290x possible bad OC?

[Mr.Meerkat]

So my single 290x which I'm currently running is OCed at 1100 core and 1350 Memory (cooled with a h75 and yes, after running it for 1 and a bit years, I'm finally going to get some VRM heatsinks but they never went over 90c anyway, usually sits at 70-80c gaming).

 

Problem:

I can easily stress test it and what ever with out artifacting (or crashing) but as I'm playing rainbow 6 siege, whenever I do something specific (reinforce a hatch so they can't drop through), it would artifact for a millisecond then be completely fine, it's the only place I can find for it to artifact. 

 

So is it a bad overclock? like it's stable, never crashed before for a year now, never artifacted until just recently playing R6S, in that specific scenario but it's only for a split second. It's more the fact that I can stress test it with out any artifacting and temps are all in check (never goes above 80c core and 90c for VRMs) and it's been fine for a year.

[CarterDavison]

As soon as you get those heatsinks, overvolt it a bit.

 

Should stable it out.

[xnoobftw]

You can always revert back to stock and test it out again! 

[Mr.Meerkat]

1 minute ago, CarterDavison said:

As soon as you get those heatsinks, overvolt it a bit.

 

Should stable it out.

Already overvolted it at 100mV with +50% power... I only got the heatsinks as I didn't want it to go above 85c stressing.

Just now, xnoobftw said:

You can always revert back to stock and test it out again! 

I could...but I'm just confused why it's only that single second of animation makes it artifact when stress testing it should obviously be more stressful. 

[CarterDavison]

Just now, Mr.Meerkat said:

Already overvolted it at 100mV with +50% power... I only got the heatsinks as I didn't want it to go above 85c stressing.

I could...but I'm just confused why it's only that single second of animation makes it artifact when stress testing it should obviously be more stressful. 

Personally, i'm only very familiar with Maxwell but if you could somehow test your temperatures lower and see if it artifacts then.

 

Max those H75 fans out and see if you artifact with lower temps (Sorry ears)

[xnoobftw]

Just now, Mr.Meerkat said:

Already overvolted it at 100mV with +50% power... I only got the heatsinks as I didn't want it to go above 85c stressing.

I could...but I'm just confused why it's only that single second of animation makes it artifact when stress testing it should obviously be more stressful. 

woah wait you're on +100mV and +50% power limit but your OC is only up to 1100/1350? On mine it went up to 1155/1580 and i thought i had a bad card...

What PSU are you using then?

Does it ALWAYS occur or was it just a 1 time thing?

[Mr.Meerkat]

Just now, xnoobftw said:

woah wait you're on +100mV and +50% power limit but your OC is only up to 1100/1350? On mine it went up to 1155/1580 and i thought i had a bad card...

What PSU are you using then?

Does it ALWAYS occur or was it just a 1 time thing?

A 850w XFX pro XXX whatever PSU and I know, I only have it there as you know, need VRM heatsinks as 5 extra Mhz raised VRM temps by 5c...and decreasing either of them made no difference in temps so may as well  

It ALWAYS occurs... I decided to leave a stress test overnight and lowered the core down to the good old 1075Mhz when I was crossfiring and it made no difference, just that specific animation makes it god damn artifact...

[CarterDavison]

1 minute ago, Mr.Meerkat said:

lowered the core down to the good old 1075Mhz when I was crossfiring and it made no difference, just that specific animation makes it god damn artifact...

In that case, test the card at stock and see if it artifacts. If so, it's not a card issue.

[Mr.Meerkat]

8 minutes ago, CarterDavison said:

Personally, i'm only very familiar with Maxwell but if you could somehow test your temperatures lower and see if it artifacts then.

 

Max those H75 fans out and see if you artifact with lower temps (Sorry ears)

Frankly, there's no point with my system as I have it in push pull and maxing it out only decreases idle temps by about 5c then underload, it takes much much longer but after 1 hour of gaming, it's one again sitting at about 70c (when my room is warm or 60c when a window is open)

 

edit, sure, will try running it at stock.

