Accelero X2 vs V1 Ultra vs VF900-Cu for silencing X1900XTX?

krille · Post by **krille** » Wed Mar 29, 2006 1:59 pm

1. What would be the better option for running an X1900XTX as quiet as possible?

a) Arctic-Cooling Accelero X2, claimed to be 0.4 Sone (how many dBA?)
b) Thermalright V1 Ultra with a Nexus 80mm, Acoustifan 80mm or otherwise silent fan to be determined and perhaps a fan controller?
c) Zalman VF900-Cu, claimed to be 18.5 ~ 25.0dB ± 10% (how many dBA?)

2. Which would turn out the most silent solution for cooling a X1900XTX (without experiencing thermal issues)?
(Cost doesn't matter, only noise and it should be as low as possible.)

3. Or would a 7900GTX be a much better choice for low noise?
I'm really itching to try an ATI card.

4. If the Thermalright V1 Ultra, any suggestions on the best (silent) fan? Should I get a fan controller?

In any case, the cooling solution MUST be a lot quieter than the Arctic Cooling NV5 Silencer Rev.3!
It's sooo noisy I just can't stand being in the same room as my current 7800GTX.

Thank you very much!

~ Kris

paapaa · Post by **paapaa** » Wed Mar 29, 2006 2:24 pm

krille wrote:3. Or would a 7900GTX be a much better choice for low noise?
I'm really itching to try an ATI card.

See this review:

http://www.vr-zone.com/index.php?i=3335&s=8

Geforce 7900GTX 89 watts
Radeon X1900XTX 135 watts

IMO that is a HUGE difference (45 watts). It would be a lot easier to keep Nvidia quiet.

krille · Post by **krille** » Wed Mar 29, 2006 2:39 pm

Things is I'm concerned with image quality. Hmm... maybe I should go 7900 GTX afterall. I really want as silent as possible or else I can't sit by my PC. I can only sit by my PC currently for very short sessions (an hour or two maximum...usually much less) and then not even every day. This is all due to noise (and my NV5 Rev 3 is noisiest component).

Any case, what would the best solution be for optimal silence?

~ Kris

NoDamage · Post by **NoDamage** » Wed Mar 29, 2006 7:29 pm

I think the Sytrin VF1 needs to be added to your list of contenders. The problem with the reviews is that many of them seem to contradict each other. though.

http://www.sudhian.com/showdocs.cfm?aid=743&pid=2929 - Shows the VF900 outperforming the Accelero X2 and the VF700.

http://www.ocia.net/reviews/vf900/page5.shtml - Shows the VF900 performing about equal with the Sytrin VF1.

http://www.ocia.net/reviews/sytrinvf1/page5.shtml - But this review shows the Sytrin VF1 outperforming the Thermalright V-1.

http://www.legionhardware.com/document.php?id=525 - And this one only shows the VF900 only marginally outperforming the VF700 while being beaten by the Accelero X1 and V-1 Ultra.

http://www.virtual-hideout.net/reviews/ ... dex4.shtml - Shows the VF900 outperforming the Accelero X1.

Granted these reviews are done on different cards for the most part, but there doesn't seem to be much consistency between them, which makes the decision harder.

If you are looking for silence, then going by the OCIA results it looks like the VF900 on low matches the Sytrin VF1 while outperforming the V-1 Ultra. And the VF900 fan should be marginally quieter than the VF1 fan so that is probably the best bet.

halcyon · Post by **halcyon** » Wed Mar 29, 2006 9:16 pm

Please don't cross-compare the results across various methodologies and thermal loads. It will end in silly results.

The only way to find out how they perform, is to see them repeat-tested by a skillful tester under a similar load as they are going to be used with (7900 or 1900 series).

Most test with nVidia 6600 or 6800. As an example, 6800GT draws half of that of 1900XTX.

Some coolers that might still perform fairly well with a 50-60W thermal load can absolutely barf under a 130W load.

Thomas · Post by **Thomas** » Wed Mar 29, 2006 11:16 pm

halcyon wrote:Please don't cross-compare the results across various methodologies and thermal loads. It will end in silly results.

The only way to find out how they perform, is to see them repeat-tested by a skillful tester under a similar load as they are going to be used with (7900 or 1900 series).

