M17x-R4 Black screen crashes ATI 7970m

I just ran the memtestCL again but with 2048 50 .

Test summary:
-----------------------------------------
50 iterations over 2048 MiB of memory on device Pitcairn
Moving inversions (ones and zeros): 0 failed iterations
(0 total incorrect bits)
Memtest86 walking 8-bit: 0 failed iterations
(0 total incorrect bits)
True walking zeros (8-bit): 0 failed iterations
(0 total incorrect bits)
True walking ones (8-bit): 0 failed iterations
(0 total incorrect bits)
Moving inversions (random): 0 failed iterations
(0 total incorrect bits)
True walking zeros (32-bit): 0 failed iterations
(0 total incorrect bits)
True walking ones (32-bit): 0 failed iterations
(0 total incorrect bits)
Random blocks: 49 failed iterations
(348560361 total incorrect bits)
Memtest86 Modulo-20: 0 failed iterations
(0 total incorrect bits)
Integer logic: 0 failed iterations
(0 total incorrect bits)
Integer logic (4 loops): 0 failed iterations
(0 total incorrect bits)
Integer logic (local memory): 0 failed iterations
(0 total incorrect bits)
Integer logic (4 loops, local memory): 0 failed iterations
(0 total incorrect bits)
Final error count: 49 test iterations with at least one error; 348560361 errors
total

 

not in stress test (I just don't do stress tests, they are just a way to kill gpu quicker), that was just my max temp ever with 7970m (and as soon as it hits 75, it cools down to 60-65C for all the games and it remains there), I hope your gaming/benching temps are also like that and only in stress test you are getting +80C?

 

Yes, as i said i was running 80+ only in extreme burn mode.

 

Just send the notebook back to Alienware and wait for GTX 680M instead.

AMD and their drivers is nothing but pain in the butt anyways

 

My fans on my M18x are acting much the same like MR Valle's. Once the gpu is around 75c they kick into medium and stay there till around 71-70c and drop to almost idle fans again.

Anytime I used the 1.075 not stable long term for me.

Maybe it's vram temps. Our gpus don't seem to track vram temps just I think the core.

I'll stick to 12.5 and test more. I play diablo most night so I do get like you a good few hours to test straight. I'll report back when I ran stock all on 12.5 for awhile. We shouldn't need an overvolt for stock clocks. I understand why it could help but shouldn't be required. Basically I know since these are 7800 series stuffed into a mobile die some cores may need more volts much like desktop gpus. Binning of these mobile cards could be better?

 

Anyone got news on how to stop the black screens?

Sent From My Rooted EVO 3D

 

Still experimenting, but not crashing at this point if I don't have any overclock. Some games will take a heavy overclock, some won't even take 25Mhz over. Will report back again after some more testing and what I've done.

 

Looks like AMD may have just jumped the gun and not had their software up to snuff before trying to beat nVidia to the punch with their 680M.

What I would do guys is keep stock clocks and voltages and wait for driver improvements. You're probably just confusing the issue with over-volting and over-clocking. From what I can tell from your conversations, even given the previous "recall", this seems like a driver issue.

I have been mentally killing myself deciding between the Samsung 7 Series Gamer with the 675M and the M17x-R4 with 7970. I know the AMD kills, but the Sammy is a better overall value. Trying to stick under 2000. Edit: my heart is with the M17x. I can add another 8GB RAM and an SSD later. Sammy only has an 8GB express cache module anyway. Here's to hoping AMD gets this sorted out quickly.


Sent from my Galaxy Nexus using Tapatalk 2

 

I agree with this above. I've been stock ever since first few issues. I tried 8.5,8.6 and 9.0 catalyst. On single card I was fine but after getting xfire cable recent everythings been unstable. I was running single card for awhile and had no issues once I cooled the cards with full fans and on stock values. Now I had one nightmare getting any driver to load on xfire. I ended up must of been 12+ uninstalls/trace file cleanups and reinstalls. Avoid 9.0 like the plague maybe it was just my M18x R1 but it did horrible with freezing and issues sometimes not even in game. I'd hot shutdown to end up at a black screen. I found two scenarios And this goes along with what you said above there are mixing probs here. One black screen if sound is in backround helps to know but one is a freeze and likely from any kind of overclock on any driver version. The screen turns black and the sound turns to like a screech and must powerdown system since even HD activity is frozen.
The second one I started to run into more so on xfire but happened on some drivers too on single card. Basically a black screen and the system is not frozen this one is likely the driver error. As I see HD activity and sound from game is still running normal just a black screen though. Now sometimes on that one I can hit CTRL ALT DEL and go to task manager succesfully and close game to end up on desktop like normal no freeze. As mentioned I ran into this one a lot more on xfire but sometimes under certain drivers or games it happened on single card. Witcher 2 no matter what driver (I'm back to the most stable 12.5) while in xfire will in just few minute temps being very cool anyway will do the non freeze black screen. So I have yet to get CCC or Radeon Pro to manually get it to only run single card while leaving xfire available to the rest. I set profile but still launches in xfire. Diablo 3 running fine in xfire on 8.5.

