No surprise here, XFX 290 DD VRM temps over 100c at stock volts

Didn't know my graphics card could double as a car hahahhahahahahah.

Anyway, people have been whining about VRM temps ever since they've included sensors that read it.

Guess what, the drivers are reading the temps too, and they'll shut down / crash if it reaches the critical temp..
 
Guess what, the drivers are reading the temps too, and they'll shut down / crash if it reaches the critical temp..
Drivers can shut down the card due tue high temperature, it's a fail-safe function.
Drivers CAN'T crash the card, it's a failure.
@Relayer: ICs running at higher temperature will have less efficiency (read: more power leakage) than ICs running at lower temperature. Therefore, your statement is completely incorrect ;). To produce the same 300W, VRM running at 70 deg. C will need 350W whereas VRM running at 100 deg. C will need 400W thus increase heat dissipation.
 
Comprehension...

Prolong driving at high gear with low RPM will put load on the engine and shorten its life. Since people don't want to kill their transmission or engine running at either extremes they prefer a mid point and with VRM temp it exists as shown. Plus, it could impact the stable operation and/or life of nearby components with lower temperature rating.

Lugging the engine (Running at high gear with RPMs under 500) only hurts the engine because of oil starvation. Your first idiotic post said save up to replace the tranny. You'll need to replace the engine, not the tranny. God fucking Christ you're dense.

So say 125c is Red line (We'll say 7k for this example), running the engine at 5600 rpms is bad for it by your thinking. Seeing as 100c is 80% of 125c, then 80% of 7k is 5600. That engine would run forever at that RPM (With proper maint), as will these VRMs (With proper maint.)
 
Drivers can shut down the card due tue high temperature, it's a fail-safe function.
Drivers CAN'T crash the card, it's a failure.
@Relayer: ICs running at higher temperature will have less efficiency (read: more power leakage) than ICs running at lower temperature. Therefore, your statement is completely incorrect ;). To produce the same 300W, VRM running at 70 deg. C will need 350W whereas VRM running at 100 deg. C will need 400W thus increase heat dissipation.

Not necessarily true, some ICs perform more efficiently when heated to a certain point.
 
So say 125c is Red line (We'll say 7k for this example), running the engine at 5600 rpms is bad for it by your thinking. Seeing as 100c is 80% of 125c, then 80% of 7k is 5600. That engine would run forever at that RPM (With proper maint), as will these VRMs (With proper maint.)
Engine wear/life span ARE relevant to rpm. If an engine with specs maxed out at 7000 rpm then running it continuously at 5.6k rpm will reduce it's lifespan comparing to running it continuously at 2k rpm given the engine temperature/lubrication are the same.
 
Not when approaching it's maximum rated temperature. Especially MOSFETs.

Eh, the on resistance goes up with temp, but the over all efficiency doesn't vary much on temp. It varies more on load. Like I said, some are more effect at a certain temp, some are less.

Engine wear/life span ARE relevant to rpm. If an engine with specs maxed out at 7000 rpm then running it continuously at 5.6k rpm will reduce it's lifespan comparing to running it continuously at 2k rpm given the engine temperature/lubrication are the same.

Can we get off the car analogies? They're fucking fruitless here. I was just pointing out to mi7chy that he'd grenade the engine at red line long before he'd ever have issues with the tranny. Of course it would fail sooner, friction will eventually win. I'd be willing to bet the engine going 5.6k rpms would have similar total revolutions to the engine that went 2k rpms its entire life.
 
Wow, did my thread go to hell fast or what.

Stupid car analogies aside, I don't want a part that runs over 100c on any form of PCB to begin with, let alone on a part mining all day.
 
Drivers can shut down the card due tue high temperature, it's a fail-safe function.
Drivers CAN'T crash the card, it's a failure.
@Relayer: ICs running at higher temperature will have less efficiency (read: more power leakage) than ICs running at lower temperature. Therefore, your statement is completely incorrect ;). To produce the same 300W, VRM running at 70 deg. C will need 350W whereas VRM running at 100 deg. C will need 400W thus increase heat dissipation.

He mentioned the components surrounding the VRM being heated by them, which is what I was addressing. There are obviously benefits of keeping components cool.
 
