We can see from the majority of our testing that Arctic have designed a solid product in the Accelero Twin Turbo 6990 VGA cooler. The product is very easy to mount and the overall design of it means it will fit in your case fine if you can already fit a HD 6990 in – if you can’t fit a HD 6990 in then you have no reason to buy this! Add to that the fact the cooler also looks great – the black and white aesthetic plays it safe but the black and white colour combination is one of the most popular these days along with red and black and a few others.
The Arctic Accelero Twin Turbo 6990 VGA Cooler was more than capable of taming GPU core temperatures, we actually saw up to a huge 20 degree decrease on the GPU temperatures under Unigine Heaven Gaming Load. Yet we still have a slightly bitter taste left after the Arctic cooler showed its inefficiency at keeping the VRM cool when compared to the stock design. Although, this was “redeemed” to an extent by the fact that we never saw the VRM temperatures get close to dangerous levels – even under Furmark it only reached 80 degrees.
The main design win that the Arctic Accelero Twin Turbo 6990 VGA Cooler has, is with its vastly superior acoustic performance. The stock HD 6990 cooler really does put your ears through torture, The HD 6990’s lowest noise output was 50 dBA at idle. While the Arctic cooler never got louder than 45 dBA under any usage scenario (including Furmark where the stock cooler hit 74 dBA). Hugely impressive stuff from the Arctic cooler.
For me the noise reduction offered by this product alone is enough to make me want to go out and buy it. The added cooling performance and clean aesthetics are both just added bonuses. Costing £76 ($96) this is a costly VGA cooling solution but considering there are no other options apart from water cooling your HD 6990, which could cost anywhere from £200 and upwards, £76 is actually a very small price to pay. You should also consider you are getting two 120mm fans, 10 heat pipes and 86 aluminium heat fins, so for £76 you actually get a lot of spec for your money.
In our opinion Arctic have created the best solution to the hot and noisy HD 6990, that does not require the excesses, risk, hassle and expense of a water cooling set up. In that sense this is one of the most innovative VGA coolers we have ever seen. Weaker-than-stock VRM cooling and the fact that this cooler doesn’t support its own weight very well stopped this cooler from snatching the Editor’s Choice award. Though we are still very impressed with this offering from Arctic and we look forward to similar product from them if AMD ever gets around the releasing the HD 7990.
lol just shut up xD.
Thanks for the detailed test. Though it looks a bit strange that with the stock cooler the voltage regulators get about 10% hotter under Furmark and with the Twin Turbo supposingly the temperature increases more than a 100%. Guys look into that again, I would bet on a simple typo.
not a typo. Tried remounting it too, no difference :/ I can assure you its legit
My VRs are despite a little overclocking never higher than 70°C. If you worry, then just push the fan from automatic control which keeps it really low to about 50% PWM. It is still virtually silent.
Claude, no need to get nervous because of this test. I got the following results and I’m a using the card every day a few hours for many months with the Twin Turbo already.
Accelero Twin Turbo 6990, Furmark: GPU 1: 55°C, GPU 2: 52°C, Vreg: 73°C
Stock Cooler, Furmark: GPU 1: 81°C, GPU 2: 77°C, Vreg: 72°C
I have no idea what was measured in this review.
Well, I’m using the AMD 6990 since a long time already. Of course the first thing I had to do is to get rid of this extremely loud reference cooler – and I can recommend it to everybody.
I don’t know how this test was conducted and at which fan speeds. It is not logic that the VR extrusion heatsink looses all the sudden massiv efficiency under Furmark which draws about 30% more power and thus the temperature should go about 30% up. But eteknix wants to tell us that under Furmark the temperature rises more than 100% – funny. No offence but something went wrong on your side.
We are currently looking into the issue. From the numbers i see below in the comments (70-85) I can see that our temperatures from Furmark are about 20 degrees higher than they should be.I will go for one last remount and refit of the card and then have another look. If issues persist I’ll use a different diagnostic program. The review will be updated accordingly.
Thanks Ryan. No offense here. Just share my own opinion. And I am wonderting why there is such a big difference between these two reviews?
We had some issues with our diagnostic program, clearly wasn’t functioning normally with the dual GPU card and all the associated temperature sensors that come with that, I think it was listing the wrong readouts as VRM but its all been retested, double checked and we think everything is okay now.
Yes I have to agree, the results look quite unbelievable. Have bought my card after I have seen the test on overclockersclub.
Those overclocking results were achieved at only 29% fan speed!
I did not bother to overclock on my card and just enjoy the silence. There is no VR cooling issue, at least not on my card.
http://forums.overclockersclub.com/index.php?showtopic=192144
Here are the results:
Overclocked speed – 940 MHz Core, 1400 Ram
Stock cooler:
Idle temp: N/A
Load temp: N/A
Fan speed at load: N/A
*Note the card ran so hot at idle, overclocking it was impossible*
Arctic Cooling Accelero Twin Turbo 6990 Cooler:
Idle temp: GPU1 – 46 , GPU2 – 32 , VREG – 52
Load temp: GPU1 – 65 , GPU2- 58 , VREG – 82
Fan speed at load: 29%
I don’t know why everybody keeps talking about the voltage regulators and Ryan for sure something went wrong on your side. I got the cooler 4 months ago and I’m super happy with it. Because of all the fuss I’m reading and this nice but little bit blurry review, I took the time and quickly ran a Furmark myself:
Vreg temp: 75°C
So no reason to worry since these chips are made for 120°C operation temp.
I will continue happily gaming in silence.
Thanks for your sharing, Jeff. That seem reasonable. I do not know why you will get these totally different testing result. But I still believe that this cooler is designer for overclocking, it is impossible to perform worse than the stock cooler. Anyway, thanks both of you.