- Jun 18, 2014
- 8
- 0
- 4,510
Hi guys, long time lurker first time poster here. I turn to you because in years of building my own rigs, I have never run into a problem that I could not solve like this.
First, let me get to the issue. I noticed artifacting, but it was not like the usual artifacts I would get from a failing GPU or heating. This is not my image, but is EXACTLY what I was getting:
http://
The links are to imgur.
While annoying, a quick alt-tab would solve it lickity split so I didn't think too much about it. I ran memtest to cover that angle too, and sure enough, I failed on test 8 once. Not wanting to chance it, I RMA'd the RAM to Corsair. (I want to point out that this solved a different issue I was having with lockups every hour or so for 20 seconds. Probably best to bring it up even if it seems unrelated). I used DDU to clean my drivers, tried a couple of revisions, made sure all my other drivers were up to date, tweaked the BIOS settings, tried with other RAM that is known to be good, no go. No matter what game was loaded up, I would get the chessboard. Sometimes it was black and white, sometimes it was colored based on what was behind it on screen. Temps were 50c under stress, sometimes a tad higher if it was hot outside, and about 30c idle.
My specs:
CPU- AMD FX 6350 at stock speeds w/ Cooler Master Vortex
GPU- Formerly MSI HD7790 factory OC @2g, now XFX DD Edition R9 270x @2g
RAM- Corsair XMS 1600mhz 2x4g DDR3
PSU- Some cheap 750w Coolmax (I'm so sorry, never again), now Antec Blue 750w
Mobo- MSI 970A-G46
Drive- Samsung 150g SSD
The list of things I tried:
-Different monitor
-Different HDMI/VGA/DVI cable
-Different power strip/different outlet
-pointing a fan at my open case (even though temps were low, you never know)
-reinstalled all AMD drivers (wiped with DDU)
-updated BIOS and chipset
-made sure my SSD was current on firmware
-cleaned DIMMs and PCI-E slot
Feeling like it was time to upgrade anyways while I was waiting, I decided to ditch my MSI HD7790 for an XFX R9 270x. While I was at it, I dropped the Coolmax for an Antec 750 Blue that was on sale. Should be the end of that, right?
I get home, toss everything together with the newly arrived Corsair RAM, and things are going great. On the second night, Mr. Boxes decides to drop in uninvited. I could feel my heart sink. On top of the boxes, I now had delayed loading of textures in a game I was playing called SMITE on max settings and some pretty aggravating z-fighting on bright things (glow, lights, fires, etc).
I then ran OCCT, which reported errors (though the test did not end, and the graphs at the end reported '0 errors' in them).
I reinstalled Windows (which was an effort and a half thanks to MS' policy to OEM), fresh installed the drivers that came with the 270x to test before grabbing the newest ones, and ran OCCT again.
This time I detected significantly less errors (900 vs 16000) and again the test did not stop automatically. This time, I saw this in the graphs:
http:// (another imgur link)
For about 15 minutes, my FPS was abysmal, which coincides with the weird flatlining of the vin0. Is this normal? I know software tests are not the best judge, but it looks to go hand-in-hand with the period of lower FPS. For some reason, my 12v rail is also as having 15v on the monitoring screen.
I am testing in games right now to see if the checkers come back, but it can take some time.
I'm at wits' end, really. The only two things left in the box that were originally there are the motherboard and the SSD. With regards to the mobo, I am unable to test the card in a different PCI-E slot due to the unfortunate placement of the SATA outlets.
I'm looking at a second RMA, this time for a second new GPU, and I'm very hesitant to buy a new motherboard, only to have another 100 and some odd dollars fall down the artifact rabbit hole again. Of course, I will if I have to. When I examined the mobo and cleaned it, I did not notice any scoring, cracked soldier/parts, or bulging/leaking capacitors.
What do you think? Is that OCCT result normal? If not, what could be the cause? Did I really get a second GPU that just so happened to result in the exact same type of artifacts at the same size, or is my mobo slowly killing parts?