Most test with nVidia 6600 or 6800. As an example, 6800GT draws half of that of 1900XTX.

Some coolers that might still perform fairly well with a 50-60W thermal load can absolutely barf under a 130W load.

Hmmm. I tend to disagree somewhat with this. No doubt, that there's a lot of factors, which will affect how well a certain cooler perform. But when a X1 is better than VF900 in one test, and in the other it's the opposite. Then I tend to say something is wrong.

krille · Post by **krille** » Thu Mar 30, 2006 3:35 am

Well of all the coolers that will provide adequate cooling performance, which one is the quietest?

I'm not sure the Sytrin VF1 is available in Sweden, but I've never heard of it before so I can't really say I know for sure.

~ Kris

halcyon · Post by **halcyon** » Thu Mar 30, 2006 10:13 pm

Thomas wrote:But when a X1 is better than VF900 in one test, and in the other it's the opposite. Then I tend to say something is wrong.

Oh, I completely agree. That's down to (imho) to non-repeat tests and the lack of skill on the testers part.

At least to me that's the most likely explanation.

Which is why I made a point of not comparing results from one review to another (methodology, tester, etc are all different).

And I do believe that V1 Ultra with any sane kind of choice of fans (2) will outperform VF900 with 100+ W thermal loads.

But that's conjecture. We'll see what the deal is, once somebody competent enough really tests them both in repeated manner (using same methodology) and with a high enough thermal load.

jmke · Post by **jmke** » Fri Mar 31, 2006 1:57 am

how much heat does 7800GT @460Mhz GPU speed produce? I'm using it for GPU comparo (which includes zalman, AC and some other solutions)

krille · Post by **krille** » Wed Apr 05, 2006 2:27 pm

What would be cooled more silently?

7900GTX SLI or X1900XTX?

If I'm going 7900GTX, I'm probably going SLI. I won't be going X1900XT Cross-Fire though.

halcyon · Post by **halcyon** » Wed Apr 05, 2006 10:12 pm

Very good question. 2 x GTX will produce more heat than single XTX, but.... it can also boil down to:

- availability of heatsinks for either card type
- space on your motherboard / in your case
- airflow in your case
- overclocked or not

I'd assume that XTX with a proper cooling could be less noisy, but it's all down to pragmatics.

If I were you, I'd pick my choice based on other criteria, as neither of those solutions is going to be particularly silent :)

krille · Post by **krille** » Mon Apr 10, 2006 5:37 am

What is quieter at same performance (as silent as possible, without GPU overheating).

VF900-Cu or Accelero X2 or V-1 Ultra?

Copper · Post by **Copper** » Mon Apr 10, 2006 6:21 am

krille wrote:What is quieter at same performance (as silent as possible, without GPU overheating).

VF900-Cu or Accelero X2 or V-1 Ultra?

I don't have an X1900XTX to try it on, but I think you're pushing your luck trying to cool it quietly with a V1Ultra. I have one on my X850XT PE. The GPU gets up to ~70C with a Nexus 80mm at 7volts where it's quiet. If the numbers above are correct, you'd be almost doubling the watts. 850 = ~70 watts, 1900 = ~135 watts. You can add a second fan to the V1Ultra, but it will take up no less than two extra slots if you do. I suspect using two fans would be the only chance of keeping the 1900 cool with the V1Ultra. But then again, they'd be two quiet fans. Best of luck. And please let us know how whatever you decide to try works out.

krille · Post by **krille** » Mon Apr 10, 2006 7:13 am

Copper wrote:I don't have an X1900XTX to try it on, but I think you're pushing your luck trying to cool it quietly with a V1Ultra. I have one on my X850XT PE. The GPU gets up to ~70C with a Nexus 80mm at 7volts where it's quiet. If the numbers above are correct, you'd be almost doubling the watts. 850 = ~70 watts, 1900 = ~135 watts. You can add a second fan to the V1Ultra, but it will take up no less than two extra slots if you do. I suspect using two fans would be the only chance of keeping the 1900 cool with the V1Ultra. But then again, they'd be two quiet fans. Best of luck. And please let us know how whatever you decide to try works out.

Oh! That was information I really needed! Thanks man. In that case I'm definitely not going V-1 Ultra, rather the X2 or the VF900-Cu (still not set). Will give you an update whenever I decide though.

