*** Official Clevo P870KM1/P870KM1-G/Sager NP9876 Owner's Lounge! - Phoenix 3.0 ***

Correct.

Mine doesn't though, or I'm not understanding where the rubber bumpers are.

Either way, even with the tape in place (kapton see through), I'm still seeing a lot of potential air pressure being bled inefficiently outside the fins and/or exiting prematurely at the sides of the fins.

The tape shows evidence of air bleeding by the dust buildup at gaps.

Middle fan is a really good example of bad airflow seal.

 

Morning all. Currently about to upgrade from an older Clevo P150SM to this model. I have a couple of questions before I jump in. I have read a few posts on the current situation of throttling at 480Ws. I am looking at getting a single 1080 configuration. Would it be prudent to get two bricks for when the modded bios comes out? Does having just 1 single brick guarantee I will never hit the full potential of this machine? Thanks for your time.

Edit: Donald answered my question via chat. Thanks!

 

Ok so I opened her up again, I do have rather large gaps between the fans and the VC, I'll throw some tape on it and see if there is any change, good or bad

Sent from my XT1585 using Tapatalk

 

Ran the 3d mark fire strike ultra and passed again with SLI turned off, the vent coverings removed and the fan to VC vent taped - CPU max 61C; GPU #1 max 58C, GPU #2 max 56C. The difference in temps are CPU dropped 1C, GPU #1 dropped 7C and GPU #2 dropped 3C.

Sent from my XT1585 using Tapatalk

 

This is a new problem on the KM1. The DM-G and DM-3 did not have such stinking MASSIVE gaps above the radiators like that. I can't believe they released it like that. It's absolutely retarded. They should have made the radiators large enough to use that space effectively and enhanced the cooling, used thinner fans that were the correct height, or at least placed a block of extra dense foam across that space to block the air leakage. Where they have worthless strips of foam now, that could have been a solid block of foam all the way across in front of each fan. I really don't see any point in the thin strips of foam being there. I don't know what purpose they are supposed to serve. It doesn't make any sense.

 

Holy mother!

I thought the DM3 gaps were bad, but this is just on a whole another level @Meaker@Sager !

Seems you got the late pattern heatsink, my KM1 has the early pattern "2 pipe bridge" that is used on the DM3.

Not thinner fans, I think a third foam block on the horizontal plane placed on the fan body would be okay.

 

Yeah, I posted something about this already and hardly anyone noticed. Don't remember what thread, but yeah... that is just totally nuts.

Photos below is how I plugged the holes on the machine I had available for tinkering for a couple of weeks. When I saw it my jaw dropped and I was like OMG, LOL!!!! What on earth were these people thinking?!?

Then, I thought about it for a few minutes and it seemed apparent that NOT THINKING might actually be the underlying issue. Maybe they hired someone from the Alienware design team.

 

Is that the KM1?

The top fin portion should be thicker due to the Vapor chamber being sandwich in the middle!

 

Yup. Sadly, it is. Not cool (pun intended). They need to ditch that extra-stupid 3rd revision of the vapor chamber. The first one was the best.

Other than the retarded Dellienware tripod piece of crap, I don't think I have seen such an idiotic engineering blunder. It's like some haphazard backyard slop job made like that. Maybe @Eurocom Support and @Prostar Computer and @Meaker@Sager can get those people to pull their heads out. Maybe @Prema can talk to the people he knows as well, but there is just no excuse for that level of half-assed amateur design to see the light of day.

 

What is that, that you used? I just did a quick fix and threw some high temp tape on it to see if there were any differences. And behold between the tape and the removal of the vent shield or whatever that crap was they used to cover the vents on the bottom plate, I dropped around 3-5 degrees C on both GPUs.

Sent from my XT1585 using Tapatalk

 

Yup, little things can make a big difference, and when they keep cutting corners and getting sloppier and sloppier, it really starts to add up. There is absolutely no excuse for that kind of half-assed nonsense.

I used some extra-dense foam like what they use for padding the inside of boxes with computer parts. I put some double-sided Scotch tape on it to hold it in place on top of the radiators.

 

There's a 3rd revision?

I thought only early pattern 2 pipe bridge and late pattern 1 pipe bridge?

Or is the 1st version the early pattern but shiny underneath?

 

Well, I am assuming so. If your KM1 is not made like what you see in the photos, and it is not made the same as the one in my DM3, the only conclusion must be there is a third revision. I only said third revision because your comments suggest your KM1 is not made that way. Had you not acted surprised by what you saw, I would not have suggested that they changed the design a third time. It really doesn't matter how many versions there are... bottom line is, what you're seeing now sucks.

 

My KM1 seems to use the DM3 heatsink/VC.

edit: I think it was bloodhawk that mentioned the DM3 two pipe bridge heatsink had a "shiny version" and a "dull version"

shiny one is supposedly better?

