Asus A8R32-MVP Deluxe w/ATI CrossFire Xpress 3200 Chipset

Low power for the extreme high end.

The "Discuss this article in the SPCR Forums" link at the bottom of the article points to the review of an AOpen Pentium M mATX board.



I believe that should read: "The ASUS A8R32-MVP Deluxe was compared against an ASUS A8N32-SLI Deluxe and a DFI LanParty NF4 Ultra D."

Those are the only major mistakes I have to point out.

Regarding the board itself, it seems a pity not to have negative offsets for Vcore. AFAIK only DFI allows this adjustment for use with CnQ. Thanks to Devon for a nicely done review, though.

Is it me or this image (http://www.silentpcreview.com/ninja-a8r32mpv.jpg) does not load?

Doesn't work for me either.

All fixed, but it will be a little while before the article gets republished.

Good to see this type of comparison.
I’m curious as to how the extra power consumption of the A8N32-SLI affects the CPU or system temps!

It works now

Nice to see this board reviewed, i've been looking to go s939 with a x2 3800+ for a while, don't have the money right now so it will be a while until i can buy the complete set. I've been looking at this board and the Abit AT8 32x with the RD580 chipset. Silence wise the AT8 has four PWM controlled fan headers
http://www.abit-usa.com/products/mb/pro ... &model=311
Fast glance tells me the Asus has the 3200, Abit the 580, Asus has 2 Gbit Lan and eSATA, Abit has 1 Gbit LAN no eSATA...
eSATA is a very nice feature, put your HD 2 m away at no speed penalty...
tough discision...

Thanks for the review.

Bought the a8r32 yesterday. Don't need the crossfire (yet), but all the other goodies will be nice.

Can we asume this is 4 phase like the A8N-SLI, not 3 phase like the DFI or 4+4=8 phase like the A8N32-SLI?

Unfortunatly pics of the A8N-SLI SE show no heatsink for the VRM so I'll probably grab a A8N-SLI on the cheap before they disapear. I do wonder about the fan control on the A8N-SLI SE if it had the better BIOS options this board has I'd consider mounting heat sinks myself...

If you don't have the money right now, is it worth waiting for AM2, or do you already have DDR RAM you plan to use?
The 3200 chipset is just the new name for the 580, the difference is purely marketing.

Thanks for the info, 580 = 3200 didn't notice that. I dont have DDR lying around tot use, will have to buy everything new. But the advantages of AM2 are not that great. New socket, new processors, makes the prices go up a lot, for little improvement. Maybe end of this year an s939 system will cost a lot less. The only thing i noticed is that the prices of DDR are getting higher every month, thats a downside.
The only thing i like about AM2 is Pacifica, i'll be working with virtualisation a lot, at my work, so it will be nice to have it at home also...

The advantage of buying AM2 is that you will have the new platform that is capable of supporting next year’s 65nm chips. It seems likely that they will offer significant performance and power efficiency gains over S939, whereas AM2 on 90nm won’t. 65nm on S939 seems unlikely.

Until the motherboards are in-stock and priced up it seems difficult to predict how much the price difference between an AM2 and S939 system will be at the release of AM2. CPU & RAM prices are already known, so it’s down to the motherboards. It may not be so easy to pick up a low-end AM2 motherboard on release, if that’s what you had in mind. They’re bound to come in droves at some point as AM2 supports Semprons also.

ASRock has one ready.
http://www.hardwarezone.com/articles/vi ... =6&id=1906

With this review focussing purely on power consumption and not looking directly at how this will affect system temperatures and cooling needs; and therefore ultimately how it will impact noise levels, I wonder if SPCR is expanding its remit to include more articles based purely on energy conservation?

I ask partly because I’m wondering how many people here are also interested in looking at power efficiency when making purchasing decisions, even when it has negligible impact on noise levels. I’d like to put together a poll to get a response to this question, but I’m not sure how to word it or which forum to place it in. I’d be grateful for some suggestions.

There is a direct correlation between electricity used inside the case and the noise caused by evacuating that heat with fans. In what case do you see power consumption not being relevant to noise?

FWIW, I'm all for knowing power consumption numbers even if they don't affect noise.

I would say that there is a power threshold for any system, below which a reduction in power consumption will have a negligible affect on noise; this is assuming that the system is designed for low noise in the first place. To be more precise, there are individual thresholds for different components and these thresholds will vary depending on the design of the system, the cooling used and the interaction between the different components.

My one quibble with the review is that it didn’t give me an idea on how the power consumption of the different motherboards affected the system temperatures and therefore on whether it impacts system noise.

