M15x & 6970M / 6990M - What you need to know - upgrading, problems, solutions

I've noticed heavy glitches in flash and other hardware based desktop apps when using memory setting other than 150 and stock; even with the 5850m. I think they just haven't been able to set the memory up so that it can switch fluidly.

 

My laptop is stuck at idle clocks 250/900 and I can't game..anyone know how to fix this? Stealth mode is OFF.

 

Use ATT to manually control the clocks. For external displays 250/900 is the default.

 

Well I know I can use that but normally my laptop runs fine; right now the GPU load reads 0%
ATT still doesn't work, GPU load still showing 0%.. sigh

 

That's weird. Did you change anything in your system prior to it happening? The only other idea I have is toggling powerplay on and off, it might be stuck off for some reason. You need to powerplay to use 3D clocks.

 

The biggest thing I can think of is repasting my computer this morning, and after that the fans died on me for a bit, leading to very hot temperatures. After unplugging and replugging in the fan, now it's just incredibly loud. My GPU worked for a bit, but then when I went to play again it's in its now-current idle state.

Like I said earlier, I ordered the copper shim mod so will be trying that when I receive it hopefully Thursday/Friday, but until then I guess I'm stuck running my computer like it's in stealth mode What causes this?

Running GPU-Z shows idle clocks 250/900, using ATT or TriXX can change the numbers to 680/900, but GPU load still 0-10%.

 

I think the shim will be the solution. From the looks of things the heatsink is tilted a bit without one, so adding it helps with contact and also raises the low end.

 

Thought this was interesting, when my fans are blowing the hell out, my temps are fine and my GPU load is crap. When my fan is completely silent (like this morning, probably will change if I restart my computer), GPU works beautifully and my temps hit 90C..

 

Right click on the desktop, screen resolution, display "1. Mobile PC Display" and choose that the desktop gets only displayed on 1.

 

250/900 is idle for using an external monitor right? My main concern is just that my clocks don't change when I try to game.

 

Yeah. If you're using an external display that's ok, else some setting is messed up. Check powerplay and stealth.

 

Those are both on, I've checked and double checked. Just gonna wait until I receive my shim and hopefully that'll fix some problems if not all.

 

waiting for a good news

 

Mother ****er! I took the computer back to my office today and plugged it into the external monitor like i always do, booted up Alan Wake and still got the damn shutdown at the same part as before.

I've made my clocks stay at 680,900 and 1.00v which prevented my laptop from shutting down when disconnected however it hasn't fixed the issue while i am connected. No matter what i try this shuts down at the same part every single time when i'm connected to an external monitor.

 

i just notice that you have different vBios version than mine which i use the 2nd post of this thread. Where did you get yours ?

 

I have the 6990m

 

Just an update; my thermal pad which I ordered had just arrived (not fujipoly due to the shipping cost is more expensive than the product) from Ebay. Took me 2 hours to set everything in place, looking good.

Ordered 0.5mm 100x100 and 1mm 100x100 each

What I did was:

-Padded the ram and remove the metal cover.
-Repadded/Repasted CPU/GPU (a combination of Mexic00ls's picture and JohnnyFlash's)
-Cleaned CPU/GPU vents.
-Recheck heatsink mod.
-Modded vBios to match JohnnyFlash's 710/1v. (will be experimenting on the numbers)

For now the temp stays at 70C-ish while gaming, and around 60C when watch streaming videos.

EDITTED: out of topic question, do you think this will work in m15x?
http://www.ebay.com.au/itm/INTEL-i7...aptopMotherboards_CPUs_CA&hash=item256a123043
was just making sure

 

Hi guys!

@Hel20eS: This CPU will work. Mine comes too from laptopmonkey.
I 've padded the ram and removed the metal cover, this didn't work: shutdown occured.
Now I've installed the copper shim mod, i hope this will solve these issues.
I've too noticed that my card throttles under heavy GPU load between [email protected](vbios' frequency and voltage) and [email protected]. Could that throttling be the shutdown's culprit? (svl7's runs @1.1v, i believe)

 

I think the throttling is a driver issue; I get it too once and a while. Waiting on the full 12.2 release to test more.

 

Shutdown is back... oh well at least I can eliminate another to-do list that helps shutdown issues.

Padding ram and GPU: BUSTED

 

hopefully the shim will be the fix....waiting for someone to verify

 

I have the 6970 no shutdowns. The only problem is I bought an i7 940 now I'm getting shutdowns again. Gonna try adding a copper shim on the processor and see if that helps. Someone posted about under volting the processor i wish I could find the post.

