Reviews - Sapphire Radeon HD 5870 1GB Vapor-X graphics card: the best in the world?

[Mathonar]

It seems to me that Hexus should possibly be querying this result with Bill Donelly then because Sapphires own Press Release states :

"SAPPHIRE’s World leading Vapor-X technology not only allows the card to run as much as 15 degrees C cooler** and 10dB quieter** than the standard model, it provides additional headroom for performance tuning (overclocking). "

Weren't the 48xx Vapor-X cards queiter than reference?

I wonder if the change in shroud from that used in the 48xx has resulted in a change in the noise level.

I'm a bit disappointed, i really wnated something that was going to be quiter than my pair of Powercolor 3870 PCS currently are. Maybe the 5850 version will be different...

[Phage]

I have the 4890 version, and it is very quiet indeed. I have never heard it.

[Zhaoman]

Originally Posted by Mathonar

It seems to me that Hexus should possibly be querying this result with Bill Donelly then because Sapphires own Press Release states :

"SAPPHIRE’s World leading Vapor-X technology not only allows the card to run as much as 15 degrees C cooler** and 10dB quieter** than the standard model, it provides additional headroom for performance tuning (overclocking). "

Weren't the 48xx Vapor-X cards queiter than reference?

Yeah Vapor-X cards have been both cooler and much quieter than reference from what I remember so this is a very surprising and slightly disappointing result. Maybe it's because the 4 series had such a bad cooler to start with. This is definitely not in line with previous Vapor-X products on the noise front anyway.

[JARSmith]

We thank the HEXUS.community members for their insightful comments.

HEXUS takes the evaluation of products very seriously. After extensively testing the the Vapor-X card from Sapphire Tech we note in our editorial that it’s slightly noisier under load than the reference card. Compounding matters, the Vapor-X’s fan has a different, more-annoying pitch.

We appreciate that this sounds quite unlikely, especially when the press release makes claims about the fan being much quieter – a fact that has historically been true of the Vapor-X cards.

Responding to your comments, HEXUS.community members, we setup both cards side by side on one of our high-end graphics platforms and listened again. Further, we also took some screengrabs showing the temperatures under load together with the fan speeds. You can find these at the following URLs:

http://img.hexus.net/v2/graphics_car...ref_load_1.png
http://img.hexus.net/v2/graphics_car...r-x_load_1.png

Corroborating our earlier findings, the result was the same, with the Vapor-X noticeably, but not hugely, noisier. We’re adamant that the noise delta between the cards is greater than the small fan-speed difference would indicate.

HEXUS has since contacted Bill Donnelly, global marketing director for Sapphire Tech, and he's providing us with a second Radeon HD 5870 Vapor-X sample. We will keep you posted and update the review accordingly, if needs be.

[Perfectionist]

Did you pass them a link to this thread? If I were in their boots I'd be pretty concerned!

[GheeTsar]

Wow. That's dedicated tech journalism. I'm impressed. I knew there was a reason I like Hexus

[Iota]

Would be interesting to compare the reference card overclocked to the same speeds, as well as fan speed and see where the noise sits at that point. The Vapor-X fan is running at 53% compared to the reference card at 30% in those screen caps (and i'd be interested to see where temps sit at that point)

Perhaps the extra noise is being generated more by the perforated cover Sapphire have gone with this time around?

[cobhc]

Yeah, I'd be interested to see how the reference 5870 performs cooling wise with a similar fan profile to the Vapor-X card. Seems like the Vapor-X version's fan is running faster, which of course would result in the card also running cooler. Maybe run a test with both cards locked to 60%?

[scaryjim]

Originally Posted by GheeTsar

Wow. That's dedicated tech journalism. I'm impressed. I knew there was a reason I like Hexus

It's one of the things I've becoming increasingly aware of, and delighted by, on Hexus. I used it as my go-to site for reviews for about 4 years before I joined the forums, mostly because the review setups and results were very clearly laid out and I found it much easier to compare between different reviews on Hexus. Since I've joined the forums I've been very impressed by just how often the staff get involved with the (often enthusiastic) members - I've had some interesting discussions with both Tarinder and Scott in the forums, and the current Batman AA thread indicates how seriously many people in the industry take Hexus.

It makes a lot of sense when you're running a review and tech news site with an attached forum though, as (I imagine) it provides Hexus with a lot of traction when looking for exclusive interviews etc. - they can just point to the forums, and their interaction with us, as being a direct line from, say, an AMD exec to their customers: something that's harder to get elsewhere. We benefit, Hexus benefits, and the tech companies benefit: good stuff all round

[Kotoa]

The AMD Radeon HD 5870 is, arguably, the best consumer graphics card that money can buy right now

But the GTX295 got higher FPS in most tests, and had the most marks?

[Phage]

Originally Posted by Kotoa

But the GTX295 got higher FPS in most tests, and had the most marks?

Hence arguably. 5870 is a single GPU, uses less power and is DX11 compatible.

[Tarinder]

Originally Posted by Kotoa

But the GTX295 got higher FPS in most tests, and had the most marks?

We think that 'best' means more than a benchmark lead, as Phage has pointed out.

[Tarinder]

Hello, peeps.

We received a second Sapphire HD 5870 Vapor-X sample yesterday and, appreciating the comments above, re-ran the noise/fan-speed tests.

A few HEXUS.community members requested that we observe the fan-speed and noise profiles when both the Vapor-X and reference card's fans were set to the same speed, and the cards at the same frequencies.

Cranking the reference card up to the Vapor-X's 870MHz/5,000MHz clocks and attempting to match fan speeds:


http://img.hexus.net/v2/graphics_car..._fixed_rpm.png - Sapphire HD 5870 VaporX idle fan-speed (HEXUS-fixed)
http://img.hexus.net/v2/graphics_car..._fixed_rpm.png - Reference card idle fan-speed (HEXUS-fixed)

http://img.hexus.net/v2/graphics_car..._fixed_rpm.png - Sapphire HD 5870 VaporX load fan-speed (HEXUS-fixed)
http://img.hexus.net/v2/graphics_car..._fixed_rpm.png - Reference card load fan-speed (HEXUS-fixed)

The speeds aren't identical, down to the lack of granularity for ATI's Overdrive panel, and the fans spin at different percentages to achieve the same rpm, but do provide a better apples-to-apples comparison.

Running them this way highlighted that the second sample's fan exhibited a noise profile that was about the same as the reference fan's - but certainly no quieter, making Sapphire's 10dbA-quieter claim seem rather optimistic.

Sapphire's engineers are now taking another look at the cooling and will report back to us.

[Phage]

Dammit. Really wanted one of those...
Actually, now that you have two samples are they for sale ? Cheeky I know, but don't ask, don't get.

[Tobeman]

Tech journalism at its finest. Keep up the pressure on Sapphire!

[Mathonar]

Thanks for going the extra mile Tarinder and also to Sapphire for enabling Hexus to do this.

Looking at the figures it's always a case of matching the stock and Vapor-X fan speeds so this really shows that the stock fan and vapor-x fan have a similar noise level when running at the same RPM.
However the vapor-x is keeping the GPU a lot cooler under load so i would have been interested to see how far you can drop the Vapor-X fan speed before the temp rises to that of the stock cooler. If you can get it a bit under 2000rpm does that make it much quieter?