This is a split board - You can return to the Split List for other boards.

R9 290X overclocking and temperature numbers

#21Xeeh_BitzPosted 10/20/2013 7:10:55 PM
Snuckie7 posted...
Nope, the fan was actually 50% in Furmark, and 43% for the first 3DMark. Use your brain a little, and think about what makes sense.


Furmark test wasn't done with an oc

All OC tests were done with 100% usage on fan

except for Metro, which had no OC either apparently.

and Furmark was apparently below 1000 on the core also, by 70

So using short burst usages from the 3D Marks with an OC with 100% speedfan isn't exactly an indicator of anything.
---
Best Windows ever, Windows 8!
#22Snuckie7(Topic Creator)Posted 10/20/2013 7:22:00 PM
Not an indicator of anything? What about real world gaming temperatures?

For those expecting more real-life scenario hereís one game test with Metro 2033. There isnít really much to say here, although I would suggest you take a look at the temperature, this could be a very important piece of information.

---
Intel Core i7 3820 | EVGA X79 SLI K2 | MSI 7950 Twin Frozr III | Samsung / 840 120GB / 8GB RAM | 1TB WD Caviar Blue | Corsair / 550D / H70 | Silencer MKIII 600W
#23Xeeh_BitzPosted 10/20/2013 7:28:56 PM
Snuckie7 posted...
Not an indicator of anything? What about real world gaming temperatures?

For those expecting more real-life scenario hereís one game test with Metro 2033. There isnít really much to say here, although I would suggest you take a look at the temperature, this could be a very important piece of information.


Not an indicator of how it behaves when in a real life scenario while overclocked

You know, what we've been talking about
---
Best Windows ever, Windows 8!
#24Snuckie7(Topic Creator)Posted 10/20/2013 7:32:24 PM
Actually, we were talking about stock voltages and stock temps. Or did you forget that already?

Xeeh_Bitz posted...
Snuckie7 posted...
39% OC including a ~20% OC by default (GPU Boost) and voltage at 1.212V (no extra over "stock").

It's really not that much when you take into consideration what the card actually runs at at "stock".


If its like the old ghz edition cards, it's already past the max voltage for Nvidia.

Temps seem to suggest it being already high

---
Intel Core i7 3820 | EVGA X79 SLI K2 | MSI 7950 Twin Frozr III | Samsung / 840 120GB / 8GB RAM | 1TB WD Caviar Blue | Corsair / 550D / H70 | Silencer MKIII 600W
#25Xeeh_BitzPosted 10/20/2013 7:46:43 PM(edited)
Snuckie7 posted...
Actually, we were talking about stock voltages and stock temps. Or did you forget that already?

Xeeh_Bitz posted...
Snuckie7 posted...
39% OC including a ~20% OC by default (GPU Boost) and voltage at 1.212V (no extra over "stock").

It's really not that much when you take into consideration what the card actually runs at at "stock".


If its like the old ghz edition cards, it's already past the max voltage for Nvidia.

Temps seem to suggest it being already high


Yeah, overclock temps suggest its high at its stock voltage.

Should I spell it out for you?

One test, the card is underclocked by 70 (Default is 1000mhz, it ran the test at 930) and the other test has a glitch in the readings so obviously that test is dismissed.

So no, I'm not referring to the normal clock test which only one would be valid and thats the 3D Mark test, the other is glitched and the Furmark is underclocked.

All in all, we're discussing overclocks. You're the one trying to derail by talking about cards and stock settings, yada, yada...

Nvidia cards do 20% on stock voltage, that card does like 12%, it's memory can't go past 6000

"n this scenario our performance has decreased. Cheap memory modules which were used on R9 290X in favor of wider memory bus are not really capable of sustaining 6GHz+ clocks. This is probably the point where you should stop overclocking any reference R9 290X card."

---
Best Windows ever, Windows 8!
#26nIMr0D888Posted 10/20/2013 8:29:24 PM
Snuckie7 posted...
39% OC including a ~20% OC by default (GPU Boost) and voltage at 1.212V (no extra over "stock").

It's really not that much when you take into consideration what the card actually runs at at "stock".


BASE CLOCK, boost runs in the high 1400s to 1500s which is getting damn close to 70% over the base.
---
CAUTION: EXTREMELY HAZARDOUS SUBSTANCE
#27Snuckie7(Topic Creator)Posted 10/20/2013 8:43:33 PM(edited)
Xeeh_Bitz posted...

So you're dismissing every test there because of some silly theories you made up? Just to reiterate, there was no glitch for the fan speed in Metro, because both GPU-Z and Afterburner report the same fan speeds. 3DMark is a perfectly fine test for temperatures too; it takes about 30 seconds to heat a card close to the maximum temperature.

Furmark is really the only useless test there, it doesn't represent any kind of realistic load at all. The card was probably programmed to throttle for Furmark, hence the lower core clock.

Nvidia cards (GK110) don't overclock 20% past the clocks they would normally boost to at stock, at stock voltage.

How well the memory clocks is almost irrelevant because of how much bandwidth the 512 bit bus will give anyways. To put it into perspective, a 780/Titan would need 2000MHz on the memory to achieve the same memory bandwidth as a 290X at 1500MHz.
---
Intel Core i7 3820 | EVGA X79 SLI K2 | MSI 7950 Twin Frozr III | Samsung / 840 120GB / 8GB RAM | 1TB WD Caviar Blue | Corsair / 550D / H70 | Silencer MKIII 600W
#28Snuckie7(Topic Creator)Posted 10/20/2013 8:38:39 PM
nIMr0D888 posted...
Snuckie7 posted...
39% OC including a ~20% OC by default (GPU Boost) and voltage at 1.212V (no extra over "stock").

It's really not that much when you take into consideration what the card actually runs at at "stock".


BASE CLOCK, boost runs in the high 1400s to 1500s which is getting damn close to 70% over the base.

Are we even still talking about stock voltage? Or do you actually think 780's regularly boost to 1400-1500 MHz at stock voltage (let alone at all).
---
Intel Core i7 3820 | EVGA X79 SLI K2 | MSI 7950 Twin Frozr III | Samsung / 840 120GB / 8GB RAM | 1TB WD Caviar Blue | Corsair / 550D / H70 | Silencer MKIII 600W