 

Alright, got the shim installed (I used the single thick piece instead of the two thin ones, I'm assuming that's the correct one based on Johnny's earlier post) and gonna do some tests. My fans are still not always turning on though which makes me have to restart my laptop, anyone know why?

Still crashed in BF3...attached is GPU-Z. Anything look strange?

Gonna try JohnnyFlash's voltage mod next. Edit:710/900 1v caused crashing for me. rawr..

 

this is one ty problem. After all this time nobody can figure out why this is happening let alone how to fix it. Luckily it doesn't happen with every game. The only game i've played that it does it on is Alan Wake and Skyrim. I've played tons of other games with no issues though so that's good i guess.

 

It only occurs for me on BF3, but that's like, the one game I don't want it to crash on..I only play HoN and LoL other than BF3 so can't contribute too much. I did play MW3 during free weekend and that ran fine.

 

It seems the cpu saps power from the gpu? This is one odd problem!

Lowering the tdp/tdc of the 920/940xm cpu's might help although considering I have read crashes occur even with the 720qm which must use quite a bit less power I am not sure it will help!

 

I don't know if my processor (i7 620m) affects my gaming results, I feel like I'm the only one with this processor though so...

 

hmm maybe the i7 620m is a bottlenecking the 6990M in BF3 it is a cpu intensive game afterall.

Wish I had the game to test myself..

 

Might be stupid of me since I just thought of this..but should I be using 1mm or 0.5mm thermal pads on the rectangular portions of the GPU heatsink? hehe...x.x I've been using 1mm and not sure if that is the right thickness.

Sent from my SCH-I405 using Tapatalk

 

Hmm I think 0.5 is adequate but I have used in some areas 2 or even 3 layers lol so it shouldn't matter too much.

I did furmark today at 715/900 at 0.96V which are stock clocks for 6990M but undervolt. I did extreme burn in at 1980x1080 and it maxed eventually at between 82C and 85C on the temp sensors. I guess this is the norm as even the most gpu intensive games load at about 5-10C lower. This was with the fans forced to max with HWinfo. Interestingly when I fold at same clocks and fan speed the GPU maxes out at 65C even after hours...20C ambient temps

 

@King of Interns: The question is at which voltage your GPU actually run... 0.96V isn't in the voltage table of the VBIOS.

 

Through exhaustive testing with my old 6970M I disagree. I tested in Crysis 2 (1920x1080 with dx11 candy on etc) at increased increments in clock speed at certain voltages and found that the card triggers its incremental voltage increases every 0.05V. This happens at 0.96V, 1.01V, 1.06V and (1.1V which is the odd one out)

This is definitely true as I found instability for example at certain clockspeeds that I tested also in 5mhz increments until unstable. 730/900 was stable at 0.96V but as soon as I went up to 735 the game crashed. This was true all the way up until I hit 1.01V which then allowed me to go up to 780mhz beyond that required 1.06V to go to 830mhz and 1.1V to go beyond that. The slightest voltage increment under those critical voltages made no difference to stability.

So 1V at 735/900 was no more stable than 0.96V but a bump to 1.01V brought stability. Maybe in the vbios the voltages are actually 0.9V,0.95V,1V,1.05V and 1.1V but the software in gpuz, att and so on report it wrongly... but these are my findings and I have to say they are pretty concrete as I took days to test it out.

 

Yeah, that's true, 0.05V increments are possible. There are seven entries in the VBIOS, two of them are below 0.9V

 

See my post 4 pages back.

 

Hi,

Somebody can use an external monitor using display port with the 6970m? i´m trying it but i only can connect the monitor with VGA.

i have Dell vbios and dell´s M17x r3 6970m driver installed on M15x

 

Oh for those with the shim mod, how much paste do you have between the heatsink and the shim?

Sent from my SCH-I405 using Tapatalk

 

The same as you would for a die that size.

Also, be sure to evenly tighten the screws starting with the bottom inner corner first, then the opposite corner and so on.

 

I'll try that when I get 0.5mm pads. I have a feeling that my 1mm is too thick, which may be the reason my idle temps are higher than normal, and possibly crashing issues. Who knows..

 

If you use the thick shim, they won't be. I'm using 1mm only on my card.

 

When I'm screwing the heatsink back in, I feel like there's resistance due to the padded areas being slightly higher, which lead me to that thinking. Do you encounter that as well?

 

A little, that's alright though, the pads will compress and make better contact. Good idea to test though, just in case. I did all my testing before putting the vents and bottom panel back on. Leaving them off makes life way easier if you want to adjust.

