Clevo P1x0HM users with 7970M/680M - Ordered? Installed? OCed? Lets have it! :)

Doesn't seem, but for the time being i bench the card and i haven't played on a long period yet ...
when i talk of consumption and temp on crysis2 it was with 3 run on all map of adrenaline benchmark maxed out ...
At this time i am not sure that my card is definitely stable at this frequencies, it require more testing.

 

Could anyone please post a Clevo 7970M HM vBIOS with 1.05V full load voltage and 950/1400 clocks? I would like to try these settings in my GX740. For some reason, my card runs at 1.0V on full load and any use of MSI Afterburner for OC causes heavy artifacts. The card does not overheat, with my laptop cooler on it heats up to 72C in Furmark. Please remember, that it has to be the HM vBIOS.

Thank You in advance.

 

Hi Tonrac,

Thanks a lot for the guide....you just made my life a lot easier. And thanks to everyone else who replied to the driver question too.......thanks a lot.

Cheers

 

That's three now with different load voltage than reference.

 

For what concerns me before modding bios, GPU temperature was below 80C. Now during gaming it reaches 90C easily on some very intensive GPU games. Yes it does start throttling at 90C dowclocking slightly. It comes from 902 to 888 and in some extreme cases if the heating is persistent gets to 836Mhz. After the fan reaches the max speed the GPU starts cooling, leading the temp down to 85C. The clock comes back to 902Mhz till it overtakes again 90C. Dowclock protects from overheating pretty good. Apparently the downcloking doesn't slow down pretty much the game that keeps beeing fluent.
In any case, i will try to repaste with IC diamond 7 and modding with the aluminum tape to check if i can run cooler. During the benchmarks, like 3dmark11, the gpu doesn't have time to throttle cause from one graphic test to the other there is the loading pause and has time to decrease temperature. This happens only during some playing sessions with games set to ultra all the way.

P.S. During repaste, is better to put a pea of thermal compound and press the plate to spread it or manually spread the compound for all the surface of the GPU?

 

So how much should a gtx680m kit(card and heatsinks) for my P150hm cost?

 

Upgradeyourlaptop sells them for $790USD right now, not including shipping and any applicable import fees or taxes. There is also no mention of a heatsink so I assume that is not included as traditionally with their MXM upgrades they were not included. I'm not sure what other resellers or hardware sites are asking, but hopefully you can find it cheaper than that.

 

I think you just need to put the heat sink on it to spread it

 

It seems that the pea size method are the best way according to that topics http://forum.notebookreview.com/sager-clevo/680771-re-pasting-methodes.html

In my case, i have manually spread the thermal paste on the gpu. But i admit that spreading IC24 is not an easy thing to do because the paste is compact and not very fluid. Some says that the pea method is better because this is the heatsink that will spread the paste and so there is less bubble air.

@Religion: If you live in US you can buy the card @RJtech for 800$ and the heatsink is included
Laptop Video Cards :: nVidia GTX 680M 4GB Mobile Video Card - R&J Technology, Clevo Barebone Notebook kits, Laptop and desktop system builder

But if you live in europe, you can also buy you card at RJtech but you need to add to the 650€ price, the cost of delivery (between 70 and 100€) and the Vat of 19%, so the price could easily jump to a 900€

The price at laptopmonkey drop to 826€ (no Vat to pay and ship inc) and i think there is no heatsink
eBay | Nvidia GeForce GTX 680M 4GB DDR5 for clevo eurocom etc upgrademonkey

Or there is the premashop with one card left ! (heatsink included) good feedback and superb business : http://forum.notebookreview.com/marketplace-questions-member-feedback/681030-premas-feedback-thread.html#post8756778

 

Every bios settles.

The 1.76v bios settles at 1.05v.

 

so i have made some first test results with power consumption and battery life on my 7970M compared to the 485M i had before it uses less power at idle (27 vs. 35W) and i get about 30-40 min.more battery life (2 vs. 2.5 hours approx.)
also, idle temps of the amd card are the same as with my old gpu before@37-40C but my fans arent spinning up as much anymore, thus the machine is way more quiet. i think thats cuz the amd card isnt heating up as fast as the nvidia card (temps would rise waaay slower,even at idle,thus no quick oscillation between fans off/on at 37-40C)

cheers!

btw: would u guys want me to include the gpu upgrade guide on the OP? its applicable to the amd card as well so it could surely help a lot of HM users of course with all thanks to where its due@tonrac

Sent from my Galaxy Nexus using Tapatalk 2

 

Another thing to be aware of is the importance of using a method of earthing yourself.

