What temp difference will the extra fan give in full load? Idle temp isn't that important. Thanks

 

It's about 4-5C difference. Previously I was sitting at about 72-73C with a 125mv undervolt. Now under full load it is about 68-69C. All other settings are stock. I unfortunately don't have recorded OC temps previously from before the mod . . . I can try to recreate that later.