desktop 64 bit ( for modders ) http://www.geforce.com/drivers/results/81390
notebook 64 bit http://www.geforce.com/drivers/results/81392
Overclocking still disabled for mobile gpus (I assume forever)
http://us.download.nvidia.com/Windows/347.25/347.25-win8-win7-winvista-desktop-release-notes.pdf
-
Overclocking is not disabled. The release notes mention it is no longer supported through NVIDIA Control Panel.
reborn2003 likes this. -
Mr. Fox likes this.
-
its disabled. Tried it before.
-
Yeah, I just checked. This driver is garbage for Maxwell. It's a pile of trash just like 347.09 and the GPUs still throttle the memory to 799MHz like Intel HD Graphics shared memory snail poop. Everyone reading this, please go to GeForce Community and start complaining. Here is a link to my post at GeForce Community... LINK
I sure hope 980M does not turn out to be another engineering abortion that never gets fixed like what happened with 880M. A GPU that throttles and can't be overclocked is as worthless as the mammary glands on a steer.
NVIDIA GeForce GTX 980M video card benchmark result - Intel Core i7-3920XM Processor Extreme Edition,Alienware M18xR2
Yes, that is watts... stock with throttling.
I also posted here... LINK
On the bright side, the new version of GPU-Z finally works right for 980M.reborn2003, Kade Storm and TBoneSan like this. -
-
I posted on the Nvidia page from my phone. I couldn't be as detailed as I wanted but we need presence there.. So folks with MXM Nvidia modules.. Get cracking!
reborn2003 likes this. -
SOrry for the stupid question, but what exactly is going on? The newest nvidia drivers for 780M's in SLI garbage? Or does this post refer to something else. Sorry im on the tail end of a 16hr shift at work and eyeballs are popping out of my head.
-
jtravapd8578 likes this.
-
still trying to get more info for single GPU laptops cos it seems that even clevo is getting gimped when it comes to overclocking
-
-
reborn2003 and Ashtrix like this.
-
The way things have been going with drivers of late I wouldn't be surprised if the throttling problem's were entirely due to Nvidia.
Something just doesn't add up when they are flooding us with all these drivers that don't overclock, others that throttle, the old ones that are fine but still don't respond great to overclocking. There's very little information to go off other than things seem 'better' with Clevo on an unlocked vbios.
I'm still on 344.91 and maintain these are the best for 980m SLI for me right now. No throttling, they overclock, but overclocking gives inconsistent results mainly due to crappy Boost 2.0
Im temped to try an unlocked vbios and see how these drivers fair. I was hoping for a working UEFI NV flash firstreborn2003, Kade Storm, Mr. Fox and 1 other person like this. -
Brother TBoneSan... Save the effort, at least until a working UEFI version of NVFLASH is available to make testing easier. (Hopefully, one will be before too long.) I have spent probably more than 100 hours testing dozens of different vBIOS mods (ranging from extremely mild to crazy aggressive) and it is a futile endeavor. Probably 30 hours of that testing was with assistance of Johnksss, svl7 and Prema. Brother Prema even did a couple of custom system BIOS mods for me to alter how the BIOS interacts with the graphics system and it did not improve the situation.
The stock vBIOS produces the most consistent results, with the least amount of throttling. Not sure why, but overclocking with an unlocked vBIOS is hit or miss. Sometimes it works awesome, but most of the time the throttling is so bad that games are unplayable (frequent dips to 135Mhz on core under load).
I used my machine with the top end torn apart for a week at one point, using a USB keyboard, and thought I finally had the problem whipped after several days without throttling, reassembled it and the next day the throttling was back with a vengeance.
I agree that NVIDIA is probably more to blame than Alienware. It does not matter that 980M seems to work better on Clevo. Stop and think about it... All NVIDIA MXM GPUs in recent history function poorly with a stock vBIOS. The M18xR1 runs flawlessly with 580M, 675M, 680M and 780M, single or SLI. M18xR2 runs flawlessly with 580M, 675M, 680M, 780M, using Clevo or Dell GPUs, single or SLI. Nothing changed on Alienware systems. They did not need UEFI/GOP BIOS settings to work right with previous GPUs. Some Clevo can use Legacy/VGA BIOS with Maxwell, some not at all. Nothing changed on Clevo, at least on the models that work with Maxwell. Something DID change on NVIDIA's end. 880M was a piece of crap for almost everyone regardless of brand. 980M has behaviors that are similar in some ways to 880M, only more erratic when installed in an Alienware. The only extent to which I think we can blame Alienware is for abandoning MXM, being extremely late to the party with BGA Maxwell GPUs and not having the influence with NVIDIA they needed to have to ensure the Maxwell GPUs would work correctly in all Alienware systems.
Had NVIDIA not broken whatever they did, or had they not added something new that is all messed up, 970M and 980M would work on Alienware systems just as well as the earlier GPUs do. NVIDIA most likely added something to make Micro$haft happy with their push for Nazi grade controls through UEFI, which in turn allows Alienware's BIOS to invoke firmware signature control whereas Clevo's less sophisticated BIOS may simply ignore that filth. (I wish Alienware's BIOS was less sophisticated in this respect.)
One Clevo owner posted in a thread I created at GeForce Community about this problem and he says his Clevo with 980M behaves like my R2. Some Clevos are not compatible at all. Again, what changed? NVIDIA changed something and it wasn't a good change. Thankfully, it does work right on some Clevos and I am likely going to buy one for that reason only. But, I am not convinced that the lion's share of the blame goes to Alienware.reborn2003, Kade Storm, Ashtrix and 1 other person like this. -
-
I've thought about moving to a Clevo but always bail on the idea. If that new machine with a 4970k was SLI I'd have been temped. The 570 ticks all the boxes pretty much except vanity actually gets the better of me and I don't want to invest in one this late in the game when x99 is a near future possibility .. It's so damn ugly... But.. My loss there .
I will probably stay with the R2 and probably not get a new machine until DDR4 makes an appearance. By that time if everything is soldered I'll simply bail on buying an expensive laptop and build a compact SLI desktop.reborn2003, Mr. Fox and Ashtrix like this. -
I am asking this because I do not know anything about this development, I mean that I have no idea and I would like to know if there is any news on that.
I am going to test this drivers with SLI680. -
[Nikos] likes this.
-
Someone with the knowledge needs to mod those versions or we will have to wait for a newer version that includes support for our newer GPUs. Hopefully, someone will mod it that knows how. I believe j95 knows someone that can, but the person doesn't want to help or something along those line. Not sure of the details. Apparently, it can be done... j95 has been trying to figure it out on his own. I'd love to help, but don't have a clue where to start or how to do it.TBoneSan likes this. -
Mr. Fox likes this.
-
-
-
The disabled Maxwell overclocking could be NV trying to solve BSOD problems. Some machines have been plagued with such problems. Some of these notebooks ship with software installed that overclocks the GPU by default (ex ASUS GPUTweak) and for all we know that is one causal factor.
-
After installing 347.25 i have noticed that in SLI my second gtx 780m has about +3-5 C in idle and +12-15 C during load comparing to first gpu. What about you guys ?
347.25 WHQL for Dying Light
Discussion in 'Alienware 18 and M18x' started by Zero989, Jan 22, 2015.