I use an Anti Static ESD Cable Kit that includes a Grounding Plug, Wrist Strap and 12ft cable. Many don't bother but why take that risk with our expensive equipment, it makes no sense.

Many seem to forget that you can also cause damage to a component that slowly causes it to fail or that is not an immediate problem.

 

@Jaybee :It could be a good idea ... and i don't say that because it's my work

I found this video to complete the tutorial and much more (Its a P150EM but you know that they have the same chassis than the P150HM):
Puget Systems Traverse Pro M550i Component Replacement - YouTube

@Laptopnut: very true, i admit that i haven't got this, and (you will laugh) i take my socks off in order to have my feet directly on the floor, and i touch a metal piece in order to evacuate static electricity from my body, before manipulate electronic component.
I don't think it's very efficient but it's my ritual.

You'd better follow laptopnut advice

Btw i notice the same level of temp and power consumption as jaybee with my GTX680M, i will test battery life later.

 

Thanks for the tip. Had already ordered one...should receive it today....like you said...better safe than sorry.

Cheers

 

lol @ 1.76v

 

I know Malibal sells them for $800 which included GPU, Heastsink and a tube of IC Diamond 7

NVIDIA® GeForce® 460M GTX 1.5GB GDDR5

Thinking of getting mine from them.

EDIT: Dont worry....the link provided is the right one regardless of what it says.

 

With the back off + cooler max temp was 80C.

 

The Saltius vbios is not stable on my GTX680M ... crysis 2 shutdown my laptop after 25 min of game. Not an overheating issue cause it reach only 86°C before the shutdown. I am the only P150HM owner who has tested the GTX680M so it will be hard to know if it's due to the laptop conception or if my gpu/laptop has damage.
I revert back to original vbios, and crysis 2 is now stable at 853/2400 (2 hours testing).
No more success with slv7 because the idle issue is back (when i launch a 3d App voltage remain at idle stage and frequencies goes only to 324 Mhz so i cannot reach 3d clock) ... Now after reflashing with saltius the idle is here too.
Don't understand why saltius was working before ...
Only the stock vbios allow me to OC to 853/2400 without issue...

 

maybe re-install the driver after flashing an other vBIOS...also both mods are based on MSI vBIOS and not Clevo...maybe request Clevo vBIOS mod.

 

I have done a lot of driver reinstall, test 304.79 304.92 and also the 302.77 provided by clevo. No success. I know that the modded vbios is based on MSI vbios, but a lot of user with clevo card use it. And the fact that i can use saltius during a certain time seem to prove that i have a problem with the combo P150HM/GTX680M and OC/OV bioses. Maybe P150HM couldn't feed correctly the GTX680M OV/OC (i doubt that cause the 7970M require more power to run when OC).

 

Meaker, that's my run with an OC'ed 7970M. It's around 15% slower than yours. The new 3740QM CPU could make a little difference like the newer 680M drivers. I plan to flash the new overvolted vbios to the card in the future, if i could be stable at 1050/1560 i would be moew than happy. Anyways 680M is a monster) Great run indeed!

 

I got a different behaviour than yours. I played crysis for more than 1.5 hours, but not system shut down occurs. On the contrary I got an issue with temperature that reaches 90C. This at 902/2200Mhz stock frequencies. There is a frequency slight downclock bringing the core to 888Mhz until the fan leads back the temperature to 85C.
The GPU frequency comes back to stock frequency and it keeps going on till temperature reaches again 90C.
So, just a slight downclocking for me. I didn't push the clock much further, cause i know the temperature easily reaches 90C and starts throttling. As I said I will try to tape mod and repaste the GPU to check this can help and push much more. So far anyway, at stock clocks i can play each game pretty good with just a little and not frequent throttle. The only game that really cooks the GPU is ghost recon. That really gets the core on fire dangerously.

 

Cough...PSU...cough...

Did you check actual draw from PSU at the outlet?

 

Probably ! With 1037Mhz/2400 I have a peak of 195w (but average stay below 180W) so i put the card at 902/2400mem and launch crysis 2, after 25 min of game and before the shutdown, the max power consumption was 177W.
BTW the card seem "rock stable" with stock bios and OC 853MHz/2400Mem. I hope that testing saltius don't damage the card or the P150HM. I hope that some other owner of P150HM with GTX680M (but at this time it seem that i am the only one) take a try with saltius. It could help me to determine if i have a failure on my card/P150HM or if it is only due to the conception of the laptop.