Catalyst beta 9.0 crashes diablo and witcher 2 no matter what eventually. Witcher 2 xfire was no different than 8.5 probs but diablo 3 had fps drops and then goes back up and eventually stutters real bad and leads to freeze. Atleast on 8.5 diablo runs fine in xfire. I don't get 100% gpu usage on both but runs maybe 30-50fps higher than single card.

So as advised above the freeze black screen means back off your overclock even if it's barely a overclock... As drivers can't handle the clocking yet. And if you get the black screen that doesn't freeze try earlier driver 8.5 and if in xfire try making just that game run in single card mode.

Btw anyone can u tell me why I can get Radeon or CCC to load xfire in single card for specific individual games. I feel like I'm not setting something right. But there have been profiles set up.

 

I am not doing any overclock at 850/1200 which is stock clocks I get black screen crashes on every game and I am forced to reboot, at 825/1200 I can play some games and at 800/1200 I am stable. So this is definetely not normal, that even with some crappy drivers I need to underclock my new card to be able to play...

 

Definitely fishy. I'd get a replacement, if that doesn't work sounds annoying to have to wait for good drivers for a released product. Happens though.

Sent from my Galaxy Nexus using Tapatalk 2

 

I would return the laptop or demand a replacement. This is total BS. No one should need to underclock or undervolt a new laptop. Period.

 

This. And I also just got my first black screen on stock clocks. Something is definitely wrong with these cards, the drivers, or both. Makes no sense that we can Furmark as long as we want with a heavy overclock and have no crashes, but as soon as we try to play games it starts black screening. It doesn't seem to be a stress on the cards or the power supply or the such, unless it's such a specific circumstance of a spike that causes it.

I'm about to send this thing back and wait for the 680. ATI's drivers just aren't up to par with NVidia, unfortunately. I've just had bad history with NVidia in the past, too, with cards running too hot or drawing too much power. Maybe we need a 3rd card manufacturer to pop up and shake things up a bit.

Last ditch effort for me - undervolting. I'll let you guys know in a few days.

 

All I can suggest is I'm on stock clocks and no black screen issue now that I am back to 8.5beta. I now am in crossfire wasn't for a long time. And still all is fine except witcher 2. If in xfire I get black screen if in single card game runs fine.

 

Might be voltage. Been experimenting some more. 1.000mV causes lots of stuttering.. no crashing but horrible performance. 1.075mV I got one crash in Diablo in a very intense graphical part of the game, bumped it up to 1.100mV and haven't had a crash yet. My other thought was, perhaps the card can't handle a sudden drop to 2D for a moment since the 2D voltage is only 0.875mV. I bumped the 2D voltage up to 1.100mV also so that there is no fluctuation if the card drops to 2D for a moment (alt-tabbing between apps etc.) Temps don't seem to be any different from stock voltage. I never get over 65deg.

Nope. Crashed at the *exact* same spot on the boss as last time, after playing all day with no problems at all.. at least 5-6 hours of play. It's either a driver issue that we have no solution for yet, or a hardware flaw. Probably going to call tomorrow and possibly just send the thing back.

 

Okay for me I'm going to see if stock volts xfire runs for your 5-6 hours. Mine was last I knew crashing like yours. I still suggest stock volts again and stock clocks, leave ulps back to standard and CCC 8.5.

 

Dell replaced my 7970 today and I no longer get the black screen crashes. I played Skyrim for hours, where before I couldn't make it 10 minutes. Guess there must have been a bad batch of cards. I suggest people call in and get the card replaced.

 

You guys need to overvolt to 1.075v it will fix the crashes

 

StZu, can you provide brief guideline of you overvolting the 7970m?

