I have been working on a pet project using spare parts pulled from various systems built in the past 3 years or so and through bargain hunting. On a cheaply acquired Thermaltake Core P5 chassis, I assembled a watercooled gaming PC oriented rig.
• AM3+ MSI 970 Gaming MOBO
• AMD FX 6300 CPU
• 2 x 4GB Corsair Vengeance DDR3 RAM
• EVGA 750W 80+ Bronze semi modular Power supply
• Sapphire Radeon HD 7950 Ghz edition
• Sapphire Radeon HD 7970
• 120GB Kingston SSD + 240GB Kingston SSD
• Custom watercooling loop (D5 PWM pump, 480 mm radiator, 2 x XSPC full cover block reference 7970 PCB design)
All components were tested independently using another motherboard (intel socket) then again together on the AM3 motherboard prior to final assembly and watercooling.
I am an occasional gamer only and was planning to mainly play Rise of the Tomb Raider, Mirrors Edge Catalyst and Quantum Break with this build. The objective was to approach 60 fps on high settings with the hope of taking advantage of Directx 12 compatibility with the older 7900 series.
Fresh installation of Windows 10 followed by upgrade of display drivers to Radeon Software Crimson ReLive edition 17.1.2 led to Solid performances in 3D Mark and Rise of the Tomb Raider with an average 57 fps using the built in benchmark. I decided to try a slight overclock on both cards which led to freezes and a crash. After a reboot and dialing back the overclock, all was good again. I enjoyed a few hours of smooth gameplay, whisper quiet and chill operation with temperatures remaining below 50 Celsius.
That’s when all started really going downhill. Turning the computer back on the next day I started experiencing flickers and temporary black screens. I was also unable to open Radeon Settings or to launch Steam (nothing would happen when icons were pressed). Then the computer rebooted on its own going through a loop of reboot to failure.
Having been exposed to driver display failures in the past, I set in action cleaning up the drivers with Guru 3d Display Driver Uninstaller and started reinstalling the drivers in Safe Mode. Half way through the installation, the computer crashed to a black screen and rebooted to Windows with uninstalled drivers. This would happen with any versions of drivers I tried subsequently, working my way back to 15.7 to no avail. Troubleshooting, I tried my trusted HD 5450 to confirm that the other components are viable, and all works flawlessly – besides gaming of course which is clearly out of scope.
I found several posts, blog online including on the AMD website with no definitive fix offered to this driver issue. With several threads being closed with a desperately discouraging “hardware failure”. I find it hard to call a graphics card dead or fried and it would still display on basic drivers.
Which all leads me to today. After 2 weeks of running around in circle, I have made no progress.
Has anyone successfully recovered from a similar situation and could share the steps taken?
Is there a sure-fire way to confirm that it is in fact a hardware failure? (visible signs on PCB or software based test). I do admit however being reluctant to the idea of moving the cards back to another system
I anticipate several comments or recommendations along the line of “upgrade to a RX 480 or GTX 1060” and would like to say that while these kinds of upgrade make sense for most gamers willing to be back in action right away, in my case they do not really apply.
• AM3+ MSI 970 Gaming MOBO
• AMD FX 6300 CPU
• 2 x 4GB Corsair Vengeance DDR3 RAM
• EVGA 750W 80+ Bronze semi modular Power supply
• Sapphire Radeon HD 7950 Ghz edition
• Sapphire Radeon HD 7970
• 120GB Kingston SSD + 240GB Kingston SSD
• Custom watercooling loop (D5 PWM pump, 480 mm radiator, 2 x XSPC full cover block reference 7970 PCB design)
All components were tested independently using another motherboard (intel socket) then again together on the AM3 motherboard prior to final assembly and watercooling.
I am an occasional gamer only and was planning to mainly play Rise of the Tomb Raider, Mirrors Edge Catalyst and Quantum Break with this build. The objective was to approach 60 fps on high settings with the hope of taking advantage of Directx 12 compatibility with the older 7900 series.
Fresh installation of Windows 10 followed by upgrade of display drivers to Radeon Software Crimson ReLive edition 17.1.2 led to Solid performances in 3D Mark and Rise of the Tomb Raider with an average 57 fps using the built in benchmark. I decided to try a slight overclock on both cards which led to freezes and a crash. After a reboot and dialing back the overclock, all was good again. I enjoyed a few hours of smooth gameplay, whisper quiet and chill operation with temperatures remaining below 50 Celsius.
That’s when all started really going downhill. Turning the computer back on the next day I started experiencing flickers and temporary black screens. I was also unable to open Radeon Settings or to launch Steam (nothing would happen when icons were pressed). Then the computer rebooted on its own going through a loop of reboot to failure.
Having been exposed to driver display failures in the past, I set in action cleaning up the drivers with Guru 3d Display Driver Uninstaller and started reinstalling the drivers in Safe Mode. Half way through the installation, the computer crashed to a black screen and rebooted to Windows with uninstalled drivers. This would happen with any versions of drivers I tried subsequently, working my way back to 15.7 to no avail. Troubleshooting, I tried my trusted HD 5450 to confirm that the other components are viable, and all works flawlessly – besides gaming of course which is clearly out of scope.
I found several posts, blog online including on the AMD website with no definitive fix offered to this driver issue. With several threads being closed with a desperately discouraging “hardware failure”. I find it hard to call a graphics card dead or fried and it would still display on basic drivers.
Which all leads me to today. After 2 weeks of running around in circle, I have made no progress.
Has anyone successfully recovered from a similar situation and could share the steps taken?
Is there a sure-fire way to confirm that it is in fact a hardware failure? (visible signs on PCB or software based test). I do admit however being reluctant to the idea of moving the cards back to another system
I anticipate several comments or recommendations along the line of “upgrade to a RX 480 or GTX 1060” and would like to say that while these kinds of upgrade make sense for most gamers willing to be back in action right away, in my case they do not really apply.
Facebook
Twitter
Instagram