Thank you greatly, for anyone who braved that wall of text. I'll hear any idea, no matter how wild. I'm darn near calling an exorcist for my poor computer.
First, let me get to the issue. I noticed artifacting, but it was not like the usual artifacts I would get from a failing GPU or heating. This is not my image, but is EXACTLY what I was getting:
http://
The links are to imgur.
While annoying, a quick alt-tab would solve it lickity split so I didn't think too much about it. I ran memtest to cover that angle too, and sure enough, I failed on test 8 once. Not wanting to chance it, I RMA'd the RAM to Corsair. (I want to point out that this solved a different issue I was having with lockups every hour or so for 20 seconds. Probably best to bring it up even if it seems unrelated). I used DDU to clean my drivers, tried a couple of revisions, made sure all my other drivers were up to date, tweaked the BIOS settings, tried with other RAM that is known to be good, no go. No matter what game was loaded up, I would get the chessboard. Sometimes it was black and white, sometimes it was colored based on what was behind it on screen. Temps were 50c under stress, sometimes a tad higher if it was hot outside, and about 30c idle.
My specs:
CPU- AMD FX 6350 at stock speeds w/ Cooler Master Vortex
GPU- Formerly MSI HD7790 factory OC @2g, now XFX DD Edition R9 270x @2g
RAM- Corsair XMS 1600mhz 2x4g DDR3
PSU- Some cheap 750w Coolmax (I'm so sorry, never again), now Antec Blue 750w
Mobo- MSI 970A-G46
Drive- Samsung 150g SSD
The list of things I tried:
-Different monitor
-Different HDMI/VGA/DVI cable
-Different power strip/different outlet
-pointing a fan at my open case (even though temps were low, you never know)
-reinstalled all AMD drivers (wiped with DDU)
-updated BIOS and chipset
-made sure my SSD was current on firmware
-cleaned DIMMs and PCI-E slot
Feeling like it was time to upgrade anyways while I was waiting, I decided to ditch my MSI HD7790 for an XFX R9 270x. While I was at it, I dropped the Coolmax for an Antec 750 Blue that was on sale. Should be the end of that, right?
I get home, toss everything together with the newly arrived Corsair RAM, and things are going great. On the second night, Mr. Boxes decides to drop in uninvited. I could feel my heart sink. On top of the boxes, I now had delayed loading of textures in a game I was playing called SMITE on max settings and some pretty aggravating z-fighting on bright things (glow, lights, fires, etc).
I then ran OCCT, which reported errors (though the test did not end, and the graphs at the end reported '0 errors' in them).
I reinstalled Windows (which was an effort and a half thanks to MS' policy to OEM), fresh installed the drivers that came with the 270x to test before grabbing the newest ones, and ran OCCT again.
This time I detected significantly less errors (900 vs 16000) and again the test did not stop automatically. This time, I saw this in the graphs:
http:// (another imgur link)
For about 15 minutes, my FPS was abysmal, which coincides with the weird flatlining of the vin0. Is this normal? I know software tests are not the best judge, but it looks to go hand-in-hand with the period of lower FPS. For some reason, my 12v rail is also as having 15v on the monitoring screen.
I am testing in games right now to see if the checkers come back, but it can take some time.
I'm at wits' end, really. The only two things left in the box that were originally there are the motherboard and the SSD. With regards to the mobo, I am unable to test the card in a different PCI-E slot due to the unfortunate placement of the SATA outlets.
I'm looking at a second RMA, this time for a second new GPU, and I'm very hesitant to buy a new motherboard, only to have another 100 and some odd dollars fall down the artifact rabbit hole again. Of course, I will if I have to. When I examined the mobo and cleaned it, I did not notice any scoring, cracked soldier/parts, or bulging/leaking capacitors.
What do you think? Is that OCCT result normal? If not, what could be the cause? Did I really get a second GPU that just so happened to result in the exact same type of artifacts at the same size, or is my mobo slowly killing parts?
Thank you greatly, for anyone who braved that wall of text. I'll hear any idea, no matter how wild. I'm darn near calling an exorcist for my poor computer.
Facebook
Twitter
Instagram