 

I can confirm that 1.075mV is stable and does not produce too much heat. The only game that still occasionally crashes for me, even with an overclock, is Diablo 3, which is a known issue plaguing random ATI and NVidia card models without much pattern.

I ran stable at 1.100mV, too, but backed off because there was no reason for it other than trying a benchmark or two. The other thing I did was set both my 2D and 3D clocks to 1.075mV because I run games at min framerates in the background, so sometimes I'd drop to 2D clocks on the desktop with the game still actually running. That could have caused potential problems with the split-second dips to 0.875mV at stock 2D clocks.

Between those two changes, I haven't had any black screens since the change earlier this week with over a dozen DX 9/10/11 games and benchmarks tried. I consider 1.075mV stable at this point.

 

I have a stable overclock at 875/1300 when I play Max Payne 3 and a few other games. The only time I got BSOD was when I was overclocked over 900. And I have an overvolted card, living in Las Vegas, with a 100F degree weather. My A/C is set to 80 degrees Fahrenheit. My card gets pretty hot at mid 70C, but it doesn't crap out like others have on the site.

 

is this the same problem the second batch of 7970M's had in the sager machines ?

 

the second batch i think was never sold, sager shipped it back to clevo/amd. they got the "new" 2nd batch that are supposed to be problem free...

 

i dont know about you guys but I have clevo 7970m and never crashed on bf3 so far, whether stock clocks or overclocked to 900/1400, it did get 95C once i think on stock drivers but seems to be running cooler on 12.7 driver. the only thing that crashed is 3dmark11 and only when overclocked. I got a bsod once but because of installing 12.6 beta the wrong way (the whole thing instead of only the driver).

 

Hi, sorry for the late responce.
Here is a link to the guide by SVL7
AMD 7970m - modified VBIOS

As he did not fuse (only link) his guide with the guide to actually flash, ill simplify it as it took me 2 days to fully understand everything and i'm sure not many will have this patience while having heart ache because of their underperforming GPUs.


Note: This is the most simplified guide for the Average first time user so more advanced individuals that have some software installed already (i.e HWinfo64, 7zip) can skip some steps and use different file names.
Also, pull this guide up on a cellphone or a separate computer to use a a reference when flashing to make things simple.

- Step 1 - download all needed software to be ready - first time users or people who are very unsure of what to do do everything in this order.

Warning: You are adding voltage to a GPU and are altering the clocks which in turn may prevent Dell from honoring your waranty if something is to go wrong. I take no responsibility for bricked cards, burned motherboards and houses.

Warning 2: I am not the creator nor the host of any of the download files so download at your own risk.

Warning 3: Even though this simple guide is meant to be as simple as possible, this is an advanced user procedure so be sure you KNOW WHAT YOU ARE DOING!

Lets begin,

What you need:

Hardware
USB Flash Drive (any size over 1GB will do)

Software
1. Extracting software (i.e 7zip) Download
2. Modded vbios files by svl7 Downloads | Tech|Inferno
a. Use the previously installed 7zip to extract the downloadedzipped file to your desktop to make things easier. If you dont know how to extract just right click the zipped file and select "extract to" and choose desktop, this will send all of the downloaded files to the desired location.
3. ATI Flash Download ATIFlash 3.95 | techPowerUp
a. This is also a zipped file so extract it to the desktop as well.
4. SVL7s sotware to make a bootable USB Drive Tech|Inferno Forum
a. To download you will need a techinferno account so take 2 minutes to create one.
b. Once again this is zipped so repeat the process, once again to the desktop to make things simpler.
5. HWinfo64 (to later monitor your temps to make sure you are safe) HWiNFO64 Download

- Step 2 - now that you are ready with all files and software standing by you need to make a bootable flash drive.
1. Plug in your flash drive into your computer (make sure its empty)
2. Load up the previously installed software to make a bootable flash drive by Right clicking it and chosing RUN AS ADMINISTRATOR.
3. Once the software is running do and make sure of the following.
a. Under Device it has to show you flash drive
b. Under File System it reads FAT32
c. Under Volume Label it reads the name of your flash drive
d. Quick Format is checked
e. Create a DOS startup Disk is checked
f. Using DOS system files located at: is checked
g. Click the 3 dots button (...) and specify the location od the dos files by chosing desktop (if you follwed my guide 1:1) and chose the file "Win98_files" and hit ok.
h. If everything is as i stated hit Start and you will then have a bootable drive! There are illustrated instructions for this step in the following guide: [GUIDE] AMD VBIOS flashing

