![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Computer is freezing on me. No blue screens and I'm not getting anything useful in the Event Viewer. Just complete freezes. Mouse locks, keyboard unresponsive, no activity lights.
No overclocking anything - CPU or GPU. GPUs were unlocked to 6970, but I restored those at the beginning of all of this.
The only culprits left I can blame are GPU (6950s in crossfire (only 1 atm)) and PSU.
I've tested all memory (both memtest and windows 8 built in - took quite a while - did each stick then started adding in channels). All passed. I've removed all but 1 SSD, and I've even swapped that drive around to ensure it wasn't an SSD issue. I've reinstalled Windows about a dozen times during all of this testing.
Both GPUs have individually passed furmark's 15m burn in. No artifacts, nothing indicative of bad overclocking (which I don't). The freezing happens with both GPUs installed, or just 1, and it happens to either single GPU (which is why I was starting to lean towards PSU).
PSU is nearly brand new. Bought it about 3 months ago after my Xion blew up. Antec HCG 900w. Should be more than enough for the full system.
About 2 weeks ago I added a Kuhler 620 for cpu. Prime95 and FFT-z both run w/o issue, load never breaks 50c. Currently idling between 28-30c.
Additionally, I know it's not the CPU / mobo (i7 3820 + p9x79) because on an impulse last week I went to MC and bought an fx8350+crosshair forumula z, which I've since been doing the bulk of this testing on.
I've also swapped nic for usb nic.
Additionally x 2, this only happens at the windows desktop (ie 0 load), never while the gpu is actually doing anything. As I said, furmark and unigine run like champs. GPU temps hover around 65c in load (32c idle) - I've left furmark and uni both just running for an hour+, no issues.
Not sure what else to do.
The only thing I haven't done is resort back to my original EVO 212 cooler. Once I had it installed on my lga2011 I tossed all of the brackets needed to install in on an am3+ board. I can switch back to my p9x79 board to stop using the kuhler, but completely dismantling the system yet again is a real pita, esp if something else (to you guys) looks immediately more like the issue. I do have the stock fx8350 fan which I can swap if I need to eliminate the kuhler.
Can the kuhler pump cause instability over 12v?
PS - I've even gone so far as doing the oven/reflow solder trick on both of these 6950s. No differences.
I don't want to buy new video card(s) if I can avoid it, haven't even had these in service for a year yet.
[apologies for the pisspoor grammar, copying/pasting this post to notepad and going fast - had 1 freeze already]
No overclocking anything - CPU or GPU. GPUs were unlocked to 6970, but I restored those at the beginning of all of this.
The only culprits left I can blame are GPU (6950s in crossfire (only 1 atm)) and PSU.
I've tested all memory (both memtest and windows 8 built in - took quite a while - did each stick then started adding in channels). All passed. I've removed all but 1 SSD, and I've even swapped that drive around to ensure it wasn't an SSD issue. I've reinstalled Windows about a dozen times during all of this testing.
Both GPUs have individually passed furmark's 15m burn in. No artifacts, nothing indicative of bad overclocking (which I don't). The freezing happens with both GPUs installed, or just 1, and it happens to either single GPU (which is why I was starting to lean towards PSU).
PSU is nearly brand new. Bought it about 3 months ago after my Xion blew up. Antec HCG 900w. Should be more than enough for the full system.
About 2 weeks ago I added a Kuhler 620 for cpu. Prime95 and FFT-z both run w/o issue, load never breaks 50c. Currently idling between 28-30c.
Additionally, I know it's not the CPU / mobo (i7 3820 + p9x79) because on an impulse last week I went to MC and bought an fx8350+crosshair forumula z, which I've since been doing the bulk of this testing on.
I've also swapped nic for usb nic.
Additionally x 2, this only happens at the windows desktop (ie 0 load), never while the gpu is actually doing anything. As I said, furmark and unigine run like champs. GPU temps hover around 65c in load (32c idle) - I've left furmark and uni both just running for an hour+, no issues.
Not sure what else to do.
The only thing I haven't done is resort back to my original EVO 212 cooler. Once I had it installed on my lga2011 I tossed all of the brackets needed to install in on an am3+ board. I can switch back to my p9x79 board to stop using the kuhler, but completely dismantling the system yet again is a real pita, esp if something else (to you guys) looks immediately more like the issue. I do have the stock fx8350 fan which I can swap if I need to eliminate the kuhler.
Can the kuhler pump cause instability over 12v?
PS - I've even gone so far as doing the oven/reflow solder trick on both of these 6950s. No differences.
I don't want to buy new video card(s) if I can avoid it, haven't even had these in service for a year yet.
[apologies for the pisspoor grammar, copying/pasting this post to notepad and going fast - had 1 freeze already]