twin snakes

On 7/2/2016 at 5:59 PM, ShatteredX said:

@J95

 

Just installed 368.25 and ran Firestrike. I got the highest score ever for my M17xR4  http://www.3dmark.com/3dm/12889451

 

 

This is with a +225 Core / +400 Memory OC.

 

I did have to use the Hibernate trick in order to prevent throttling. I'm also using 330w PSU as I suspect my 240w may be causing throttling too but I'm not sure.

OMG the Hibernate trick did work! Never know about this, you save my life! 

got over 19k graphic score with newer driver! never been so happy!

@J95 Do you have an inf file of the latest driver for 980m on win10?

/or can I use the one from 352.84?

/nvm, i figured it out

I'm on win10 with 352.86, working great, but how can I update newer game profile (like MGS V) into older driver?

yes i notice that too, 2nd gpu has higher vcore as well as lower asic quality

so do you get what you need?

there ya go, hopefully l'm doing it right

first i run about 10 sec of 3dmark11 (entry profile), take screenshot of 2 gpu-z status

then i ran render test in fullscreen mode, exit render mode and take asic quality of 2 gpus

final step is run render test in windowed mode and take asic quality at the same time (same with previous post)

pls note that left panel is GPU0 and right one is GPU1

/if there's anything i can help with, pls inform me

well of course i've tried the fullscreen mode, but the problem is render test itself, it cannot fully utilize the 2nd GPU, it still idle, pls have a look

GPUz of 2 gpus

@J95 yes there's no way to fully loaded two GPU at same time, with SLI disabled, the 2nd gpu still sit idle, i even think to swap 2 gpus, but seem to be useless since 2 gpus behave exactly same way, 2nd gpu still idle anyway

Here my result, stock cpu& gpu clock, stock vbios, stock bios, driver 353.06, with VDDC set to highest value

GPU0

GPU1

GPU 0

GPU 1

Well i still can;t run render test with secondary gpu, only primary one is fully loaded

here's mine, there's 2 problems:

- i cant "render test" both gpus (SLI), whenever windowed test or fullscreen test, only primary gpu is working, the other one is having rest time

- my machine misidentified primary GPU for GPU#1 and secondary for #0, not a big problem but i wonder wth happened, well since 680m-ism time long ago

VDDC is 1.05v during render test (main GPU), ASIC quality is 71.3% and 66.5%

Well i'm on holiday too, will help you on this sunday if you still need it

thank you

btw your 353.38 throttle like hell when benchmarking with 3dmark, but in games (GTA V, Witcher 3) still ok, no throtling

stock bios, stock vbios, stock GPU & CPU clock,

i installed exactly from that link you give, after clean windows installation

btw i will try 353.38 later, thanks for responding

/damn shame, i made a mistake, sorry pal

//have tried 353.38 PEG mod, still no luck, but i think there's sth wrong with my setup

///Well it seems that MFAA didn't support SLI, cos when i disable SLI, MFAA show up

why didn't i have the MFAA option in the driver setting? i using J95's 353.06

donated, great work

/damn paypal dont accept my credit card, i'll find another way, sorry

can 7970m goes undervolt lower than 0.95v?

I'm new to these forums and I was just wondering would Alienware M17 M18 VGA Upgrade Kit Dell HD7970M | eBay work in my M15x? I'm still using the GTX 260m and it needs to be gone.

yes that kit will work, but be sure to check the backplate's screw holes, sine the default backplate from m17x/m18x is way to high for the m15x's heatsink screws, therefore the gpu's die cannot make contact with the heatisnk

i guess it should work with fan control automatically, since my m15x works great with clevo 6970m

First of all... don't use furmark if you want to have an functional GPU.

Then... your temps are quite high for this card, you shouldn't get past 85°C, really. I hardly hit 80°C in long gaming sessions. You're sure the heatsink fits well? Enough pressure? Is your system flat on the table or do you lift it a bit below the battery? Cause that helps a lot...

thanks for quick reply

as i said, not only furmark cause high temp, but in heavy task games like sleeping dogs or MoH warfighter the system still shutdown midgame

i tried repaste using Coolaboratory's Liquid Ultra and Laird's thermal pad on Memory module and VRM... but still no use, idle around 45-50, load increase slowly to 85-86* then shutdown immediately!

I have even tried lift the battery side much higher.

any suggest? i'm really frustrated now

sorry for bumping this thread up, but here's my problem

i have installed 7970m on my m15x system following svl7's guide, including backplate grinding and heatpipe modding. So far everything went ok, 3dmark score ok, game playing ok.

But when i play some heavy game like sleeping dogs, medal of honor 2012 and stress test using furmark, the system automatic shutdown when gpu's temp goes beyond 87 degrees!

have tried undervolt to 0.95 but nothing change, still shutdown when temp reach 86-87 degrees!

please help me!