- Step 3 - chose the bios flash you want, rename and throw it onto the bootable flash drive.
1. From SVL7's modded vbios files chose the one you want to flash to. I reccomend the Overvolted 1.075v with 900/1300 clocks which will give you a nice performance boost too, but you may chose another file at your descretion, remember, this is extra strain on your GPU! So dont get greedy!
2. The file we are flashing to is located under "overvolted & modified clocks " folder from the files you previously downloaded The name of the file is "Dell7970m-900-1300-ov1075v.ROM" Right click it, rename it as 1075vbio.ROM (make sure you add the .ROM to the file name as I have.)
3. Drag and drop the renamed 1075vbio.ROM file to the root of the flash drive (root is simply the flash drive itself with no folders.)
4. The ATIflash software we downloaded earlier? Drag and drop it into the root of the flashdrive as well (just the ATIflash file). Now your bootable flash drive should have 2 files - ATIFlash and 1075vbio.ROM.
5. You are ready to flash.

Step -4- To be on the safe side lets prepare a bit. Make sure to know and do the following.
1. The checksum for the 1075vbio.ROM file is 0x5700 (if you dont know what this is dont worry, you will know when we flash, but basically its meant to verify you are flashing the right file which we will check later)
2. The file size for the 1075vbio.ROM file should be 128kb when looking into you flash drive.
3. Make USB Flash Drive as your primary boot device just in case something fails and you need to do a blind flash (when the display does not work due to a bad flash and you need to guess what you are pressing to revert to normal)
a. Restart the computer and go into the bios by pressing F2 (with your bootable ah drive in your computer.
b. Under boot options make the USB Flash Drive your primary and first boot option.
c. Save and exit. Now everytime your usb is in your computer it will boot from the usb giving you a command line on a black screen background.

Step -5- flash
1. If you followed my guide 1:1 after you saved and exited you are now seing a command line, if you did not make the USB drive as your primary drive and want to take a risk of having a pain in the a** blind flash then restart your computer and hit F12 to go into the boot options menu and chose USB Flash Drive.
2. If everything is good you are seing a command line
Type: dir hit enter
3. You are now seing All of your files on the USB Flash drive which we know are atiflash and the 1075vbio.ROM file
Type: atiflash -i hit enter
4. You are now seing all of your adaptors (which should only be one as we are flashing an m17x r3/r4 with one video card) Now you will save the stock vbios from your card to later be able to revert to original vbios. To do so (the 0 is a zero);
Type: atiflash -s 0 stokvbio.ROM Hit enter
5. You have now saved your original vbios file to your flash drive with the file name "stokvbio.ROM", now verify your checksum on the modded vbios file we are flashing to.
Type: atiflash -cf 1075vbio.ROM Hit enter
6. You should be seing 0x5700 somewhere in the messege as your checksum, if you are, proceed, if not, review all steps and DONT FLASH.
7. Take a deep breath and;
Type: atiflash -p 0 1075vbio.ROM Hit enter and DONT TOUCH ANYTHiNG!!! Until the flash is complete, should take a few seconds.
8. When prompted, restart your computer by pressting Control+Alt+delete or pressing the alienware head powerbutton, and take out your flash drive.
9. If you flashes successfully you will load into windows normally.
Congradulations, you have flashed.

- Step 6 - Although you flashed you still need to make sure you are running cool and well.
1. Fire up HWinfo64 and check out your temperatures under the sensor button.
2. Look at the temperature of your GPU Thermal Diode, If you are idling under 60c you are good.
3. Now play a game while running HWinfo64 in the backgroud. Play for like 30minutes and look at the sensors again,
4. Check what was the maximum temperature for the thermal diode, if it was under 67c , you are good to go.


The end


Note: I wrote this at work from my iphone so if anyone sees an error please let me know, ill correct it. Thanks

 

You can do the same as above, but much more simplified with ATIWinflash. I've used it over 20 times on this machine alone without problems.

Download ATI Winflash 2.0.1.18 | techPowerUp

 

How do you use this one?

 

Black screen during games? I have same problem with all my laptops when I overclock to much.
Stock clocks are to high?

