cancel
Showing results for 
Search instead for 
Did you mean: 

R9 280x DirectCU II Top Artifacts!!

d1versify
Level 7
HEllo i just upgraded from an 7870 to the R9280X-DC2T-3GD5

I followed these steps for drivers

1. Ran Catalyst control center from C:/ AMD and did a custom uninstall. Removed all
2. Restarted computer
3. Ran Driver Fusion and deleted everything from AMD
4. Restarted computer
5. Installed latest 13.12 R9 drivers from AMD site

I get artifacts (shapes, lines , squares) flashing everywhere in screen , when in Chrome, in Dota 2 and Planetside 2. Haven't tested in more games , but I guess it will be the same.
The card will need a replacement? I have a Corsair CX600 , AMD FX 8350, Corsair Vengeance 2x4gb 1600 mhz
I read somewhere about GPU bios update. I ran the GPU ASUS TWEAK software that came in with the CD, and i tried to update something but i got a message ROM programming fail. But then it seemed to installed the update and forced me to restart the computer.
But still i get artifacts.
What do you suggest?
604 Views
678 REPLIES 678

70°C is the maximum operating temperature for those hynix memories. We could achieve better results just by putting custom heatsinks over the memories or using a waterblock.

Shadow of Mordor gives me artifacts with textures on ULTRA but if I define the FAN % to 100% those artifacts go away (an indication of overheating issue).

Games that uses less than 2gb of vram do not cause me problems BUT programs like MSI Afterburner and EVGA Precision (I thought it is pretty curious) makes the memories overheat somehow and I don't know why...

dotachin wrote:
Here's the bios with raised up timings to try for anyone who wants to experiment. With this bios card performs about 5% slower on same clocks, but helps to reach higher memory clocks without artifacting and without having to lower core clocks (helper to get +100MHz more on memory in my case, but may help better for others). The changed timings are only those altered by The Stilt in his optimizations.


Hello! I've passed my night to read all pages...
I've a technical question for you: how a "temperature" problem can be solved from "latency" tweaks? I don't get it.
Today I'm testing fan speed at 55% on GTA V, and VRam seems ok (max GPU temp 68°c, 69°c vRam).
Case HAF XM with a 20cm fan on the side 😛

psychok9 wrote:
I've a technical question for you: how a "temperature" problem can be solved from "latency" tweaks? I don't get it.
Timings specify how many cycles are given to the memory between operations. Raised timings reduces the amount of operations per same period and reduces heat. But slightly, of course,- it does not solve the problem, just allows a bit higher frequencies in my case.

psychok9 wrote:
The bios appears the oldest, and change my card frequency: 1100MHz on GPU and 1400MHz on Vram!
-200MHz on Vram and I don't see the latencies...
I've posted it to let test the raised timings - you easily can (and have to) change the clocks for your own card. About the bios version - I didn't notice any difference between different bioses in performance, however, i didn't focus on that, so it's possible that I've missed something.

psychok9 wrote:
The "optimized" bios hurts a lot! -15% of FPS!

Very interesting, didn't test it with GTA, as moved back to official bios. The Stilt's tunings reduces maximum clocks as bit. WIth GTA I'm using two cards, one on 1100/1850, second with 1150/1350. Yeah, GTA forced to lower memory clocks of problem card to 1350 MHz.

dotachin wrote:
Here's the bios with raised up timings to try for anyone who wants to experiment. With this bios card performs about 5% slower on same clocks, but helps to reach higher memory clocks without artifacting and without having to lower core clocks (helper to get +100MHz more on memory in my case, but may help better for others). The changed timings are only those altered by The Stilt in his optimizations.


The bios appears the oldest, and change my card frequency: 1100MHz on GPU and 1400MHz on Vram!
-200MHz on Vram and I don't see the latencies...

dotachin
Level 7
So the problem is/was basically bios conflict and not an overheating issue?
No, just read the comment above the attachment to see what that bios actually gives.

dotachin
Level 7
I define the FAN % to 100% those artifacts go away
Do artifacts go away in your case once appeared but just with changing fan speed - without having to restart game or PC, without having to start game with 100% fan initially?

dotachin wrote:
Do artifacts go away in your case once appeared but just with changing fan speed - without having to restart game or PC, without having to start game with 100% fan initially?


Nope, once the artifacts appear, the only way to fix is really closing the application and reopen it with changes. For me all artifacts gone with a modified bios (1070mhz core. 1500 vram - I modified the bios offered to download via Asus website) and custom fan settings, I only use the fans with 100% to determine that it all was really an overheating issue.

dotachin
Level 7
Nope, once the artifacts appear, the only way to fix is really closing the application and reopen it with changes
Same as in my case.

I only use the fans with 100% to determine that it all was really an overheating issue.
100% of fan is an issue itself.. Anyway, 1070/1500 is almost close memory clock to safe 1000/1500 or 1200/1400. I would like to use some watercooling for VRAM, but it's not possible to be sure - what if even with watercooling it will not be possible to reach 1600 on memory with 1100 on core? As I've said above, it's better to use 1200/1400 than 1000/1700 (i try to not refer to memory clocks without core clocks, as high core clocks are important to load the memory enough to make the artifacts appeat)

With my custom tool it's clear that from 960x3MB surfaces only 5-6 ones fail, the rest ones keep performing stable on intensive read/write up to 1850 MHz. And so far I've found no way to properly isolate bad areas to let the other 2990MB work on top frequencies.

dotachin wrote:
Same as in my case.

100% of fan is an issue itself.. Anyway, 1070/1500 is almost close memory clock to safe 1000/1500 or 1200/1400. I would like to use some watercooling for VRAM, but it's not possible to be sure - what if even with watercooling it will not be possible to reach 1600 on memory with 1100 on core? As I've said above, it's better to use 1200/1400 than 1000/1700 (i try to not refer to memory clocks without core clocks, as high core clocks are important to load the memory enough to make the artifacts appeat)

With my custom tool it's clear that from 960x3MB surfaces only 5-6 ones fail, the rest ones keep performing stable on intensive read/write up to 1850 MHz. And so far I've found no way to properly isolate bad areas to let the other 2990MB work on top frequencies.


I purchased this waterblock, but I didn't get it yet. When I do I'll update here with news about how the card performs! (:

delbiolima wrote:
I purchased this waterblock, but I didn't get it yet. When I do I'll update here with news about how the card performs! (:
€ 104.94 is a half of the card price, wasn't there anything cheaper?