Using current generation low power CPUs, non-gaming VGA cards and efficient fanless power supplies, the major remaining issue for silent computing is hard drive noise and not power consumption. In this case moving to a 2.5â€

Yes, we are interested in looking at products solely on the basis of efficiency, since it represents an area that is relatively unexplored. As you mentioned, the only real obstacle to creating a quiet PC these days is the HDD, which doesn't leave much room for new research. Quiet PCs have become a mature market where the only real question is "which components are quiet", not "how do I make my existing system quiet". That question has already been answered.

Regarding this review in particular, I would love to be able to explore the effects of power consumption on thermals, but how do you suggest I go about doing this? Even a 15W difference is unlikely to make much more than a degree or two of difference, especially since it is likely to be spread over the whole motherboard. There are real questions of methodology and relevance that would need to be answered before I could answer questions about the direct effect of MB power consumption on noise levels. Suggestions?

Excellent news.

I thought about why the review hadn’t looked at thermals before posting on this thread and decided that you must have considered that power differences of between 17W and 24W weren’t significant because the power sources on a motherboard are not localised in the way that say a CPU is; as you yourself also mentioned.
That seems a reasonable assumption to me, but, I do like the rigor of testing assumptions to see how they actually stand up.
Looking at the A8N32-SLI Deluxe, one of the chipsets and the heat-spreader are close to the CPU socket, so I guess it will have some affect on CPU temps.

One idea for a thermal test would be the following:
Build a system using a small ATX compatible HTPC case and use a low power mATX motherboard. Add the hottest components that you can whilst still being able to cool the system (virtually) silently. Then swap the motherboard for the A8N32-SLI Deluxe and see if the temps vary. If the temps climb enough so that the fans need to be turned up above their acoustic sweet spot then we have an issue.
The hardest part is in choosing the maximum safe temperatures for the system and CPU that will be used for calibrating the baseline power level that the system can cool quietly. It’s a very subjective issue.
If the A8N32-SLI Deluxe doesn’t have a negative impact in this scenario, then it’s unlikely to prove an issue in the real-world, as the above setup isn’t likely to be commonly used.

That could be done, but the real issue would be ensuring that the measured temperature results in each case are comparable. I've seen enough variation from diode to diode, even on boards that are supposedly identical, that I wouldn't trust that sort of thermal comparison. I think it would require an external thermistor to pull off, but even then there's always the unanswerable questions about what postition on each board is a fair comparison.

Thank god there’s at least some things left that aren’t so easy to do, otherwise it might get a bit boring
I think an external thermistor makes sense; for measuring system temps I would suggest using about 3 of them then you have more of a chance of finding hot spots. Deciding on where to locate the 3 sensors is very open to question; I think that comes down to experimentation. Maybe what is ideally needed here is a ridiculously expensive thermal imaging device that uses infra-red or something!

For the CPU temp measurement is the following practical?
Measure the CPU temps from both motherboard sensors whilst they are located in an open test-bed with all other components being low power. The idea is to get baseline CPU readings from both motherboard sensors whilst they are in an environment that helps mask any differences that the motherboards power consumption have on the readings. Hopefully, the idle and load temp readings for each sensor will allow you to calibrate the sensors accurately. If this works, you can then apply the calibration to the in situ CPU readings.

Yup! As you can see from this append which includes a thermal picture of a machine in operation, thermal imaging would make an excellent toy tool for the SPCR review team to use. Just think how much easier it would be for us to plan fan placement if we had access to thermal pictures of suitably similar systems.

The cameras are indeed hideously expensive. But they'd only have to buy one, wouldn't they?

I would be very happy to contribute something towards the BuyDevonAThermalCamera fund.

I assumed they are expensive, but I haven’t checked prices, yet.
Any idea how much an entry level system would cost?

I did look, back six months or so. It was surprisingly hard. They're not consumer items. I did get a reply to an email from some manufacturer or other saying that their entry model was - around $8000 I think. But they were rather cagey as I have a UK email address and their camera was subject to export controls. (!)

I've seen this board on sale for what looks to be a fair price, and I was tempted to buy me one to replace my MSI mainboard. It seems quite a solid overclocking board for my Opteron 165.

Are there any issues with this motherboard that I should be aware of?

One area of concern is support for the ULi southbridge, considering nVidia has bought them.

latest BIOS 0602 has CnQ issues in Vista. Other than that, it has been a perfect board for me. Asus needs to release a new BIOS badly.

I've been perusing Asus' forums and I've been scared out a bit by the sheer number of people complaining about the ULi southbridge, the same issues seem to be shared by many other mainboards using that IC.

While I was pondering, the board went out of stock so I'm now looking for alternatives ( = A8N32-SLI Dlx) Thank you for your reply, anyway.