Jump to content
EwinRacing Flash Series Gaming Chairs
kxplus

Windows crashes when eGPU/external monitor is utilised

Recommended Posts

Hi all, I have the following setup:

early 2015 rMBP 13"

Win 8.1 on external SSD

akitio thunder2 + gtx1060 6gb

400W gold PSU, with a 4pin EPS remade to barrel for the akitio, and a 6pin VGA into my GTX1060

external HDMI monitor connected to eGPU

 

Beyond the usual booting trouble, everytime I go into Windows and want to utilise the eGPU, whether by just displaying my external monitor in control panel or running benchmark apps like unigine, a black screen occurs.

Windows boots up with only my internal display showing, and control panel is able to reflect the name of my 2nd monitor properly.

However if I ever try to switch to extend/duplicate/2nd screen only, Windows crashes. I know it has crashed because the light on the caps lock key wont toggle when I press it. The only thing I can do at this point is force shut down.

 

In device manager, Windows shows my GTX 1060 just fine and even shows its status as "working properly". Yet, I cant utilise my eGPU.

I have the v327.54 geforce drivers already installed. I also tried v327.70 to no avail. The only time anything ever appeared on my external display during this eGPU thing (like, ever) is a flash of a white screen on both internal and external displays, before the macbook/windows crashes.

I have tried restarting countless times, either plugging thunderbolt cable in from cold boot or only when the windows insignia shows.

 

I am clueless as to what's wrong. Does anyone have any idea? I will greatly appreciate it!!

Edited by kxplus
Typo errors

Share this post


Link to post
Share on other sites

I use the switch on the psu. Paperclip trick done on the 20pin mobo cable. The akitio lights up and my 1060's fan spins fine though, even after windows crashes they continue to be lighted up/spinning. Does that mean anything?

6 hours ago, MVC said:

Sounds like power issue.

 

How is 400W PSU triggered on? 

Share this post


Link to post
Share on other sites

new progress: i can get my external display to show for about 10-20 seconds after windows boots up (i.e.. login screen, desktop), then it hangs after that and becomes a white screen, and the system is unresponsive (caps lock light doesnt toggle). I can consistently get this result each boot.

 

Since you suggested its a power problem, here are more details of my PSU:

FSP 400W aurum Gold 80+

ATX 12V V2.3 & EPS 12V V2.92

4pin EPS cable modded with 2 yellow and 2 black cables into screw terminal barrel plug for akitio. 6pin for GTX1060 comes either directly from PSU, or via molex to 6pin, no difference at this point as load is low.

 

I also tried using the akitio 60W power supply to test if it was the instability of my PSU, but the 60W one is worse and throws up a Code 43 every time.

 

The 400W PSU is old, but it is a high quality PSU.

My question is, the PSU can provide a stable voltage to keep the akitio and graphics card lit up consistently, so why cant it handle more than 20 seconds of just showing the external display?

Also, I have used the mobile and desktop 372.70 drivers. Both give the same result.

Any idea, @MVC?

Edited by kxplus

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Similar Content

    • By xerman
      Hello all! Im posting this here in hope to receive some orientation or help regargind my egpu project.
       
      I have my old and trusted Acer 5750G and I’m still using it everyday for work. I’m a photographer and it has worked quite well with some upgrades:
      -Upgraded i5 to i7 2630qm.
      -Upgraded ram, 8gb to 16 gb.
      -Upgraded HD to SSD.
       
      Some facts:
      – I have the unlocked bios so I was able to set the mpci port as “hot pluggable” and I can also deactivate the dedicated nvidia 630m from the bios itself.
      – I have bough this case [PCIe PCI-E EXP GDC External Laptop Video Card Dock / Laptop Docking Station (Mini PCI-E interface Version) ] that will arrive home in 10-15 days.
      – I have ordered a wifi usb card.
      I was thinking in getting a nvidia 1060 6gb and plug it to my old laptop using the mpci port. I ask for your advice and knoledge regarding this:
      1- Is this a good card choice?
      2- Will I face any problem with my setup or there is something important I should take into consideration?
      3- Will software that uses CUDA (photoshop and premiere, for example) get reinforced with the new card?
      4- How much processing power will get lost for the “small” port I will be using?
      Thanks a lot again, and sorry for my english mistakes!!
    • By endray
      Hey guys,
      I recently upgraded my M15x with a i7 920XM and a GTX 780M.
      Flashed the modified vbios, got a 330W brick, pulled the pin, installed newest drivers and did some gaming.
      Everything worked pretty well, only sometimes the 780M reached 93C and clocked down, but still ran stable, never crashed.
      Therefore I decided to get a finer TIM and repaste. Bought the Noctua NT-H1, carefully repasted and voila, temps dropped around 15C.
      But here comes the catch: when trying Furmark or some very intensive game, the GPU reaches 88C and screen immediately turns a random color. No downclocking, no limiting, just a sudden crash. Only thing I can do is a hard reset. This doesnt happen on any other occasion, can game on 80C for an hour.

