Alienware M18xR2 New/Updated BIOS Mods - NOW WITH MAXWELL LEGACY SUPPORT - Testers Needed

I can also confirm that I have no throttle, windows 10 latest driver, but you do need to toggle in your bios from igfx to peg save and restart, then switch back to peg again, will run throttle free till you shutdown and unplug your system. BTW This has only been tested and work's/fix's the M18XR2, not tested on the others.

 

Yes I use HWINFO64. I wasn't aware of this!

 

I talked with the guy over at HWINFO I had this problem as well till he told me about that, after I disabled the battery monitoring my lock ups vanished, I donated moneys to him and thanked him.

 

I tried that driver as well the clock speeds drop and get stuck at low states when a throttle is triggered and no hwinfo installed. Back using the latest driver for now.

Only thing I can think of is a bios config file is not loading correctly on cold boot and this triggers throttling but flipping the gfx switch resets it.

Pity swick hasnt been around lately but it would be nice if someone could check the bios power configs between a cold boot and reset and see if theres a difference.

 

I am using HWINFO from day one I had the laptop. From my CF 7970m I had the fans at max speed to keep them cooled, but they died anyway!

 

I had a couple of freezes today, one was because I forgot to disable the battery monitor on mine, the other... not sure why? I disabled drive scan to (I think I had it disabled last time I used it).

If you still have lockups on HWINFO (I know this is what caused my freezing) this is the config I use that seemed to stop the freezing. https://app.box.com/s/z9cfku4c9xr8nbvacfirof3ko9mpse5g I uploaded it to my box account, it is setup to auto launch with windows and automatically bring up the fan profile as well. Replace the file in C:\Program Files\HWiNFO64 with the one I put in that link, lemme know how it works for you .

 

I will try it thanks. Though I never had lockups to the system, only 'hard' throttling with the 353 modded drivers.

Now with this procedure we are throttle-free

 

I know it says for R2,but can I use this MOD for my R1?

 

No way unless you'd like to brick your computer. The flash sizes aren't even the same....

 

No, we are still waiting if anything can be done to make Maxwell usable in M18xR1

 

Hi guys @Swick1981 has asked me to let you know that he will be back on this board in approximately a weeks time. As Mr fox said hes been on Vacation and is looking forward to getting back and helping out the NBR community. Have your questions ready as i think there could be no better welcome back than to bombard him with requests for help.

That will teach him to go away lol.

 

Just buy him a welcome beer lol

 

Guys I dont know if anyone else has noticed this... When I get throttling with my 980m, I simply restart the laptop, go into bios and change from PEG to IGFX. After booting into windows I restart again going into bios and set back to PEG. Then the throttling is gone!

 

Few of us actually posted this already lol . Only problem for me is it seems the os goes tits up after a few days on me with corruption issues of flipping gfx forcing me to restore from a clean image.

 

I already have done this a lot of times but never had any problem with windows corruption. Everything works smooth so far!
But I have another problem, I always have to use hwinfo to set the gpu fan to the max to get good temps on the GPU while gaming. (around 75c max). If I dont use the hwinfo then it will reach 92c after a few minutes and throttles due to high temp. Anyone else has this problem?

The fan works and raises RPM as it gets hotter but never reach maximum speed!

 

Which laptop and vbios are you using? Single card or sli? And are you hitting 92c with HW info?

 

Has anyone tried this BIOS with Win 10 x64? And did Nvidia fix the problem with 18" monitor "burning out" (corrupting the LCD firmware)?

Seriously considering taking the plunge with the 980m SLI. (And Finally got my 1TB Sammy 850 pro in so I need to redo the OS anyway to get off the SATA 2 BS Raid-0 that dell "tricked" me into running)

 

EVGA Precision/Nvidia/Windows 10 LCD Bricking Bug not fixed as far as I know - best not to run EVGA Precision to be safe, and even then there's been a few (not many) failures with people who have never run EVGA Precision.

 

M18x R2 with single 980m. I am hitting 92c without hwinfo. With hwinfo if I set the fan to max I never get more than 75c. The fan works without hwinfo but it never reach the maximum speed as I can see.

 

The bricking issue was with evga precision corrupting the EDID. Im using legacy win10 64 and apart from the occasional derps outta it it seems to work mostly.

Only problems I seem to have with mine is the OS will sometimes corrupt for whatever reason forcing me to use an image to restore it. Ram and SSD seem fine from testing. It seems hit and miss with windows 10 depending on who you are.

 

Some people have had issues with the secondary card fan not working properly... first time I have heard of it in a single card setup.
No need to set fans to max, you can make a fan profile in HWiNFO so that it ramps up properly until you find a fix at least.

 

Thank you for the reply, I'm in the market for a pair of 980m's talking to Woodzstack currently, Awaiting his reply. How do you tell if the cards are made by EVGA Precision anyway?