So for shits and giggles I bought a cheap APEVIA X-SNIPER2 and tossed another Apevia 3 pack of fans in, as well as one on the side panel. This was the result of a cheap laugh, quite effective. VRM's down to 91c and core dropped 7c, I should also probably bump my OC to 4.5GHz+. lol
http://www.newegg.com/Product/Product.aspx?Item=N82E16811144286

http://www.newegg.com/Product/Product.aspx?Item=N82E16835228044
apevia_zps83be904d.jpg~original


Screw Lian-Li :D

Having the 120mm side panel fan helped drop the VRM's quite a bit, but it does sit about an inch from the cooler/pipes.
 
Last edited:
People still buy XFX? :D

C7uL90v.png


After 10+ hours mining, fan @54% 290 Tri X :)
 
Last edited:
People still buy XFX? :D

C7uL90v.png


After 10+ hours mining, fan @54% 290 Tri X :)

Yes, because I purposely wanted to buy a card from the same manufacture that sold me a 7950 that died in 3 days. :rolleyes: :D

Having said that though, my new setup works, and I'm probably going to keep the card. ;)
 
People still buy XFX? :D

C7uL90v.png


After 10+ hours mining, fan @54% 290 Tri X :)

No doubt the Tri-X is an awesome cooler. Sapphire hit a home run with that one. I'd be curious to see the DD at the same settings. Might have to adjust fan speeds because fan speed % does not always coincide with RPM.
 
Actually i get even lower temps, this was on hot day lolz. My core is around 64 believe it or not, and vrm is around 70c with fans on 55%

Tri x is not so good only because its very cool, but also because is dead quiet and has guaranteed Hinyx memory with thermal pads on them, and very high quality iR mosfets, outstanding card.

Please do test with same rpm/settings, but i think your pc will crash, DD cards (vrm) run at 100c with fans at 100% afaik :(
 
Do you plan to slap a block on it? If so check this out. I heard that people that have the EK got lower VRM temp by replacing the thermal pads with Fujipoly.
 
No doubt the Tri-X is an awesome cooler. Sapphire hit a home run with that one. I'd be curious to see the DD at the same settings. Might have to adjust fan speeds because fan speed % does not always coincide with RPM.

Actually i get even lower temps, this was on hot day lolz. My core is around 64 believe it or not, and vrm is around 70c with fans on 55%

Tri x is not so good only because its very cool, but also because is dead quiet and has guaranteed Hinyx memory with thermal pads on them, and very high quality iR mosfets, outstanding card.

Please do test with same rpm/settings, but i think your pc will crash, DD cards (vrm) run at 100c with fans at 100% afaik :(
If I have some time tonight, I'll get on it, and see what it does.
The 290 tri-x is a reference board.
I was referring to XFX, I wanted a Tri-X but wasn't going to pay for it heh.
Do you plan to slap a block on it? If so check this out. I heard that people that have the EK got lower VRM temp by replacing the thermal pads with Fujipoly.

Wasn't planning on it, although who knows what will eventually happen with this card. ;)
 
Ah yes, XFX's patented water boiling feature. I had an XFX card once and it was awful, glad I didn't pick up an XFX 290 when I could have, though I was tempted! :D
 
Had to RMA my first XFX 290 (reference) due to super annoying coil whine and got a new R9 290 Double Disspiation Black Edition in return, retail packaging and everything. I was pretty pleased with XFX, even though their website is totally sketch, what with the poor English and community college web design capstone support page.

Anyway, I've been very pleased with the cooling the card provides, as it maxes out around 80 deg., though VRM1 does max at 90 deg. It's quiet, though, and it runs faster at 980MHz than my other card at 1050MHz. Go figure. Anyway, can confirm that VRM1 gets hot, dunno really the implications, am pleased with the card, it's super nice and fast, etc etc.
 
Hey guys, We have heard this complaint as well and we are currently investigating it to find out the cause. There were reports that the thermal tape nearest the DVI ports had dried up, but replacing it resolved the issue.

We will be reporting the results as soon as we find out more.

Thank you

edit- Grammar
 
Last edited:
Hey guys, We have heard this complaint as well and our currently investigating it to find out the cause. There were reports that the thermal tape nearest the DVI ports had dried up, but replacing it resolved the issue.

We will be reporting the results as soon as we find out more.

Thank you

Thanks for dropping in.

I'm glad to see you guys are actually investigating the situation, if indeed the VRM's are capable of running cooler than they currently are with this design, that would be awesome!
 