 

I'd say the 680M is very picky especially when you overvolt it - you probably need a 220W PSU (that actually has a slight higher voltage) to maintain your o/c and o/v, especially with such demanding games as Crysis.

Unlikely Saltius vbios would have damaged your card, no worries.

Also keep in mind that these P150HM were not designed to accomodate such a card and even though the new gen GPUs seem to consume less and draw less heat, they are still very picky voltage wise (cf. Meaker's 7970M in MSI 16F2)

 

But would it shut down if it is not receiving enough power , it could just downclock also he did say Max was 177 W so it couldnt be the psu , also if it the psu is rated at 180 W then it should be able to do 180 W so it might be something to do with the repadding he did.

Tonrac you try making sure the pads are making contact or try the stock pads see if it solves the issue

 

To fully utilize the 680M with Saltius vbios I think you need a bigger PSU than 180W and most importantly a faster CPU than 3610QM since these clocks are bottlenecking the CPU

 

Since you are using a Sandybridge CPU it might have been drawing more W than the Ivy bridge CPU, there you might need a bigger PSU.

 

I guess you might need a bigger psu since the effenciency level is V so its around 87% efficient


So will this psu work on the p150hms?

Clevo Laptop Parts :: P150HM/P170HM :: 220W AC adapter for Clevo P170HM/P170HM-3DE - R&J Technology, Clevo Barebone Notebook kits, Laptop and desktop system builder

 

Thanks for all these replies

I think that PAu1ow is in in the true ...
@killer: I first think that the issue was due to re-pad with phobya thermal pad so i had re-install the original pad (my first issue was that the Gpu "saltius bios" keep his idle clock of 324Mhz when 3d apps were launched) .
After that i flashed with saltius and made a lot of test (3dm2011 , 20 minutes of Unigine Heaven benchmark, Crysis 2 adrenaline benchmark) but no real game.
Result of these tests are that the card perform extraordinary well and seemed stable.
So i launch crysis 2 to confirm the result of adrenaline bench, and this is when the problems began

I confirm that thermal pad are well applied.
For the P170HM PSU, she is incompatible because the connectic is not the same as the P150HM PSU. But if you have an idea of where i can buy a 220W psu compatible with P150HM it could be great.

 

How thick are your pads tho? I was using 1.5mm to make sure they are well contacted.

 

It could be the psu but it doesn't make sense that it would shut down when your max was 177 W on crysis 2 and then when you were benchmarking it was 195 W , also these 2 articles show that the max they recorded was near 170 and 180 for the gtx 485M which runs as 1.03 V :

Power, Efficiency, And Battery Life : MALIBAL's Lotus P150HM: GeForce GTX 485M Gets Its Game On

Review Schenker XMG P501 Gaming-Notebook (Clevo P150HM) - Notebookcheck.net Reviews

I get that the issue might be psu but things don't point seem point to it , also look at HTWingNut's results with the GTX 680M, he had a max comsupmtion of 189 W, granted he is on the newer system with a Ivy Bridge but the PSU is the brick that P150EM and HM share , Their efficiencies should be similar:

http://forum.notebookreview.com/sag...em-review-lots-680m-benchmarks.html#peakpower

 

The phobya thermal pad? I have took the size of all the original pad (1 mm memory, 1.5mm for vrm and 2mm for those in the border) and then subtracted 0.5 mm (0.5 memory, 1mm for VRM, 1.5mm for those in the border) Then i have put the heatsink and remove it (you could see some mark on the pad if they made correctly the contact between heatsink and components).
I must say that with phobya, the card was stable with the original bios and 853Mhz/2400 clock, temp were correct.
But like i said, i had re-pad with the original thermal pad

@killer: What you point make me think that the problem is more complex that only replace the PSU ... That 's why, i hope some other P150HM owner will test GTX680M in a near futur.

 

Yea I noticed with 1mm pads on the vram you can still move the pads using a screw driver when covered up, 1.5mm is the right thickness for vram.

 

There are a million different universal PSUs that can work with any notebook out there.
Shouldn`t be problem finding a 220W one

 

As you go higher its quite hard to near improbable to find one that is 150 W and above but if someone can find one let us know bit his issue doesn't seem to be a psu issue

 

When you bench the power draw is quite consistent, when you game you have fluctuations and spikes which work harder on the whole system & PSU.