      Do you guys have any sort of idea why the laptop crashes at 88C, while before it worked at 90+ without a problem?

      Things I tried:
      Used DDU and reinstalled the drivers, also tried older version.
      CMOS reset.
      Cleaned everything, no dust left anywhere.
      Carefully checked the GPU for any visible damage or old TIM residue.
      Tried undervolting and underclocking, only took longer to get to 88C and still crashed.
      For the sake of it I actually strapped a Corsair H45 AIO to the GPU heatsink. Now i get to max 55C with furmark, but the laptop still cashes after around 2 minutes.
       
      I would be happy for any suggestion, thanks!
    • By Quackingplatypus
      Hey tech inferno,
      I started my eGPU/docking station project a little while ago, using a cooler master elite 120 white as the enclosure, an akitio board, a gtx 970, and a 1tb samsung ssd (for external windows 10). Everything is powered by a 650 watt evga power supply, and the akitio is connected with a soldiered barrel plug to a chopped molex cable. The first time I tested it with a 750ti (before buying a 970) everything just worked. I pugged the thunderbolt in while windows was booted, and installed the latest drivers. The screen flickered, and that was it. Done. Everything worked. Optimus worked flawlessly, external monitor worked flawlessly, frame rates were amazing (for a macbook) and my computer itself stayed cool. I immediately returned my friend's 750ti and bought a 970. Once I got around to installing it I was at college, but it was the same as before. Once I got the computer to recognize the card (in device manager), I installed the drivers, the screen flickered, and everything worked, just faster. I had a blast playing some games that I had been missing while working on the project. For several days I was able to either connect the egpu by plugging it in while the computer was off, turning on the power supply, waiting 15 seconds, and then booting, holding alt, and booting onto my external ssd; or just booting into windows and then hotplugging my egpu. It always worked, and it was better than I ever expected. Then, one day, I came back after class, and nothing worked. No hotplugging, no successful boots, and no accelerated video. Once in a while I could get it to boot, but when it did that it was only because it didnt recognize the card at all, and just used iris pro. Can someone help me get my setup working again? My laptop is the 2015 macbook pro 15" (for which there are few guides). I'm hoping someone else has overcome this problem. I have tried reinstalling drivers, but it no longer freezes the screen and works, it instead requires me to restart, and when I do, it breaks. Major karma to anyone who helps me out. I can put pictures (or video of boot failures) if someone wants me to.
      thanks!
      -Hunter
    • By poboy975
      Just like the title says, I cant get the driver to install. I'm on Windows 10 x64 release 160906. The card works in 1920 x 1080 with the windows vga drivers. I've disabled driver signing, I'm in test mode. I've tried the nvidia driver versions, 365.19, 372.54, 372.70, 372.90. With modded inf's for each from laptopvideo2go.com They all start the install, but when it gets to the driver install, the screen goes black for about 30 seconds, then the computer just reboots. The drivers don't get installed. I've seen a bunch of threads around about Windows 8, but nothing that seems to answer for windows 10. Is there something that I'm missing to get the drivers installed?
       
      Update, found this thread. 
      I got the drivers installed following the directions, but after rebooting back into normal mode, the computer just reboots. I see the windows logo, screen goes black, then after about 30 seconds, it just reboots. This was using the 372.90 and inf files from the thread.
    • By BAZZINGA
      Hi everyone!

      As described in the title I have  a
      1) MBP (mid-2012) connected to an
      2) Akitio Thunder2 which holds a
      3) GTX 750 TI SC.

      Further specs are:

      HARDWARE
      4) Original Apple TB-Cable
      5) new PSUs
      5.a) 120 WATT
      5.b) 150WATTS
      6) various LCD-Displays
      6.a) 1280x1024
      6.b) 1920x1080
      6.c) 1600x1200
      6.d) three 1600x1200 at the same time 

      SOFTWARE
      7) latest NVIDIA-Driver (346.03.15f02)
      8) automate-eGPU.sh v0.9.8
      9) MAC OS X, El Capitan 10.11.6

      FURTHER INFO
      1) I have the MBP in clamshell mode
      2) so external monitor(s) is/are the main screen… not the MBP
      3) startup is running without any problems

      THE ISSUE
      1) MBP crashes 
      2) at random events
      3) even at low performance... browsing in chrome or anything else normal
      4) yes gaming works but also crashes… which is not surprising because 3)
      5) all of the configurations mentioned above are not stable
      6) even with a small monitor and the 150 WATT PSU
      7) Also no difference with three 1600x1200 displays and a 120 WATT or 150 WATT PSU. Also unstable.

      MY QUESTION(S)
      1) So what am I missing here?
      2) Any pointers?


      Many thanks in advance!
      Bazzinga

       
×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use. We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.