 

I just read this entire post. I'm a bit nervous now for my M17x to come in...I really don't want to deal with any issues

*prays for a completely perfect R4*

 

Most R4s are perfect. The reason you are seing many here with issues is because most people only post when they have issues, when everything is running smooth they have no need for the forum.

 

I asked for a replacement of my card and now everything is perfect 0 crashes so far and my card now is even cooler. It is a bad batch of cards, so if you are having any issue ask for a new card.

 

agreed.
makes me think whether we need to make a polling post on who got the perfect R4 w/ 7970M and who don't, so to help a lot of ppl not to judge before even owning or trying.

 

Elkay the dos bootable usb flash method is the safest. you may find you had success with flashing with atiflash... but someone could easily have a issue in windows and could brick the cards from a bad flash. SVL7 is well experienced and been doing this for a long time. Safer method is the usb bootable drive method.

okay been playing crash free (on the games without noted issues) on stock or greater clocks instead on an undervolt. strange but I have no issues in diablo 3 in xfire on 12.5 7900 mod driver from guru3d. At stock clocks I get 180fps peak in Diablo 3 in xfire. when i use 1.025v stock clocks i now have 215 fps peak. Why i don't know. seems some of our cards need more volts and some (maybe just mine) need less volts. 1.075v is fine for benching for me but if I use it for long time gaming i get the crashes. this has been testing in 5-6 hour game sessions. Currently my card seems to be able to actually overclock for full time gaming overclock once i undervolted. not sure why and not sure if I am the only one to have found this to be true. SVL7 has easily used stock clocks at .975v so it doesn't seem to require an overvolt for all cards. seems strange so many variations in cards.

 

Is everything still working well with your replacement GPU? What driver are you using?

 

... good question MrFox. If there really is mkre than just a couple of bad batch cards they are hiding that info well and we are all left guessing if we recieved a bad card or if all this is driver issues new screen booted stock vbios and bios to a red screen of black backround and red twxt in lines it looked like the matrix scren but in red. I captured it kn my phone camera. I had tk shut the entire system off.

 

Yep everything is still working fine for me. Right now I am using official 12.6 drivers but I have also played with 12.7 beta and with the old 12.5 beta with no problems.

 

Hi all,

I'm new to this forum, but since I experienced the same problem I decided to share it in here. Also sorry for incoming english mistakes it's not my first languange so I'll try my best .

I bought an m17x r4 with a 7970m for my girlfriend just when it went out. The first game she played on it was D3, which never crashed. Then came Skyrim, which maybe had a single crash on more than 170h of playing, the crash not being a black screen but a freeze (maybe a game crash).

The first game where she experienced the black screen + sound loop was Warhammer online. We renewed our account to play a bit awaiting for GW2 and she couldn't play more than half an hour without experiencing the black screen crash. Sometimes it happened after less than 2m of gameplay.

Meanwhile she went through the 2 first public Beta of GW2 with no crashes at all, since we both aimed to play this one, everything was still ok .. But since the last two stress tests, she now has black screen crashes on GW2 too :'(. During the firsts Beta A.Net communicated on the fact that the client wasn't yet optimized for using the GPU. Now that the release approaches, they surely made some, and I can't help but thinking this is related..

As many of you I tried running multiple 3D mark 2011 bench to stress the card and see if the crash came from this, but never did. It looks like this only happens in game, so I was also looking into a driver issue. Reading this thread started to confort me in that path and I was hoping to solve the issue by changing the voltage or underclocking the card. But then came the posts with those who got a working replacement card.. So I'm not sure what should I do.

I find it hard to believe that a hardware faulty card only crashes on certain games. So either Alienware decided to change stock voltage and they send back the same cards or they apply a driver fix ? I'm getting kind of confused and a bit disapointed by the lack of communication from them. Lucky for me these kind of forums are up .

I'll try to perfom the GPU mem test this evening, change the voltage and/or clock and do my test on Warhammer Online since this game has always crashed for me.

 

I have had my m17x for about 3 months now. And in the past month i have started noticing black screen crashes on certain games. Warhammer Space Marine mostly, but i just bought Darksiders 2 today and the same thing happened. Im wondering if i should contact Alienware and ask for a replacement card?