On VF900-Cu vs Accelero X2 only thing I can make out is:

Zalman about VF900-Cu wrote:Bearing Type : 2-Ball
- Speed : 1,350 ~ 2,400rpm ± 10%
- Noise Level : 18.5 ~ 25.0dB ± 10%

http://www.zalman.co.kr/eng/product/vie ... 2&code=013

Arctic Cooling about Accelero X2 wrote:Extremely Quiet
A large fan and optimized fanblades offer a minimum noise level at a high air flow.

The Fluid Dynamic Bearing keeps oil inside the bearing and thus reduce friction inbetween and keep the noise level to a minimum.

http://www.arctic-cooling.com/vga2.php?idx=90

According to this site 0.4 sones equals ~23.5-24.5 dB(A). But this should be at 100% I assume. At lower, I have no clue.

On a side note, this X1900XTX is to replace a 7800GTX with the NV Silencer 5 (Rev. 3) measuring in at 0.9 Sone by Arctic Cooling. 0.9 Sone would be ~29.5dB(A) so it should at least be more silent than what I have currently. And the noise of the NV5 is really pissing me off.

I don't think you can convert dB to dB(A) though, because you would need to know the intensity of all frequencies. Which we don't.

Anyway, if these specs are true and directly comparable (which they probably aren't, since they were probably measured very differently at different distances) then I reckon they'd be about the same at full. At anything lower, we don't know for sure.

jmke · Post by **jmke** » Mon Apr 10, 2006 7:17 am

Sone is a subjective noise measurement, making it not so easy to rely on; but dBA ain't perfect either so we're still left in the dark in the end.

for what it's worth, the fan on the X1 (for nvidia) is really quiet imho compared to the stock cooling.

krille · Post by **krille** » Mon Apr 10, 2006 7:37 am

True... so which is quieter IRL under same conditions and measuring precedure?

jmke · Post by **jmke** » Mon Apr 10, 2006 10:03 am

I can only speak for X1 vs Z900 - both at max speed, installed inside a case with system ambient noise (without the VGA fan) at 37.8dBA at ~10cm

the dBA meter reads:

40.1dBA for Z900 and 38.1dBA for X1.

krille · Post by **krille** » Mon Apr 10, 2006 10:44 am

Interesting, since you can use the Accelero X2 in conjunction with ATI's dynamic fan controller, I think it may be the better option here. Thank you ever so much!

jmke · Post by **jmke** » Mon Apr 10, 2006 11:16 am

do note that my readings were taken with the dBA next to the case, the side panel closed -- I had to put my ear next to the case in order to make out the fan speed difference on the X1 between 100% and 25%. With the Z900 I could more clearly hear the difference in fan speeds (12v vs 5v)

krille · Post by **krille** » Mon Apr 10, 2006 1:44 pm

jmke wrote:do note that my readings were taken with the dBA next to the case, the side panel closed -- I had to put my ear next to the case in order to make out the fan speed difference on the X1 between 100% and 25%. With the Z900 I could more clearly hear the difference in fan speeds (12v vs 5v)

Do you mean as in the X1 was already quiet, or do you mean the X1 was very noisy even at 25%? Which one was quieter... X1 @ 25% or Z900 @ 5v? What about 7v?

jmke · Post by **jmke** » Mon Apr 10, 2006 1:51 pm

X1 very quiet

krille · Post by **krille** » Mon Apr 10, 2006 2:16 pm

jmke wrote:X1 very quiet

Excellent! I'm convinced. I'll be ordering the X2 along with my X1900XTX. Again, thanks very very very much for all valuable information and help!

~ Kris

thetoad30 · Post by **thetoad30** » Mon Apr 10, 2006 3:39 pm

Be forewarned: I think Arctic Cooling's service and quality is suspect.

I had a fan (ATI5 R2) that was clicking really bad (worse than normal) on my ATI card. So I did an RMA. Took a while for them to even respond, so I went back to my original retailer like they recommend. They almost didn't replace my fan since I was outside their 30 day policy.

Then, the new fan somehow popped off its shaft and stopped working one day. I didn't notice until I wanted to see what the card was set at for fan speed since the noise went away. How surprised was I to see the chip reading 96C!!! I thought, for sure ATITool has to be bugging out. NOPE. Opened the case, and the fan was at a dead stop.