 

You can't find any 180W+ PSU with 5.5mm x 2.5mm power tip - this is just too small to handle over 10A+ @ 19VDC so the only way to get a 200W+ PSU is to mod it yourself (as I had done for my 16F2 with the GTX580M in the past)

 

Or use a second PSU. Would also require some modding though.

 

so it might just be his system because HTWinghut also has 180 W PSU but can use the OC/OV? also if it is the PSU why would it shutdown?

 

Oh, the shutdown is simply an overload safety measure, build into the PSU.
Honestly it may be that the board BIOS doesn't work so well with the MSI vBIOS.
Why don't you try the non modded MSI 680 vBios to see if you run into the same problems?

 

"Overload protection means that if the PSU detects a large spike in current, aka a short, or it exceeds its rated current rating it will shut off. Over voltage protection means that if something bad happens and the voltage starts to go out of spec, say your 12V rail hits 12.5V, it will shut off the PSU automatically to protect the components. Any halfway decent PSU should have both of those, particularly overload protection. "

I got that from this site:
Overload/Over Voltage protection - Power-Supplies-PC-Cases-Case-Mods - CPU-Components

so that was what Prema meant by spikes in comsumption can damage the PSU and explains why playing games would shut off the PSU vs Benching/Stress Testing

EDIT:
here is good page to read too

Power Supply Overload Protection Function | FAQ | Singapore | Omron IA

 

@killer: got an headache

@Prema: It could be a good idea but there is two problem:
1) I don"t know how to find this bios .. maybe Pau1ow could give me an unaltered msi bios
2) The non-modded msi allow only a max of 853Mhz core so i could'nt realise test in the same condition ...

I found an other issue (minor for me)(bios or driver related) : when i closed the screen in order to quickly put the laptop in hibernate mode, it works ... but when i try to wake it up ... the screen remain black.
I need to do a hard reboot ...

I appreciate a lot your help and effort ...

My message for P150HM owner is:
Remember the card is compatible and work stable with standard vbios with 853MHz/2450 clock. At this frequencies it beat a desktop GTX580.

 

I have never had a shutdown even at a solid 1.075v.

 

i really wish someone could test whether or not you can have 2 external and the internal monitor all running at the same time(on the 680m), imagine how awesome it would be to have you laptop screen running at 1366x768, with a couple of these on the sides Hanns-G HL161ABB 16ms Widescreen 15.6" LCD Monitor (Black).
if your lucky maybe that monitor uses the same type of lcd as our laptop and you could pop in some 1080p lcds, that would just be amazing, imagine 5760x1080 on bf3 running compactly on your desk.

 

The HM and EM power circuitry are different - was it rock stable with demanding games such as Crysis or BF ? Because I doubt you can properly game at such voltage - at least in terms of heat generated.

 

I love to help you but i haven't got the screen

 

This should probably cover your question:

From an article about 680M:

Quadro K5000M (680M), tested by our forum member "Bokeh":

 

I think hes trying to make a point that he uses a 180W psu and that the power consumption of the GTX 680M @ 1.076v would be much larger than anything that tonrac is seeing, Also both P150EM and HM are built for 19V @ 9.47 Amps so overvoltage/overcurrent should have kicked in shutdown his computer. They would not suddenly decide to loosen the voltage or current tolerance for the EM series. But you do make a good question about gaming on that voltage.

 

"Also both P150EM and HM are built for 19V @ 9.47 Amps so overvoltage/overcurrent should have kicked in shutdown his computer. "
Not if the P150EM has a better build power circuitry that could be more tolerant on overload ... or maybe the ten month power circuitry's of my P150HM begin to be a little tired and don't support "this shock treatment".
But if i were a P150em owner, i keep an eye constantly on my wattmeter because i definitely think that the extreme OC is only available on a 220w laptop designed ... and probably, they will shorten the life of the power circuitry.
BTW i confirm that the card is "rock stable" at 853/2450Mhz with clevo vbios ...
after 1h30 of game no shutdown, i measured a peak of 87°C (without notebookcooler and without FN+1, i test fn+1 and in few seconds the temp drop down to 80°C); I also measured a peak power consumption of 175W (so the max on 900/2400 and OV couldn't be 177W) ... but average is more 165W.