 

I did the memctlCL test and ended up with 50 failed iterations on Random blocks (25885781 total incorrect bits).. I'm inclined to call Dell support to ask for a replacement but I would like to try out the GPU voltage increase to see if it get things more stable. I wonder if doing so breaks the guarantee ? Will Dell refuse to perform the replacement if they see that the BIOS have been modified ?

Edit:Ok, since the original vbios ROM was provided with the vbios pack, I took the risk and changed the vbios to : Dell7970m-900-1300-ov1075v.

I still get the 50 iterations error when runing memctlCL, but I'll try running Warhammer (and GW2 during this evening stress test) and see how it goes.

Edit2: Just ran 3Dmark 11 test again. The tests went smooth, I went from a score of P5937 (original clock and voltage) to a P6255, temperature in max load between 68 and 72° . So far so good, lauching War online now.

Edit3: I played 5m of War and got the BSOD again :/. I'll maybe try and undervolted setup and if it doesn't work, I'll put the original vbios and ask for a replacement :'(.

 

So i called Alienware support last night. The guy remoted in to my laptop and ran a bunch of tests, basically two stress tests to see my max temps for the CPU and GPU. I told him the problem seems to be my GPU and brought up points from this thread. He said it sounds like overheating, and figures that the cpu is the cause of the problem since my max temp reached 90. The GPU temps were normal maxed out at around 74. So he set up an appintment with a tech to come and put in new fans and heatsinks. I really dont think that is going to change the problem, but if the black screen contiunes to happen after at least then i can say we tried your way, now give me a new card!

 

Hi axedatta,
Just for curiosity, did you run memtestCL ? Do you also have those iteration errors on Random blocks ? I ran it on 3 different configs (gtx 470, gtx 560 and the alienware 7970m), only the alienware gpu gives those errors.

 

I dont think we ran memtestCL. He ran unigine heaven and some CPU stress test.

 

The 7970M cards are seeming to be very inconsistent. Some are great, some are mediocre and some are terrible. There are several M18x R2 owners in our community that have had their GPUs replaced, as well as M18x R1 owners that upgraded to 7970M.

Use Alien Autopsy and run the Gaming Test (CPU, GPU, RAM, HDD) under custom scans. You may find the GPU fails. My VRAM fails the Gaming Test and I get a stop error now and then that references the video memory. I also see the issue with the screen going black in DX11 games and benchmarks that are more demanding (probably using the area of VRAM that has an issue).

 

Thanks for your answer Mr Fox,
I just ran the custom gaming test with Alien Autopsy and everything went fine. It is odd that the memory test doesn't fail when memtestCL shows errors :/


Just to be sure I'll run the test a couple of times.

 

Did anybody figure out a fix for this yet?? It's really getting annoying crashing after I get 5 stacks of nephalem valor!!!

 

Hi, I just spend about an hour with a Dell tech on phone, and after conviencing him that the problem comes from the graphic card, they will send a tech on tuesday to replace the card.

I'll be sure to redo all the basic tests: memtestCL, 3Dmark2011 etc, play WarOnline for more than an hour and give you feedback in here. I really hope it'll solve the problem .

 

Same thing with me, the tech came yesterday to put in new fans and a heatsink. 20 mins later the same problem still occurs, so then i finally called and asked for a GPU replacement. They agreed, hopefully this new GPU will not be defective as well.

Might have to wait a bit longer now, just got a robo call from dell stating that my parts have been delayed in processing.....

 

With the problems that come along with the 7970 right now, is it worth it to even order a new one with that configuration? Seeing as the 680 isn't being offered, I don't really want to settle with the 675 or 660 options. My warranty may be replacing my computer and I was planning on getting an R4, but now I am not too sure.

 

From everything i have read, these problems probably stem from a particular batch of early 7970's. I ordered mine in late May and received it early June. I think most people that have had problems were the ones that ordered relatively early. So i am hoping/assuming the newer batches of 7970's don't have that problem as evidenced by some people that have gotten replacement cards later which work perfectly. So i think you will be fine.

 

Just to give you an update:
The Dell tech came this morning, and replaced the card. After rebooting I got some pixels artefacts everywhere. We launched 3DMark11 and after half the first test we got a black screen. After this the laptop didn't pass the boot process and we got 8 bips (which is LCD problem if you look at Dell docs). However, after putting back the old card, everything went back to "normal". He called his support and stated that the new card was also faulty :/. He planned another intervention for tomorrow afternoon, hope that I'll get a working card this time !