Hey guys, We have heard this complaint as well and we are currently investigating it to find out the cause. There were reports that the thermal tape nearest the DVI ports had dried up, but replacing it resolved the issue.

We will be reporting the results as soon as we find out more.

Thank you

edit- Grammar

XFX - any updates on this or ETA? i'm planning a new build and these would look mighty fine...:D please advise!
 
I just got 2 xfx 290s and they seem to run way hot. Cranking up the fans to 100 percent doesnt seem to do anything. I cant even tell that they are going faster so not sure whats up with that. I even have a giant box fan blowing over them and the top card will still run up to 94/95 and stay there. I expected that with the custom dual fan cooler that it would do better than this. Wondering if I should have bought the Trix.
 
I have 2 XFX R9 290's in Crossfire with the double dissipation heatsinks.

My top card hits 75C on the core, VRM 1 hits 82C and VRM 2 hits 64C. Bottom card hits 62C on the core, VRM 1 71C and VRM 2 55C. Unigine Heaven 1080P 8XAA Extreme setting for 1 hour.

This is with the fans at 100%, cards are not overclocked.
 
That must be incredibly noisy.

Nope, you can barely hear these cards at 100%. Core only gets to around 75 when gaming, mining gets a fair bit hotter though. Of course, now I have to see how hot my vrm's are after coming across this thread.
 
Nope, you can barely hear these cards at 100%. Core only gets to around 75 when gaming, mining gets a fair bit hotter though. Of course, now I have to see how hot my vrm's are after coming across this thread.

VRM temps would be good :). And how is your case cooling setup?
 
I've never met a fan over 2000 RPM that I wouldn't classify as "Noisy as hell".
3300 RPM sends shivers down my spine. No ty.
 
I've never met a fan over 2000 RPM that I wouldn't classify as "Noisy as hell".
3300 RPM sends shivers down my spine. No ty.

The fans on the new DD aren't noisy at all, at 100% its quieter than my old reference 660Ti. Its definietly under 30DB since I can hear case fans over it heh.
 
My bros GTX 660 Ti blower is noisy even at low RPM.

I've never met a fan over 2000 RPM that I wouldn't classify as "Noisy as hell".
3300 RPM sends shivers down my spine. No ty.


Yep. Especially since it has two larger fans. Even windforce cooler at 3K RPM was loud as heck, this can not be quieter at 3300 RPM. This MSI would sound like a hairdryer at similar speed. If people think that it is not loud I wonder if they have shitloads of >1500 RPM case fans... No matter what fan, 3K is not acceptable. I have open headphones so I want my rig to be as quiet as possible.

Luckily this twinfrozer does rather good job and when overclocked to +38mV 1,1GHz GPU and 1430MHz RAM the max GPU temp is around 80c-83c and RPM 2100-2200 in 100% really stressful load. In lighter load where it does not have to blast full speed temp can be much less and fan speed less than 2K. VRM1 temp is always less than GPU temp and VRM2 is not an issue.
 
Rev 2.0 290 Double D VRMs seem pretty warm to me.

10min Valley Extreme HD with a mild overclock on the core/mem and stock voltage/power limit. Ambient a little warm at 78F but nothing crazy.

 
Rev 2.0 290 Double D VRMs seem pretty warm to me.

10min Valley Extreme HD with a mild overclock on the core/mem and stock voltage/power limit. Ambient a little warm at 78F but nothing crazy.


This is not an unusual temp reading for VRM for cards that are running ~250W (or more). While not ideal, and will likely contribute to limited O/C potential, it's not a sign of early failures or anything.

FWIW Guru3D has started using thermal imaging because they caught some card company purposely falsely calibrating their sensors. They won't say who, though.


index.php

Asus 780ti DCII
 
My Ver 4.0 290x DD hits 101+ no problem any time the load stays at 100% for more than 2 minutes. I opened a ticket with XFX support today.
 
Was going to purchase one of these this week, came here and saw this thread... now I'm undecided :( Any recent datapoints?
 
Was going to purchase one of these this week, came here and saw this thread... now I'm undecided :( Any recent datapoints?

Purchased XFX 290 Black DD 3 weeks ago from Amazon. VRMs reach about 80C while playing BF4, according to GPU-Z. Not sure what version it is, can check when I get home.
 
Back
Top