-edit- The above comment was brought to you by an nVidia noob - I was corrected, EVGA Precision is not a card but a OC program

 

Might be the vbios too. The AW18 980m one would cause the 2nd fan on mine to not spin up but the thing works fine with the clevo version instead

 

I had the EXACT same issue with windows 10 corrupting. I kept having to restore it every 3 or 4 days. Turns out I had a bad Image. I downloaded via torrent (using my product key). After downloading another image. I have had absolutely no corruption issues. Hope this is the same for you buddy

 

Evga precision is a OC program. Its what was the primary cause of the corruption to the LCD EDID. Ive used windows 10 for months but my monitors fine.

 

Okay! I understand now, Thought you had a windows 10 issue. I was like "i know thisssss!". haha. So does MSI afterburner not work on these cards? Is that why people are using eVGA Precision?

Also do you think these were the issues Mr. Fox had? he had the burn out happen on 3 different laptops. I seems weird that he would install the cards in another laptop then immediately over clock them while at the same time trying to find out why his monitors where blacking out. Are we sure the EVGA Precision is the ONLY problem?

Asking because I do NOT want to "burn out" my panel here in the next couple weeks when my cards come in

 

Mr Fox had the problem appear both with & without EVGA Precision, he was one of the rare ones where there was a problem even without EVGA Precision (if I recall correctly) - it's a bug primarily limited to EVGA Precision/NVidia/Windows 10/Alienware (that combination mostly, but has happened to other notebook manufacturers and operating systems in vastly smaller numbers). Best to use NVidia Inspector as your OC program, it offers & exposes a few OC variables that aren't there in some other programs (you can control clocks/voltages/TDP limits/throttling temperatures).

 

I downloaded mine with the tool off the microsoft site. Unless thats corrupted. I also made a new usb bootable recently for windows 10 for the legacy install and thats also showing corruption with a /sfc scannow run.

 

SC doesn't lie. I would double check your hard drives Integrity if you're absolutely sure about the image hash check. You either have a bad image. Bad hard drive. Dieing/OC ram or a mean virus. I hope / experiences

 

I just heard from @Swick1981 and he is finally back home in the USA. We should expect to see him in the forum pretty soon. It may be a day or two before he can get online.

 

Ran all week doing the switching from peg to igfx and back to peg at cold boot trick, no throttles each time so thats gotta be are bug

 

Just a guess on my part, but I suspect it is something in the EC monitors power draw on discrete graphics and does not with Intel HD Graphics and switching back and forth temporarily confuses the system enough that it works correctly when power draw is ignored. I think we would need to locate and destroy whatever piece of firmware in the BIOS or EC needs to mind its own business and ignore GPU power draw in order for everything to work correctly 100% of the time. It may also have something to do with the vBIOS, Intel ME and BIOS/EC collaborating in an unfavorable manner. With the industry devoting an unhealthy amount focus on accommodating the severe power handling restrictions and thermal limitations of pathetic turdbooks, it is a wonder anything works correctly. When everyone is limited by the shortcomings of the lowest common denominator nothing turns out right unless you're one of the lowest common denominators.

 

I took the dive, I'm running 1 980m I bought from @woodzstack and I'm on Windows 10 64 bit, I'm also using swicks bios. Switching the bios from peg ---> IGFX save and exit bios and then enter bios again and go from IGFX ---> peg save and exit will cure the throttles till a permafix is found Also if you need help with it don't be afraid to drop me a line or the forum I'm sure everyone will help out I've torn mine apart way to many times because I was learning, a tip if you do get it, don't put the thermal paste on the gpu till you have your heat pads in place, a good way to check if they are making contact with the gpu components is to take a piece of paper and tear it up in a few sections, place them over the top of the heat pads and put the heatsink on, hold the heatsink on and tug on each piece of paper, if it's tight then the heatpads are touching the components, if it's loose you need to make that area thicker , <--- lots of trial and error but I figured it out.

 

Not sure if it would be vbios though I think some are using prema's vbios, I've been using slv7's vbios (helps keep my temps in better check) and the throttle seems to occur on both versions until we do the bios toggle, I'm thinking Intel ME and BIOS/EC area's need to be checked like your stating.

 

Which vbios are you running? I run slv7's vbios it keeps the card at 1v and tends to help with temps . 75c isn't bad I think under load the hottest I've seen mine hit is 73ish? but I'm using hwinfo as well, default fan controls can't do it.

 

Ive been using premas v2 bios on my sli 980m's since the 361.xx series began as apparently stock vbios and those drivers would cause a blackscreen while booting. Also have to toggle the gfx as well to get rid of throttling (when win10 is playing nice). Did that issue get fixed btw?

Also I hope we can find the issue thats causing the throttlng as well. I think its safe to say were close to solving it the fact that we can actually run our machines throttle free with a simple gfx toggle at least would swick or prema or anyone with knowledge some idea on what to home in on. If we can nail the bad code then we'll finally be able to put this behind us.

Could be something like the bios thinking the 690m is there instead of the 980m and basing power draw on the former instead of the latter. Find the part of the EC that regulates the a18's power draw and patch that to the m18xr2 and were free.

 

I'm using Windows 10 64 bit with mine but I'm running on Legacy mode, I have yet to go back and try UEFI again, but I think the black screen thing is fixed, but I do remember them saying you need prema's vbios for it to work, I've been using slv7's and its been working fine haven't seen a need to go back to the stock firmware yet.