 

Ok so all of my steam games seem to have a crazy lag. Where I used to get 30fps I get 5fps, and the only way to increase the fps is to turn down the resolution. Anyone know why I am getting lag due to 1920x1080p? It used to play 60fps but now bfb2 only plays at 10fps unless I lower res. then it goes to about 30fps. This is extremely annoying. Also the weirdest thing is all of my non steam related games (minecraft cryengine 3, world of warcraft, EVE, so on) play fine. Please help this is getting on my nerves.

I also first noticed this after I updated AVAST.

BTW these are my specs radeon 6990m i7 740qm 4gb ram 500gb HD

 

Log with GPU-z to see how it's downclocking.

 

Date,Time,"Virtual Memory Commited","Virtual Memory Available","Virtual Memory Load","Physical Memory Used","Physical Memory Available","Physical Memory Load","Core #0 Clock","Core #1 Clock","Core #2 Clock","Core #3 Clock","Core #0 Thread #0 Usage","Core #0 Thread #1 Usage","Core #1 Thread #0 Usage","Core #1 Thread #1 Usage","Core #2 Thread #0 Usage","Core #2 Thread #1 Usage","Core #3 Thread #0 Usage","Core #3 Thread #1 Usage","Total CPU Usage","On-Demand Clock Modulation","CPU#0 Core0","CPU#0 Core1","CPU#0 Core2","CPU#0 Core3","Temp 0","Temp 1","Temp 2","Temp 3","Temp 4","Temp 5","CPU","GPU1","PCH Temperature","CPU Core","SAMSUNG HM500JJ [S2APJDRB311257]","GPU Thermal Diode","GPU TS0 (DispIO)","GPU TS1 (MemIO)","GPU TS2 (Shader)","GPU Fan","GPU Clock","GPU Memory Clock","GPU Utilization","GPU Fan Speed","Battery Voltage","Current Capacity","Current Capacity",
26.02.2012,12:40:00,3700 MB,4467 MB,45.0 %,2084 MB,2000 MB,51.0 %,1956.5 MHz,2035.2 MHz,2049.1 MHz,2105.7 MHz,21.8 %,15.3 %,14.2 %,0.5 %,0.0 %,10.6 %,19.5 %,0.0 %,10.2 %,100.0 %,47.0 °C,47.0 °C,48.0 °C,47.0 °C,59.0 °C,50.0 °C,51.0 °C,53.0 °C,44.0 °C,54.0 °C,4007 RPM,4448 RPM,62.0 °C,25.464 W,42.0 °C,60.0 °C,61.0 °C,61.0 °C,62.5 °C,0 RPM,100.0 MHz,150.0 MHz,9.4 %,30.0 %,12.410 V,35.676 Wh,100.0 %,
Date,Time,"Virtual Memory Commited","Virtual Memory Available","Virtual Memory Load","Physical Memory Used","Physical Memory Available","Physical Memory Load","Core #0 Clock","Core #1 Clock","Core #2 Clock","Core #3 Clock","Core #0 Thread #0 Usage","Core #0 Thread #1 Usage","Core #1 Thread #0 Usage","Core #1 Thread #1 Usage","Core #2 Thread #0 Usage","Core #2 Thread #1 Usage","Core #3 Thread #0 Usage","Core #3 Thread #1 Usage","Total CPU Usage","On-Demand Clock Modulation","CPU#0 Core0","CPU#0 Core1","CPU#0 Core2","CPU#0 Core3","Temp 0","Temp 1","Temp 2","Temp 3","Temp 4","Temp 5","CPU","GPU1","PCH Temperature","CPU Core","SAMSUNG HM500JJ [S2APJDRB311257]","GPU Thermal Diode","GPU TS0 (DispIO)","GPU TS1 (MemIO)","GPU TS2 (Shader)","GPU Fan","GPU Clock","GPU Memory Clock","GPU Utilization","GPU Fan Speed","Battery Voltage","Current Capacity","Current Capacity",

Thats what I get while playing bfb2

 

Please.... how long have you been around here? Attach the log file to your post, this is totally useless.

 

I am sorry I don't really understand what you mean. This is the log file. it is short witch is kinda weird cause I logged for 15 mins. Sorry, but I just don't get what you mean

 

You need to go to advanced reply and attach your log text file as an attachment

Sent from my SCH-I405 using Tapatalk

 

And you need to check "keep refreshing this screen while running in the background" in GPU-Z, sensor tab, bottom.

 

@Johnny, currently I'm using Thermal Conductive Silicone Pad TP-250 (5.0W/mK); 1.0mmx4"x4" (076783016996) | eBay so I don't know how compressible it is considered compared to say, Fujipoly. I'm gonna buy the 0.5mm version of these and see how that turns out most likely.

 

That stuff is 5.0 (W/m.K), Fujipoly is 6.0, so 20% better thermal transfer....wether that'd actually make a huge difference or not though, I'm not sure.