[CarterDavison]

Just now, Mr.Meerkat said:

Frankly, there's no point with my system as I have it in push pull and maxing it out only decreases idle temps by about 5c then underload, it takes much much longer but after 1 hour of gaming, it's one again sitting at about 70c (when my room is warm or 60c when a window is open)

That wasn't my point at all..

[Mr.Meerkat]

Just now, CarterDavison said:

That wasn't my point at all..

I know, but temps are only 5c lower to start off with then it slowly increases over a period of one hour to the same temp  

[CarterDavison]

Just now, Mr.Meerkat said:

I know, but temps are only 5c lower to start off with then it slowly increases over a period of one hour to the same temp  

So what are your temps after an hour? You said 80c earlier and now you're saying 70c...

[xnoobftw]

6 minutes ago, Mr.Meerkat said:

A 850w XFX pro XXX whatever PSU and I know, I only have it there as you know, need VRM heatsinks as 5 extra Mhz raised VRM temps by 5c...and decreasing either of them made no difference in temps so may as well  

It ALWAYS occurs... I decided to leave a stress test overnight and lowered the core down to the good old 1075Mhz when I was crossfiring and it made no difference, just that specific animation makes it god damn artifact...

Just give stock settings a try and report back!

[Mr.Meerkat]

Just now, CarterDavison said:

So what are your temps after an hour? You said 80c earlier and now you're saying 70c...

80c during the summer, 70 right now or infact 60 if I let the cold air from outside in

[BarisERY]

Could be a driver issue try a fresh install of drivers or old ones if youve just updated

[Mr.Meerkat]

Just now, BarisERY said:

Could be a driver issue try a fresh install of drivers or old ones if youve just updated

Last update I did was before Christmas when the crimson drivers came out...

[BarisERY]

Just now, Mr.Meerkat said:

Last update I did was before Christmas when the crimson drivers came out...

My point still stays drivers do get corrupted sometimes it cant hurt uninstalling and reinstalling them right?

[Katorice]

R9 290 here,

+25mV  +44%p

1100 (OC)core, 1250 (stock)mem

core--idle 38°c, load 68°c

vrm--idle 37°c, load 80°c

cooler: H55+Kraken G10

fan speeds: min 40%, max 85%

 

I never touched my mem's clock speed. but would doing so yield any more fps?

I don't want to add any more heat because my H440 is suffocating my rad

[Guest]

could just be the game

 

some games are finicky with OC's, if that's the only game that you're having issues with.. 

[Pesmerga]

Tri-x 290x here

Current core clock: 1098

mems: 1400

power limit: +20

voltage: stock

 

It's perfectly stable, works good on 1100 as well and 1105 but during a long folding session (~10 hours) crashed so I dialed back to 1098 and seems totally stable, I'd add voltage, but I'm scared to break something

 

Temps while folding: 69-74°c depending if case fans are full blast or at minimum.

[Morgan MLGman]

11 minutes ago, Katorice said:

R9 290 here,

+25mV  +44%p

1100 (OC)core, 1250 (stock)mem

core--idle 38°c, load 68°c

vrm--idle 37°c, load 80°c

cooler: H55+Kraken G10

fan speeds: min 40%, max 85%

 

I never touched my mem's clock speed. but would doing so yield any more fps?

I don't want to add any more heat because my H440 is suffocating my rad

R9 390 is a 290 with better and faster memory. See the difference in performance?

[xnoobftw]

1 hour ago, Katorice said:

R9 290 here,

+25mV  +44%p

1100 (OC)core, 1250 (stock)mem

core--idle 38°c, load 68°c

vrm--idle 37°c, load 80°c

cooler: H55+Kraken G10

fan speeds: min 40%, max 85%

 

I never touched my mem's clock speed. but would doing so yield any more fps?

I don't want to add any more heat because my H440 is suffocating my rad

From my experience OCing the mem did yield more FPS but not of a significant value (~1-2FPS) 

OCing mem is much more risky because mem clock speeds are more "fragile" as compared to core clock speeds