-
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.
Installing an Arctic Accelero Xtreme IV to Sapphire 290
- Thread starter Beerbaron
- Start date
I have no idea, I wanted the III but could only find IV. I thought it was a new release. I actually received an email from Arctic saying the new version isn't even released yet
Points I would like to add from the install:
Removing the ref cooler from a Sapphire 290 is pretty easy once the screws have been removed from the metal frame. You dont need to remove the coolers plastic housing first. The thermal tape and adhesive still hold it on so slowly twist the cooler off. There are loads of vids online to watch first if needed. The PCB will move/bend a little. I honestly found this the worse part. Once removed clean the GPU with some alcohol. I used Vodka but some may want to use something more pure, isopropyl or ethyl alcohol.
The Alpenföhn Thermal Adhesive is easy to spread, I used a penknife and some paper towel. It is not as messy as some people may think. A nice even thin layer is fine. Don't over coat the coolers as it may leak over the edge of the chip. I believe its non conduct so won't cause damage just a mess (correct me if I'm wrong). I haven't used the Akasa AK-TT12-80 Thermal Adhesive Tape people have recommended so can't compare. I easily removed a Alpenföhn cooler from a chip after 2-3 hrs curing. When it cures it has a rubbery type consistency. I slowly twisted the cooler off, cleaned the chip and cooler and reset it with no issues. The Alpenföhn Thermal Adhesive does cure somewhat in the tube after use. I stored mine lid tightened in a sealed bag but it does start to cure. I noticed this with mine about 6 hrs after I opened it. I was able to use it after I removed the part cured adhesive at the front of the tube.
Don't feel you have to test the card as soon as you have assembled the cooler. The thermal adhesive needs time to cure. I would say 2-3 hrs minimum to be sure but over night if you can. Although the coolers may be stuck the adhesive may not have cured fully in the middle. I found this when removing a cooler. I was very eager to make sure my card still had life but had to resist.
I used GPUz to test my VRM temps.
The Acrtic Accelero Xtreme IV is definitely wide. I have an MSI P67 GD65 motherboard which this fits on a single card. The motherboard is a good few years old. I did try to have 2 Sapphire 290's in crossfire but there is only an inch gap between the cards.This is not sufficient for crossfire and this cooler. I opted to send one card back, sorry OcUK, as the heat and noise of a ref card is too much. In hindsight I could have bought a non-ref card or even a new motherboard.
The protective sheet of plastic that sits in between the rear PCB cooler and the card I found was pretty useless. By the time I had cut out all the areas I had stuck the blue thermal putty there wasn't much of it left. Its held in place by a few plastic clips that pin the large heatsink and the heatsink/fan around the card.
Associate
- Joined
- 10 May 2012
- Posts
- 1,748
Give her a go in furmark.
Monitor temps on vram during 15min test,we see how high they'll go.
My first water block (xspc razor) hit over 100 degrees after few minutes.
Monitor temps on vram during 15min test,we see how high they'll go.
My first water block (xspc razor) hit over 100 degrees after few minutes.
Hello everybody, just read the 3 pages on this discussion, but I have a question: has anyone some data with ONLY the Accelero Xtreme IV as it comes out of the box? I mean, from what I understood, people here who own an Accelero IV also bought some VRAM and VRM heatsinks..
Since a GPU cooler is supposed to be tested by the company, I also suppose that temperatures only with the backplate which comes from the box would be at least sufficient to provide good temperatures.. Instead there would be absolutely no reason to buy a product which can cause big damages for a 400€ card. And it would be also stupid for a company to sell a product like that...
Since a GPU cooler is supposed to be tested by the company, I also suppose that temperatures only with the backplate which comes from the box would be at least sufficient to provide good temperatures.. Instead there would be absolutely no reason to buy a product which can cause big damages for a 400€ card. And it would be also stupid for a company to sell a product like that...
Caporegime
it should come with its own heatsinks?
No, that would be the problem.
It comes only with the backplate, which assumes that they studied it to make VRM and VRAMs cooling working properly only with the 3 fans. Hoping it would not be worse than the III
Caporegime
yeah I read the first post,it should be ok but for good measure some extra vrm/memory heatsinks wouldnt hurt
But I would be quite disappointed. I mean, this is known as a great cooler for high-end VGAs, also allowing OC'ing... If someone (like me) wants a completely working cooler out of the box and instead it needs additional heatsinks it wouldn't be the right choice for me. Also because here in Italy it's priced at 70€ which is not very cheap..