 

That's great. They all should. That defective new design should be retired and pulled from shelves immediately.

Edit: Bear in mind @Mobius 1 - that air gap you are seeing shows how much smaller the radiators are compared to the previous version. I challenge anyone to explain how that can be viewed as a good thing.

 

Manufacture cost is less

And clevo at one point said that 97c was ok with the GPU?

 

Maybe I should have qualified that by challenging anyone to explain how that can be viewed as a good thing from the customer's perspective, LOL.

 

I noticed that this one has a copper plate that runs the length of the green line, would there be any benefits that would come from making a small copper plate to sit where the blue square is? I am not very computer savvy when it comes to mods, but I know with cars something like tape/foam would cause turbulence and potentially make the air flow worse.

Sent from my XT1585 using Tapatalk

 

with yours for now, try focus on closing up the holes first.

 

Missed this one personally, I am asking.

For now if you have an old finned heatsink you can cut down, flatten and thermal tape you could actually increase the proper effective fin area.

 

Currently the holes are closed with high temp tape, I am going to grab some foam like Mr Fox did, but I want a fix that won't hamper the air flow if the tape/foam is causing turbulence.

Sent from my XT1585 using Tapatalk

 

So why doesn't that happen on factory?

Hmm...

 

It should not hinder the air flow or create a lot of turbulence. It might have a little bit, but the air will take the path of least resistance (through the radiator fins) if the holes are plugged. The way it came from the factory, the path of least resistance would have been through the gaping holes over the top of the radiators.

 

I am not responsible for the design or the manufacturing at Clevo This is why I am asking.

 

Patiently waiting.

 

Anyone that may be thinking about buying one of these beautiful beasts, if you are a student you might like to know that from now until July 31 Eurocom is doubling its normal everyday student discount from 10% to 20%. http://www.eurocom.com/ec/student()ec - that's a pretty sweet discount.

 

That is a big of useful information, thank you very much for informing us, @Mr. Fox !

 

Ill be putting up my 2 x Extra VC heatsinks (Original DM3 ones) for sale soon, if anyone is interested. For a super cheap price if i may add

 

Is it compatible with p870km1?

Sent from my XT1585 using Tapatalk

 

Should be. Maybe @Mr. Fox will know.

 

How much are you asking for one?

Sent from my XT1585 using Tapatalk

 

Can you share what the answer was?

 

I believe it will fit, but I never tried swapping them out. @Mobius 1 has a KM1 that may have a first generation vapor chamber. It does not have the huge gaps over the fans. Maybe he can verify that, as I could be mistaken.

So, the best answer I can give is "I think so" since I haven't tried it. I cannot see any visible reason why it would not fit and I have had both machines apart. I don't have a KM1 in my possession right now to try swapping them.

 

Eh I figured no one would really know for sure lol.

Sent from my XT1585 using Tapatalk

 

Yeah, I am kicking myself in the butt for not trying it while I had a KM1 in my possession for a couple of weeks. I was planning to for my own curiosity and got busy with other things, then it was time to send it back. There is nothing I can identify that is different in screw layout or dimensions. The only visible differences that were obvious are the motherboard having an extra M.2 slot and the metal plate under the keyboard designed to accommodate that change. The power button switch is different, but that is not visible from the outside. That said, there could be subtle differences that one might not notice at first blush, like the pockets in the chassis where the radiators fit (for example) could be different. I really doubt it, but without checking I can't say for sure.

 

Sorry - I had read the following posts (from earlier in this thread) incorrectly before I had asked the question:

The answer was: Based on estimated power draw the 330W power supply should provide ample power under most if not all conditions for a single GTX 1080 card configuration. 200W(GPU)+91W(CPU) = 291W total with 39W leftover for everything else.

The above threads, as I understand it, discovered that Clevo set a hard limit of 480W total power draw thus preventing a machine configured with 1080s in SLI from ever reaching its full potential utilizing the stock BIOS. (please correct me if I am wrong). Cheers!

Edit: clarified where these posts came from

 

And to add, the reason why I was concerned is on my old laptop (Clevo P157SM), my 780m was power starved from the 180W power supply. Once I got and modified a Dell 220W power supply, I stopped getting dim screens etc from when the card would spike. It appears that issue was avoided in future laptops.

 

How is there not an intake grill over the cpu fan?

Sent from my LG-H918 using Tapatalk

 

I'll be stalking the marketplace.

DM3 heatsink fits in KM1, no problem.

39w leftover for the rest of the system is not that much.

There is.

The intake grill that is blocked is the middle fan near the power connector.

