• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

7970/50 Arctic Accelero Users

MOOGLEYS said:
The VRM's are rated to 125.c so you still are well within limits.

I have no idea what mine are as I cannot read them but tbh I'm not that bothered. If they overheat the card will shut down and I have not had that yet so all must be within tolerances.

I would not worry about the VRM's too much tbh and enjoy a nice silent card.
Click to expand...

I know you are right Moog but i can't help it.

I wonder if the 7970 stock VRM pads would be better? Maybe i could put those on top of the accelero heat sinks? There was a long grey strip from my reference blower that rested over the VRM's.

Uncle Petey said:
Some cards play differently like I said :)
If your VRM's are at 90°C odd and you're not seeing any ill effects, then be happy.
Some cards will throw a complete wobbly with artifacts and glitches galore.
Click to expand...

Its sleeping dogs SSAA what heats them up. Its brutal on the VRM's. So i use that as my safe maximum. Just wish i could keep the vrms down just a tad more than what they are currently.
 
I used the little right angled heat sinks on my VRM's. They are slightly bigger than the smaller straight ones you are meant to use. You should have these in the pack with the cooler. Could be worth a go... ?
 
LtMatt said:
Its sleeping dogs SSAA what heats them up. Its brutal on the VRM's. So i use that as my safe maximum. Just wish i could keep the vrms down just a tad more than what they are currently.
Click to expand...

Yeah, Sleeping Dogs Extreme was my test of choice for VRM worst case.

I know what you mean when you say you can't help it :)
I have a real thing about my CPU not going over 70°C for example (It doesn't). It was the same with the GPU core and VRM's.
The only difference was that I actually saw issues when the VRM's went over 75 or so, but that could've been down to a slightly dodgy card. Ymmv...

MOOGLEYS said:
I used the little right angled heat sinks on my VRM's. They are slightly bigger than the smaller straight ones you are meant to use. You should have these in the pack with the cooler. Could be worth a go... ?
Click to expand...

When I was researching coolers, I did read of a few guys that had a bit of success using those sinks. Worth a shot as they are bigger after all.
 
Uncle Petey said:
Yeah, Sleeping Dogs Extreme was my test of choice for VRM worst case.

I know what you mean when you say you can't help it :)
I have a real thing about my CPU not going over 70°C for example (It doesn't). It was the same with the GPU core and VRM's.
The only difference was that I actually saw issues when the VRM's went over 75 or so, but that could've been down to a slightly dodgy card. Ymmv...



When I was researching coolers, I did read of a few guys that had a bit of success using those sinks. Worth a shot as they are bigger after all.
Click to expand...

I know what you mean. Keep my cpu under 65c. Core stays under 60c now and vrms well under normal games under -70c 75c, but sleeping dogs 80-90c depending on the overclock.

I will reseat when i can and try those other heatsinks. Wish i could do it now but i have no adhesive or anything.

I'll also need some new adhesive rings for the spacers, all 8 of them. Suggestions regarding that?
 
Last edited:
I can only suggest asking Arctic for more adhesive rings and/or glue.
If you're gonna try installing new sinks, I'd try the Alpenfohn ones which come with a tube of pre-mixed thermal glue (It's not permanent. It dries like rubber sealant and just takes a little patience to thoroughly remove). You could also just use thermal tape.
Can't see them dropping more than 5°C or so, but that's pure guesswork :)
 
Uncle Petey said:
I can only suggest asking Arctic for more adhesive rings and/or glue.
If you're gonna try installing new sinks, I'd try the Alpenfohn ones which come with a tube of pre-mixed thermal glue (It's not permanent. It dries like rubber sealant and just takes a little patience to thoroughly remove). You could also just use thermal tape.
Can't see them dropping more than 5°C or so, but that's pure guesswork :)
Click to expand...

I'm quite happy to use the G1 stuff again or akasa tape.

Is this the G1 stuff?

http://www.overclockers.co.uk/showproduct.php?prodid=TH-005-AC

I'll probably try the larger AC heatsinks, on a budget atm.

EDIT

Using vsync has lowered my VRM temps a fair bit. Saw a max of 60c in black ops just now.

If i limit fps to 45 with sleeping dogs id imagine they'd come down a fair bit. I was using vsync in sleeping dogs previously.
 
No idea mate. I'd imagine it isn't, but I guess Arctic or google would be the ones to ask.

Tbh, if you're not seeing any issues and Sleeping dogs is the only one that pushes the temps up, you're probably ok to just leave it as is.
 
Last edited:
Uncle Petey said:
I can only suggest asking Arctic for more adhesive rings and/or glue.
If you're gonna try installing new sinks, I'd try the Alpenfohn ones which come with a tube of pre-mixed thermal glue (It's not permanent. It dries like rubber sealant and just takes a little patience to thoroughly remove). You could also just use thermal tape.
Can't see them dropping more than 5°C or so, but that's pure guesswork :)
Click to expand...

Petey, how would I go about removing the Arctic vrm heatsinks and how does the thermal glue clean off. Really not happy with my vrm temps which are going over 90 and sometimes touching 100. Going to try the Alpenfohn 1's. Great cooler otherwise.

Although strangely I was able to run a constant oc of 1185/ 1625 no problems. Then my ssd containing windows failed. So bought new ssd, clean install of windows (with same gpu drivers) and now I can't get that clock to run stable. Getting artifacts in Heaven now, whereas none before.