Hello everybody, just mounted the Xtreme IV on my 290X without any heatsink (only with pieces provided by Arctic), run many Valley benchs consequently for about 20 minutes at Extreme HD preset. Temps are great. I don't know temps from AX III, but here with fan at 100% (which is basically not udible, especially with game's sound/music) I got after 15/20 minutes 55°C on GPU core, 60°C VRM1 and 50°C VRM2.
It's clear that in this case fans are the most important part in cooling VRMs, in fact lowering them from 100% to 50% GPU core is at about 60°C while VRM1 goes to about 70°C, VRM2 stays about the same. Lowering fans to 20% GPU core goes up to 70-75°C, VRM1 to 82°C and VRM2 at about 60°C.
So, after 20 minutes Valley bench preset Extreme HD:
Fans 20%:
GPU, VRM1, VRM2: 73°C, 82°C, 60°C
Fans 50%:
GPU, VRM1, VRM2: 60°C, 70°C, 58°C
Fans 100%:
GPU, VRM1, VRM2: 55°C, 60°C, 50°C
So, if you want a cold card using an Acc Xtreme IV just put your fans to 100% and you'll be great even without heatsinks on VRAMs and VRMs Or just use AB to create a personalized fan curve where if temp goes to 55°C fans will go up to 100% and keep core and VRMs at great temps (for those who do not like the almost non udible sound of fans spinning at 100%)
Hoping this would be helpful for someone
It's clear that in this case fans are the most important part in cooling VRMs, in fact lowering them from 100% to 50% GPU core is at about 60°C while VRM1 goes to about 70°C, VRM2 stays about the same. Lowering fans to 20% GPU core goes up to 70-75°C, VRM1 to 82°C and VRM2 at about 60°C.
So, after 20 minutes Valley bench preset Extreme HD:
Fans 20%:
GPU, VRM1, VRM2: 73°C, 82°C, 60°C
Fans 50%:
GPU, VRM1, VRM2: 60°C, 70°C, 58°C
Fans 100%:
GPU, VRM1, VRM2: 55°C, 60°C, 50°C
So, if you want a cold card using an Acc Xtreme IV just put your fans to 100% and you'll be great even without heatsinks on VRAMs and VRMs
Or just use AB to create a personalized fan curve where if temp goes to 55°C fans will go up to 100% and keep core and VRMs at great temps (for those who do not like the almost non udible sound of fans spinning at 100%)Hoping this would be helpful for someone
Caporegime
Yes i remember how effective the accelero was. Nice piece of kit and so so quiet.
Soldato
- Joined
- 7 Aug 2012
- Posts
- 2,647
VRM temps were the thing that put me off my 7970 Accelero Xtreme. 90 degrees+ a lot of the time and occasionally passing the 100 degrees mark. I know they are rated to 125 degrees, but I didn't like seeing temps as high as that and wondered if it possibly removed any additional headroom I had for overclocking, since the Xtreme kept the core so cool.
That is interesting that VRM temps were better without the heatsinks. Probably wont use it again unless I build a second rig and stick 1 of my 7970's in it once I upgrade my main rig, but will defo keep it in mind if I do.
That is interesting that VRM temps were better without the heatsinks. Probably wont use it again unless I build a second rig and stick 1 of my 7970's in it once I upgrade my main rig, but will defo keep it in mind if I do.
Soldato
- Joined
- 28 Oct 2011
- Posts
- 8,553
Why not just buy a non ref 290?
Soldato
- Joined
- 7 Aug 2012
- Posts
- 2,647
Most will, but those that bought reference - and remember reference cards were the only 1's available for a couple months - may feel that the reference cooler is either not good enough to prevent thermal limits being reached and throttling or far too loud, or both and thus require an after-market cooler.
Arctic happen to be probably the best manufacturer of after-market gpu air coolers.
That's the point. I bought a reference card (I was buying a GTX770, then for some other problems I tried a R9 280X, then the news of the 290(X), so I waited a couple of weeks and bought a reference card).
It's clear that now everyone buys custom cards, but for those who bought a reference one I think the AX IV is the best solution for now.
Soldato
- Joined
- 28 Oct 2011
- Posts
- 8,553
Soldato
It doesn't cover all of the core due to the offset of the 7970 chip and remember the 79's have a raised sink, you can use a brass shim, EK do one for a couple of quid.
If you want to try it out, I imagine it's fine short term(a week later when I wasn't lazy I added the shim when I bottled it
as I was waiting for a gpu bracket from Hong Kong for my 5870 Xtreme) as long as you don't oc it.Because they are still too loud and can't compete on core thermals as well as the reference boards are better than almost every non ref boards-the best cores generally are the first batch of cores, lastly, AMD's vrm design can take a severe beating on the voltage.
£320 290 that unlocked to a 290X.