 

Okay guys, I managed to install it.
I'm not sure what to think as it says B173QTN01.4 on the back of the screen, EDID says version 1.4, but panel reports itself as AUO 1296 instead of AUO 1496 like expected.
Anyways, I installed it while the laptop was in sleep mode in case I ran into whitelist issues in the vBIOS. It started up fine and looked normal, totally uniform backlight was a pleasant surprise lol.
No interlaced banding at all!
nVidia Control Panel still reported that G-Sync was supported although this is probably because I installed it from sleep mode.

....and then I shut down the laptop completely.

Turned it back on and now the screen ONLY works from BIOS menu/Style Note splash, but turns off as soon as I enter Windows.

Conclusion: No problem, I'll just carry both screens around with me along with my trusty plastic pry tool and philips M2 head screwdriver and boot from 01.0 then sleep mode and replace with 01.4!

 

Does safe mode display?

 

See if the screen works in Safe Mode. If it works in Safe Mode, DDU the drivers, reboot and reinstall the drivers. It's not uncommon for goofed up or confused GeFarts drivers to turn off the backlight. For a while I has having issues with that on El Cucuy after upgrading to an aftermarket 120Hz screen, then NVIDIA finally got their act together. Before they corrected their issues I had to use only older drivers because their "latest and greatest" garbage made the screen go black.

 

Ah, so the screen indeed does work on normal boots after uninstalling the Geforce drivers, and works in Safe mode due to lack of Geforce drivers.
DDU & new driver install unfortunately doesn't fix it - as soon as the driver is installed the screen goes black.
I noticed that without drivers the screen will run at 64 Hz and will have occasional vertical blue lines, didn't have that with my 01.0 revision panel.
If I do boot with the Geforce drivers and then swap the panel in sleep mode, the 01.4 panel has no issues. Working brightness adjustment and everything, even the Fn+F2 backlight off key combo works to turn it off and on. I just tried G-Sync in World of Warcraft and it is working as well.

The panel unable to start issue appears to only affect full reboots/fresh power ons/fresh driver installs.
I reallllllllllly realllllllllly want this screen to work, as its backlight uniformity really is pristine and the color production is better too, my 01.0 revision I was drastically reducing blues (down to 70%) but on this one I'm only slightly reducing the blues (down to 95%).

Let me know if any more information can help, here's some panel info from madVR:

Code:

edid version        1.4
manufacturer id        AUO
manufacturer name    AU Optronics
product code        1296
data string        JYWWFB173QTN
manufacture date    2016

raw edid data
00 FF FF FF FF FF FF 00 06 AF 96 12 00 00 00 00
00 1A 01 04 A5 26 15 78 02 9D 75 A4 57 56 9C 27
0C 50 54 00 00 00 01 01 01 01 01 01 01 01 01 01
01 01 01 01 01 01 28 C3 00 A0 A0 A0 5A 50 30 20
35 00 7E D6 10 00 00 1A 28 C3 00 A0 A0 A0 54 56
30 20 35 00 7E D6 10 00 00 1A 00 00 00 FE 00 4A
59 57 57 46 14 42 31 37 33 51 54 4E 00 00 00 00
00 02 41 2F 28 00 01 00 00 0B 01 0A 20 20 00 F0

The AUO 1096 01.0 reports a serial number, this one (AUO 1296 01.4) does not.

I'm tempted to maybe try writing my AUO 1096 EDID to the AUO 1296...

 

Try installing some drivers that are older, from like about Q3 2016 or Q1 2017 and the screen might light up for you. That's what I used to have to do when I first installed my 120Hz screen. 375.70 is a good one.

 

No luck with the older drivers, @Meaker@Sager are you able to fwd a request to Clevo to include better support for this screen? Seems unlikely given that the EDID has a Dell part number programmed into it. AFAIK it works fine on AW 17R4 with latest drivers, so it likely is something vBIOS related? I'm very hesitant to try flashing Alienware GTX 1080 vBIOS on a Clevo manufactured card unless someone has tried this before?

 

I have one coming in as well, later this week (hopefully).
AFAIK these were specifically made for Alienware's since i have not seen / heard of these being by used other OEM's. But i could be wrong.
Check with @t456 , if there is any way to splice things together using the 01.0/01.1 panels firmware dump and modify, this panels firmware.

 

No.

Safe mode will work. Tried multiple driver and no go :|

 

Well shoot... that sucks. Wonder what the deal is? Is it just me, or does anyone else think an LCD panel being able to light up having a dependence on NVIDIA drivers just stupid or what? It's so retarded that it makes me want to say really nasty words that my mama would have washed my mouth out with soap if she ever heard me say them.

 

G-sync module not recognized for certain laptop combination I think.

Similar to INF file mod needed to make an old laptop with new mxm card upgraded.



Maybe it will work if I take out the slave card?

 

It might. Try it and see.