I'm getting desperate. Hoping you could help me out a bit. I'm looking for thoughts as what to do with my video card situation... Any feedback would be appreciated.
While playing some games, the video will freeze, and I'll get a popup in the lower-right that says that the display driver has stopped responding and has recovered. Some games can recover after a short freezing period, otherwise I either get dropped to desktop or have to end game's task to get out.
Examples of games that crash: Heroes of the Storm, Battlefield 4, CS:GO, PvZ: Garden Warfare, Insurgency
The event viewer has this after a crash:
Log Name: System
Date: 9/5/2015 8:12:41 AM
Event ID: 4101
Task Category: None
Display driver amdkmdap stopped responding and has successfully recovered.
- I've been living in this article
- Tried all of these reg settings:
Increasing the timeout just increases the time between when the video freezes and the popup appears.
Disabling the TDR system just makes the computer crash instead of just the display driver.
- Ran memtest86 for almost eight hours, RAM checks out fine
- Updated every mobo related driver I could, as I saw conflicts with Realtek audio/network drivers when googling.
- Tried all of the driver versions I can get my hands on, including betas. Haven't tried the very latest that came out this week because of the memory leak warning. (I use DDU to install as cleanly as possible.)
- Tried disabling on board nic, using wireless desktop card
- Thought it may have had something to do with the google hangout that we use for comm when we play, so I tried different browsers, no browser running, etc.
- Tried uninstalling my vpn software for work, closing everything except game related programs.
- Tried not using front 3.5mm ports for my headphones/not using headphones at all. New USB only headphones also don't stop error.
- Replaced my 620W Antec PSU with a 850W.
- Ran WD Lifeguard's extended test (3.5 hours) on the secondary HDD with no errors.
Samsung's Magician doesn't seem to have a test like that, but SMART reports no errors. Instead, I used it to set the drive up with "Max Reliability" settings (other two choices were performance and capacity). Not sure SSDs do a big sector scans anymore, as it may lower lifespan. At least, I only saw a performance test, anyways.
- Reinstalled Windows 10 with a clean install.
And yes, I RMA'ed my video card. Sapphire sent me a new one, and told me that my old one was missing a capacitor and this new one should work better. But I still had the same problems with the new one, so they said that they would send me a 2nd RMA card, this one a "Vapor-X" version that was 100 mHz faster than the original.
Installed the 2nd RMA and it still does the same thing. They told me to scrap the 1st RMA, so technically I have two GPUs here that do the same thing.
Just so you get an idea how off the rails I've been lately, I actually got out a spare HDD and tried installing SteamOS on it to see if I could play CS:GO there. SteamOS had horrible download speeds and ran the game like garbage. So I don't think I learned much except that I don't like SteamOS.
So there are some things I'm confident in:
- HDD - passed tests
- RAM - passed tests
Things I think are OK:
- SSD - no errors detected
- MOBO - older, but on latest BIOS, not much else I can do here
- CPU - passed a short stress test w/o error
- GPU - at least one of these two cards has to work...right?!
- Windows 10 - had same issues in 8.1
Things I'm less sure about:
- AMD - drivers... kind of held hostage here
- Some stray BIOS setting? I've looked, and tried a bunch of changes, don't see anything obvious.
- Something else that a kind, friendly reader of this post will suggest?
Like I said, any thoughts appreciated.
My full hardware PC specs can be found here:
Thanks in advance,