R9 290/X AIB Cards

Powercolor 290.
Flashed with XFX290X BIOS unlocked to full 290X.
Arctic Xtreme III cooler, fans running at 12V (still silent).

The Xtreme III cooler isnt designed for the 290 series (even though they make it seem so), one memory chip can only fit a 1/2 size heatsink because the GPU cooler overhangs it and there arent enough memory sinks. The 1/2 size memory sink hasnt affected memory overclocks badly though, I still got to 1620MHz with +100mV on the core (quad pumped to 6480MHz).
There are a lot of heatsinks in the package, you get to decide which you want to put where.
I used L shaped heatsinks on the VRMs.
Paired up some 1/2 size heatsinks and used a double size heatsink for the remaining memory chips.

All sinks glued on with provided glue.
 
Nenu said:
Powercolor 290.
Flashed with XFX290X BIOS unlocked to full 290X.
Arctic Xtreme III cooler, fans running at 12V (still silent).

The Xtreme III cooler isnt designed for the 290 series (even though they make it seem so), one memory chip can only fit a 1/2 size heatsink because the GPU cooler overhangs it and there arent enough memory sinks. The 1/2 size memory sink hasnt affected memory overclocks badly though, I still got to 1620MHz with +100mV on the core (quad pumped to 6480MHz).
There are a lot of heatsinks in the package, you get to decide which you want to put where.
I used L shaped heatsinks on the VRMs.
Paired up some 1/2 size heatsinks and used a double size heatsink for the remaining memory chips.

All sinks glued on with provided glue.
Click to expand...
Load core/VRM temps? Core overclock?
 
Gumbi10 said:
Load core/VRM temps? Core overclock?
Click to expand...
Heh, I'm having a bit of dejavu, already posted, Afterburner graphs are on previous page too.
I'll make it concise.

Stock volts, stock 290x speed (1000MHz core), Max GPU and VRM temps are around 55C (this measurement was taken weeks ago so may have changed a little).
Stock volts, 1050MHz core, Max GPU temp is 59C, max VRM is 64C, measured today.
+100mV, 1155MHz core, Max GPU temp is 64C, max VRM is 72C, measured today.
All these are with room temp around 18C.

It will clock to 1070MHz core on stock volts, I dont have the temps for this.
 
You want best vrm? - reference.

You want cool card and no noise - custom.

You want best of both wordls? only 2 picks for now. TF4/DCII. Sapphire cards have crap vrm's they had on all cards for like 2 years already. Even DCII this generation is not so great regarding vrm's, Matrix and TF4 maybe, only good picks i guess...
 
I have a GB 290, not flash-able with an AC Extreme III as well, Load temp maxes out at about 66-67 and VRM (one one of the ones measured) seems to to to about 82-84 max
 
Digital Viper-X- said:
Load temp maxes out at about 66-67 and VRM (one one of the ones measured) seems to to to about 82-84 max
Click to expand...

You are on the edge, of the recommended temps. Thats why referece design is best imho.

With fans at 55-60% you won't have throttling, maintain ~85c, and yet have vrms around 60-70c max. Its a bit loud, but not big issue imho, if you are not using open bench.
 
psyside said:
You are on the edge, of the recommended temps. Thats why referece design is best imho.

With fans at 55-60% you won't have throttling, maintain ~85c, and yet have vrms around 60-70c max. Its a bit loud, but not big issue imho, if you are not using open bench.
Click to expand...

Nenu said:
Are you running your fans at permanent 12V?
Click to expand...

Yes, that is mining actually, gaming it doesn't get that high, Closed case with a cooling fan blowing air from the bottom right to the card, I almost wonder if it is not seated properly since the other VRM sits around 72deg.

I'm going to look @ chopping up the factory heatsink to get the base plate off, use it the same way I Did with the 4890 :p
 
My VRMs show vastly different temps under high load.
During my over volt testing earlier today when one hit 72C, the other was almost 20C lower!

At idle now one is 25C, the other at 27C.

Note that my earlier tests were running Unigine Heaven 4.0 benchmark.
If you also run that, we will have a basis for comparison.

ps I'm using GPU-z to get the VRM temps.
 
Nenu said:
My VRMs show vastly different temps under high load.
During my over volt testing earlier today when one hit 72C, the other was almost 20C lower!

At idle now one is 25C, the other at 27C.

Note that my earlier tests were running Unigine Heaven 4.0 benchmark.
If you also run that, we will have a basis for comparison.
Click to expand...

I'll run it tonight.

What do you use to monitor VRM temps, I'm using GPU-Z.
 
I will try it tonight :)
Thanks! if not I will go and cut up the baseplate from the original cooler to get a nice big VRM heatsink out of it.
 
Its normal to have big temp difference, relax.

And always try to stay under 65c if you want max stability/efficiency of the vrm's.
 
Can you state the voltage used as reported in MSI Afterburners graph as well?
Its possible yours is applying higher voltage than mine.
As temps increase, it increases voltage further so you can get an early runaway effect.
 
Nenu said:
Can you state the voltage used as reported in MSI Afterburners graph as well?
Its possible yours is applying higher voltage than mine.
As temps increase, it increases voltage further so you can get an early runaway effect.
Click to expand...

My Voltage is roughly 1.109 pretty stable and once in a while it will blip to 1.165. Not very often though.
 
Assuming those are under load, mine is reported as higher than yours, at stock 290x clocks mine blips up to 1.19V
What is your ambient temp btw?
 
There is something a bit baffling.
Mining with scripts is more memory heavy than GPU heavy, which might explain your slightly lower max voltage even though we are both recording 100% GPU use.
But your GPU and VRM temps are way higher despite theoretically less GPU use, much more so than the ambient difference.

Did you glue your VRM sinks on?
Are you 100% sure your fan is running on 12V? It can be fed 7V or 12V direct from the PSU with the provided connector.
 
Nenu said:
There is something a bit baffling.
Mining with scripts is more memory heavy than GPU heavy, which might explain your slightly lower max voltage even though we are both recording 100% GPU use.
But your GPU and VRM temps are way higher despite theoretically less GPU use, much more so than the ambient difference.

Did you glue your VRM sinks on?
Are you 100% sure your fan is running on 12V? It can be fed 7V or 12V direct from the PSU with the provided connector.
Click to expand...

I'm pretty sure I am, :) What are your idle temps? IT could be the case, I have mine in a bitfenix ghost, Something tells me that if I remove my intake fan filters, my temp would drop, or If was to crank all the intake fans to full power.
 
My idle with only a single monitor selected is 33C.
If I duplicate to another display or use a 2nd display in any way, it goes up near 38C because ULPS gets disabled.

My PC case is an Antec 900. All fans running at lowest speed setting.
fyi
 
Nenu said:
My idle with only a single monitor selected is 33C.
If I duplicate to another display or use a 2nd display in any way, it goes up near 38C because ULPS gets disabled.

My PC case is an Antec 900. All fans running at lowest speed setting.
fyi
Click to expand...

Interesting indeed! my idle temp is usually around 40-41, Here is a mining shot.

Also yes, I did glue my VRM heatsinks on, should I not have?

 
You must log in or register to reply here.
Back
Top