I am going to give watercooling a try in the New Year at some point I think. Need a project.
 
Five.Stars said:
Petey, how would I go about removing the Arctic vrm heatsinks and how does the thermal glue clean off. Really not happy with my vrm temps which are going over 90 and sometimes touching 100. Going to try the Alpenfohn 1's. Great cooler otherwise.

Although strangely I was able to run a constant oc of 1185/ 1625 no problems. Then my ssd containing windows failed. So bought new ssd, clean install of windows (with same gpu drivers) and now I can't get that clock to run stable. Getting artifacts in Heaven now, whereas none before.

I am going to give watercooling a try in the New Year at some point I think. Need a project.
Click to expand...

What asic, voltage and gpu and memory clock settings are you using ?
 
I have no practical experience with the Accelero mate :)
My 'knowledge' is based on what I learned with an Alpenfohn Peter and from what I could gather when researching VGA coolers beforehand.

Petey's theory:
The common theme between them is relatively poor VRM/PCB cooling due to (I believe) having a much thicker/denser main heatsink which leads to less air getting through at the pace required to cool pcb components.
Since the main heatsink provides better core cooling, I'd also theorize that the extra heat being dissipated by the main heatsink means that that extra hot air is also being pushed through and onto the board.
Combine that with the thicker sink and you have poor pcb cooling.
This theory isn't backed by Newscientist or anything, but it makes sense in my head ok? :D

I was never any good at cleaning I'm afraid. I hear that the arctic glue peels off fairly easily and you can just use some achohol (Not Jack Dan ;)) and/or Arcticlean or something along those lines. Others can advise you better here :)
 
LtMatt said:
What asic, voltage and gpu and memory clock settings are you using ?
Click to expand...

Asic is 79 (i think). Will need to check that when I am on main pc either this evening or tomorrow.

Voltage for that overclock was 1212mV according to Trixx although monitoring with gpu-z showed that to be a bit lower. The oc was 1185 core and 1625 memory but as I say, this is now not giving a single artifact free run on Heaven. Possibly increasing voltage further would help, but I am actually not even comfortable with vrm temps at that voltage. I run fan at 100% when gaming/ benchmarking at that overclock.

Abother thing I have noticed is benchmarking sleeping dogs & hitman absolution I will often see dips into the 40's, yet gpu load is never over 70%. Also playing Need for Speed The Run which uses Frostbite2, again I am seeing dips into low 50's regularly now whereas before it was buttersmooth at 60 (d3doverrider forced triple buffering which meant the occasional dip to 57 or so which was unnoticeable without fraps running). Again gpu load is maybe only 50-60%.

I can't recall this being the case on previous Windows install although it had been a while since I monitored gpu-z. Cpu is an i7 920 @ 4 GHz so I can't see that being a bottleneck.

One thing i didn't do was change bios to ACHI mode before install and did this afterwards (but I am not sure I even did this with old OCZ SSD either). Can always try a fresh install but bit of a pain reinstalling updates etc. again so soon.
 
Why is everyone so concerned by vrm temps ? Even at 100.c you have 25.c left before you reach the safe Max. These things are DC to DC power converters they are designed to run ruddy hot.

Tbh unless the temps are going wildly out of limits I would leave things as is and play some games.
 
I know i know, but i strive for perfection.

I'm going to rebuild mine as soon as i can find some suitable replacement gear. I'm going to use those slightly bigger AC heatsinks on the vrms i think. Even if it only brings temps down a few c ill consider it worth it.

@Five stars i would try a fresh driver install, do a atimanuninstaller driver sweep first.
If that doesn't work go for a windows reinstall. It could also be your memory or gpu core is too high for the voltage supplied. Try lower one then the other until you find the culprit. If i had to guess i reckon it would be either the gpu or memory is slightly too high.

QUESTION

Would this akasa tape be ok to use on the spacers if i was to rebuild? I have no double sided adhesive and there is no way i can reuse my current ones. I'm amazed they even worked the first time after i butchered them.
 
Your card is running perfect it's well within the tolerable limits. There is nothing to gain from running them a few .c cooler, they will not perform any better or effect the rest of the card. The power delivery system has a max temp it will work at and unless you are hitting temps beyond this and experiencing issues there is nothing to gain from removing the heatsinks you have fitted other than possibly damaging the card in the process.

There is far too much worry about VRM temps as people have this conception that if it's near 100.c it's about to self destruct. These are high temp components, They are rated to run at 125.c constant for the entire life of the card if needs be.

Having a higher core or memory voltage does not always mean a higher VRM temp.
 
Last edited:
MOOGLEYS said:
Your card is running perfect it's well within the tolerable limits. There is nothing to gain from running them a few .c cooler, they will not perform any better or effect the rest of the card. The power delivery system has a max temp it will work at and unless you are hitting temps beyond this and experiencing issues there is nothing to gain from removing the heatsinks you have fitted other than possibly damaging the card in the process.
Click to expand...

I don't think ill damage it by reseating the heatsinks, the G1 stuff comes off pretty easily it seems.

I'll just use the rubber/alcohol on the vrms to clean them up then use some thermal tape to re-apply. ;)
 
You must log in or register to reply here.
Back
Top Bottom