 

I'd like to know how or why a stock vbios would cause an issue and which stock vbios too. I ask this because we have roughly 100 vbioses available. none are REALLY stock.. as they have ALL been modified to some degree, for better handling.

 

Looks like you were right I got another iso off the microsoft tech site and managed to repair my installation's repair image with DISM and finally fixed the errors with sfc /scannow. My Win10 install should be fixed for now. Might've spent a few hundred Euro changing the RAM for nothing XD. Of course the new ripjaw ram might be better than the old 4 year old stock ram in my machine as well.

Edit: Corruption came back and had to download the tech iso again to repair it. Apparently theres an issue with one of the files in windows 10 going wierd.... -.-. Also looks like imma have to do this everytime I update the drivers as well for whatever reason. Least for now a sfc check has turned up no corruption or integrity violations.

http://www.tenforums.com/performance-maintenance/31695-sfc-corrupt-issue-regarding-opencl-dll-3.html

 

I haven't tested the stock vbios with latest drivers since we found the problem, when i get a minute I'll test it, I know j95 said something about some of the laptops getting a black screen problem with stock vbios and the fix was going to prema's vbios.

 

Did you see my post on HDTune pro and the memtest? http://www.hdtune.com/ and http://www.memtest86.com/ hdtune pro will tell you if there is any kinda of problems between your sata drives, controllers, and cables. Memtest86 will spew out errors in red if you have a bad stick of ram, you can pull one and run it again till it stops spewing errors to find your bad stick, or just test 1 stick at a time with it.

 

No it wasnt the ssd or memory (I actually replaced the stock memory yesterday with ripjaws). Its apparently an issue with windows 10 itself that others had as well. The corruption was a single file related to the graphics. Ive fixed it now after digging some more and downloading a technet copy of windows 10 and repairing dism with that. It might be an issue with the current release of windows 10 as well.

Edit: the alienfx went wonky and wouldnt work even after a reinstall so used the image again and updated the new drivers + fixed the dism again. Noticing some programs dont seem to shut down correctly on shutdown (program.exe for example thats part of my g13 and sometimes .net programs which may be related to command centre). Gonna try keeping an eye on it and when they release redstone 1 in a month or 2 try and see if that fixes it. Very odd because the ram isnt the issue since its brand new and the msata ssd was checked and is 100% healthy.

 

Excellent tips right there. Thank you for the advice!

And I am chatting with woodzstack as we speak. Hashing out the details

And welcome home Mr. FOX!

 

Has anyone tried using MSI afterburner on those 980m to force the clocks to stock keep them from Auto overclocking would that keep it from throttling? I mean keeping the clocks down to stock... shouldn't that lessen the TDP keeping them from hitting the "power limiter" of the Mobo (or what ever is causing it)?

I realize this the answer is probably "no" as it seems to be too easy of a workaround, but I had to ask anyway

 

amazing work man, thank you for the solid mod!

 

All the "stock" vbioses I have heard of see or used or tested never had any such issues. This is why I am curious. Editing a vbios is not a difficult thing to do, there are programs out there that are almost so easy, its just a few clicks, and enter a value, and you can change settings like voltage and clock speeds. The stock vbioses are usually the same as the overclocked and overvolted ones, except usually more stable, not less. UNless someone can provide irrefutable evidence that suggests otherwise, I have to think a stock vbios is always best, for compatibility. Overvolting a vbios, while sure almost anyone can do this, might provide some stability at the cost of the cards life, in the event it is not installed properly, and cooled efficiently or provided power properly. But its all just cheaps tricks. Its snake oil. You can use MSI after burner to overclock, or nVidia inspector, theres others. Just saying, but in terms of a bios not being good, I would have to know more before I believe that, is it a first iteration a very old card, like for example if it is a 980M from early september 2014, we had a different set of vbioses for a few weeks, small tweaks to how UEFI dealt with the cards. But no body would know this because it didn't sell or leave with that stock vbios, and was updated before the run was made. This is because they're all tested. I'm not going to go into infinite details, but again, flashing vbioses is like buying snake oil. Now a system bios, those are deliberately screwed around with, to prevent you from doing things, but I've never seen this done to a videocard.

 

Ca

Can you tell me where are downloads for R1? Thanks

 

Dell stock bios is compative with all drivers. No black screen or any other. The difference between premas and dell bios is performance.

Stock bios fluctuates clock and voltage beetween 1037 to 1127 and 1,000v to 1005v depending on load.

It isn't perfect, but for gaming it keeps the GPU at lower tems. Mine even without having to max out the fans, I get 65 degrees. If the GPU reaches 70 the fans kick in according to the stock fan table.

Dell's bios, notepal U3 and maxed fans never get my GPU above 60 degress.

Premas bios was better for benching but while gaming without any help to colling, my GPU easily was reaching 75 degrees!

 

@stefan063

Still nothing for M18xR1

 

This might be a minor detail but whenever i switch gfx the screen takes about 5 to 6 seconds to come on unlike the cold boot which is only 1 or 2. This could be a reason for the throttling as the system could be loading incorrect setting which doesnt happen on a gfx switch.