That's the last Arctic Cooling product I use. Great ideas, but not great quality. I would recommend the V1-Ultra. I almost bought that, but then I decided to just go water.

NoDamage · Post by **NoDamage** » Mon Apr 10, 2006 9:57 pm

I have compiled a list of user reports of the X1900XTX coolers over at AnandTech:

http://forums.anandtech.com/messageview ... erthread=y

The general trend seems to be that the Accelero X2 is a better cooler than the VF900 for this card. The Zalman at 12V does perform better than the stock cooler and the Accelero when they are set to auto fan control. However, when they are set to 100%, both the stock cooler and Accelero outperform the VF900. The Accelero at 100% also tends to equal or outperform the stock cooler at 100%, so it seems like the best choice for this card. The reports of the VF900 causing crashes or overheating when paired with a X1900XTX are also fairly discouraging.

In terms of noise, I think the Accelero has a huge advantage over the VF900, V-1 Ultra, and Sytrin VF1, because it is the only cooler that is powered by the video card. It is supposed to be fairly quiet at 100%, but you can have ATI Tool dynamically adjust the fan speeds between 2D and 3D mode and throttle down the fan to be completely silent when you are not using it.

The auto fan control issue also helps explain the conflicting results in the earlier reviews. The performance of these coolers is highly dependent on the fan speed, and the default fan header voltage may vary from card to card. For instance, on the X1900XTX it is usually under 50% most of the time, only cranking up to 100% when the temperature surpasses something like 100C, which virtually never happens. With a cooler like the Accelero connected to it, it may not perform too well since the fan is turned way down by default. The VF900 on the other hand goes directly to the MB/PSU, so its fan speed is generally consistent across all reviews. So on a card with default low fan voltages the VF900 will tend to perform better, while on a card with higher voltages the Accelero will fare better.

halcyon · Post by **halcyon** » Tue Apr 11, 2006 12:28 am

Very nice roundup damage.

This will help us all until somebody (SPCR) does a definitive temp/acoustics testing on these.

thegrommit · Post by **thegrommit** » Tue Apr 11, 2006 8:09 pm

None of the above? link. The review isn't particularly good, but the cooling system looks intriguing.

jmke · Post by **jmke** » Tue Apr 11, 2006 11:10 pm

thegrommit wrote:None of the above? link. The review isn't particularly good, but the cooling system looks intriguing.

intriguing in what way? that it cools but as good as the stock cooling and generates only 2dBA less noise?
http://www.anandtech.com/video/showdoc.aspx?i=2737&p=4

thegrommit · Post by **thegrommit** » Wed Apr 12, 2006 1:02 am

jmke wrote: intriguing in what way? that it cools but as good as the stock cooling and generates only 2dBA less noise?
http://www.anandtech.com/video/showdoc.aspx?i=2737&p=4

while running at slightly higher speed. Granted, we're only talking about 14 MHz while overclocked. It's unfortunate that AT didn't bother to take noise measurements at stock speeds.

[edit] never mind, the fan is not temperature controlled, so the "low" number presumably represents operation at stock speed.

krille · Post by **krille** » Wed Apr 12, 2006 9:28 am

Thank you ever so much NoDamage this was exactly what I was looking for!

I have now ordered a Sapphire X1900XTX + Arctic Cooling Accelero X2 for $753.64 + $29.28 + shipping = $782.92 + shipping.

I had a look at Sapphire's Blizzard version, but it wouldn't come in stock till June and judging by the reviews it's much louder and noisier than the Accelero X2, so that ruled the Blizzard out.

Thanks again to everyone contributing to this thread! Big thanks!

jmke · Post by **jmke** » Wed Apr 12, 2006 9:46 am

Accelero X2 vs V1 Ultra here: http://www.legionhardware.com/document.php?id=537

Accelero X2 vs V1 Ultra vs VF900-Cu for silencing X1900XTX?

Which will allow the most silent cooling for X1900XTX?

Accelero X2 vs V1 Ultra vs VF900-Cu for silencing X1900XTX?

Re: Accelero X2 vs V1 Ultra vs VF900-Cu for silencing X1900X