• 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.

4850 - Stock vs. Accelero vs. DuOrb vs. vf900

Soldato
Joined
17 Dec 2006
Posts
8,190
25iu9ep.png


These are my own tests.

I've owned a VF900 for a while, and my friend owns a Accelero S1 so I used both those two coolers in this test. We didn't have a good fan to hand for the Accelero so I had to zip tie a 120mm TriCool Antec fan with it's switch set to "M" for the tests. The DuOrb I received this morning and added it into the tests as well. I'd like to point out however the Accelero S1 was attached to his card in the tests (he also owns a 4850) however we tested all the cards in my room in the same environments. If you want to discount the Accelero results because of that please do (I wasn't really inclined to take his cooler of his card and attach it to mine when they are the same cards).

The VF900 and VF1000's fan speed's was set to midway on a Zalman Fan Mate 2. This would make it about 1850 RPM.

Case cooling: Antec P180, not top exhaust, rear exhaust @ low, intake from middle HDD bay @ M, lower PSU cage fan @ M. CPU cooler is Zalman CNPS9700-LED attached to a Fan Mate set to medium.

Ambient room temperature at the time of testing was: 27c - 29c. We couldn't keep the temperature in the room stable due to lack of air conditioning so take that into account on the results. Something else to note also is that the temperatures will probably go down a couple of degrees across the board when the paste had set but it never got time to do that during testing as it normally takes weeks.

You might wonder why my results with the Accelero S1 are so much higher than that chart floating around with the 32c idle, 39c load. Well that's probably because he used RTHDRIBL which doesn't really stress the card to levels that ATI Tool seems to do, kinda seems like when you are testing your CPU and you set the priority from 1 to 8. When I ran RTHDRIBL with the DuOrb the load temps never went above 44c, compared to 57c with ATi Tool.

Hope this helps anyone at least, was just attempting to show the difference between the coolers since I was lucky enough to have all of them to hand :)

Links to the coolers:

Thermaltake DuOrb - £30.54
Zalman VF900 - £23.49
Arctic Cooling Accelero S1 Rev. 2 - £15.26 ** Fanless **
Zalman VF1000-LED - £31.71
 
Last edited:
Soldato
Joined
2 Sep 2006
Posts
13,483
Location
Portland, OR
Very nice job...i can see your graph being referred back to on many occasions in the future.

As expected, clear winner is the accelero s1 as you say based on price/perofrmance.
 
Associate
Joined
3 Sep 2003
Posts
238
Location
hampshire
Ah just what i was looking for, im getting a new system with a hd4850 and was trying to decide what aftermarket cooler to use :)

Bit off topic (sorry)
Ive never fitted a aftermarket cooler are they easy to install? also if i was to buy an accelero s1 would i need to buy extra ram/vrm heatsinks?
 
Last edited:
Soldato
OP
Joined
17 Dec 2006
Posts
8,190
As expected, clear winner is the accelero s1 as you say based on price/perofrmance.
Yep, it might even match the DuOrb cooling or suceed it with a better fan option. The TriCool is not the greatest fan in the world, and it was only set to medium. I should have tested it at high setting as well but it slipped my mind as we were in a hurry to get through the tests. Better 120mm fan, or maybe 2x 92mm fans if they will go on and I reckon it'll supass or match the DuOrb. But if people already have a DuOrb I wouldn't bother buying a Accelero.
 
Soldato
OP
Joined
17 Dec 2006
Posts
8,190
Bit off topic (sorry)
Ive never fitted a aftermarket cooler are they easy to install?
The DuOrb and vf900 are easy to fit, my friend stated the Accelero was easy to fit. I would like to point out that they come with 2 sets of screws for different heights of the card from the PCB. If you are using a DuOrb make sure you use the 11mm screws and not the 13mm screws. Also one more note on the DuOrb is that it'll only go in one direction. Technically you should be able to have it in 2 directions, but if you use the alternative the copper hits the container housing the S-Video port and it won't go on.

All in all they were all extremely easy to fit, the only annoying thing is getting the ram sinks to stick properly, had a few occasions where I turned the card over and the RAM sinks fell off and I had to reattach and press down to get some cohesion. I'd imagine after a blast in the machine and when the card is loaded the compound will melt a bit and they'll stick a lot better.

sorry for double post, accidently created a new post when attempting to edit my original to answer the above questions
 
Associate
Joined
24 Oct 2005
Posts
2,047
Location
Lincolnshire
I've got the Accelero S1 on my 4850 at the moment, and would agree with the results posted above.

The only problem for me with the S1 is that I plan to get another 4850 and crossfire them, in which case the cooling fins on it will stop me connecting the crossfire bridges.

I can't see any way round it unless you can buy extra long flexible bridges?

I think I'm going to try a Zalman VF1000 also which I've got stuck in the cupboard.
 
Associate
Joined
3 Sep 2003
Posts
238
Location
hampshire
I've got the Accelero S1 on my 4850 at the moment, and would agree with the results posted above.

The only problem for me with the S1 is that I plan to get another 4850 and crossfire them, in which case the cooling fins on it will stop me connecting the crossfire bridges.

I can't see any way round it unless you can buy extra long flexible bridges?

I think I'm going to try a Zalman VF1000 also which I've got stuck in the cupboard.

good point :( i plan on getting another one for crossfire at some point too
 
Associate
Joined
1 May 2007
Posts
1,901
Location
London
When i put the DuOrb on my 3870 I left the reference RanSinks on as they were a lot better. The DuOrb isn't the quietest cooler in the world, but compared to the reference one, it's a step in the right direction.

And for anyone interested, it has the 'fancy' blue led's on it.....
 
Associate
Joined
31 May 2008
Posts
1,049
Location
York
I've got the Accelero S1 on my 4850 at the moment, and would agree with the results posted above.

The only problem for me with the S1 is that I plan to get another 4850 and crossfire them, in which case the cooling fins on it will stop me connecting the crossfire bridges.

I can't see any way round it unless you can buy extra long flexible bridges?

I think I'm going to try a Zalman VF1000 also which I've got stuck in the cupboard.

Very good point, i probs get vf900 for both and there quiet coolers and hlaf decent and i dont plan to overclock the 4850's that much.

~Slash
 
Caporegime
Joined
6 Dec 2005
Posts
37,553
Location
Birmingham
Nice comparison fibby. :)



My Akasa Voertex Neo results. :cool:




Idletemp-1.jpg


Maxtemp.jpg






Should have a Zalman VF1000 early next week to do a review on so watch out for that folks as that is apparently a better cooler than this Akasa one.
 
Last edited:
Soldato
OP
Joined
17 Dec 2006
Posts
8,190
For the sake of completeness the VF900 fan speed should be mentioned in the first post.
Middle ground on a Zalman Fan Mate 2. Not sure what that speed would be, I didn't record it, but "medium" sounds about right :p

[EDIT]: Checked around, that would make it about 1850 RPM.
 
Last edited:
Back
Top Bottom