It's like AMD has a very good answer for INTEL and VIA, if you look at the power consumption numbers:

http://www.tomshardware.com/reviews/Atom-Athlon-Efficient,1997-5.html

ATOM eats energy almost exactly as on TR review so it looks the numbers are comparable, and the performance of the Athlon 65 2000+ is better then the ATOM 230, may be AMD Athlon 65 2000+ == VIA Nano L2100

As always, very good review, when can we see AMD numbers included?

You should not thank for this OCZ suply. It skewed the power consumption tests. Next time get an 90W pico PSU or something like that for those low power tests. ;)

Simply going from an 550W to an 220W PSU shaves 20W from the total system power consumption. More here: http://www.xtremesystems.org/forums/showpost.php?p=3176757&postcoun...

It would be good to underclock your Pentium M to 1.8GHz, to give an better perspective of Nano architeture efficiency per clock. The use of an low cost celeron would be even better.

Besides that, great review as always! Talked about all atom and nano line-up, links to architeture, and remembered that this is the "ugly end of the clock frequency/voltage curve" for Nano.

I'm impressed with Nano's results. Atom really had no chance most of the time. The power efficiency, even though not as good as the Atom, is put into perspective with how fast the chip is, performing the tasks and overall ends up being as power efficient as the Atom. definitely good stuff and competition is spiked in this area aswell, which is always a good thing for us.

Yeah, what really matters is price. If a Nano + Motherboard is under $100, I'll bite for sure, maybe even a bit more because the motherboards are going to be better featured. I've used an 800MHz VIA C3 in the past, and for everyday tasks at the time it was fine (but I never used it for video, etc) but that wasn't using Windows.

How fast is the Padlock encryption? You could test with OpenSSL I think, as that supports it.

if the Atom platform cost $69 why is the EEE $700?
why 45nm Atom is still behind a ~5 years old Pentium M ?

In my book both processor fail to deliver , on power consumption, performance.In the year 2008 , in the age of 45nm transistors, in era of multi core , hiperthreading , atom is a shame.
Even Intel hates it!

And now the wait for the Nano powered HP 2133 starts.

Hey TR,

Thanks for this really informative review. If a Nano-based office desktop comes out, I could easily recommend it over an Atom-based desktop if it costs less (desktops aren't greatly concerned with power consumption, anyways).

69 bucks for a board + processor is pretty damn cheap, tho.

I think it's fairly safe to say that what this review really shows us is how much butt the atom will kick in the proper device. something where the CPU load is the most significant power draw.. :)

I currently have a laptop with VIA C7 1.5 w/ 2 gigs ram. And i have to say that it really isn't that good. I can't even surf the web and watch a movie at the same time, or download at top speed from comcast connection.

To see the results of this chip are kinda disappointing, it really reminds me of a celeron(which IMO is a slow chip and i wouldn't buy).

Not to mention i would NEVER buy first gen of anything, especially since they reworked this chip from the ground up.

I look forward to Nano 2

Nice work to VIA on updating their CPU line, Good luck on optimizing/polishing this gem.

Scott,

Something caught my eye in the first page of the article:

Windows installer is looking for a CPUID string that says either "GenuineIntel" or "AuthenticAMD". When it runs into "CentaurHauls", it keels over and dies

This CPUID seems to be a bit of a problem for some companies. Instead of checking for feature compatibility (I'm really surprised to hear that the Windows installer isn't doing that) they're just looking for a string.

Ars Technica had some very interesting findings regarding this CPUID and the PC Mark benchmark app:

http://arstechnica.com/reviews/hardware/atom-nano-review.ars/6

Via doesn't lock the CPUID on their processors and so it can be changed. I'm wondering if this same thing could be done for installing a 64-bit Windows? I ask mostly out of curiosity, as it seems like this is just showing us that there's some really sloppy programming out there.

Anyways, re: the rest of the stuff. I'm far more interested in Nano than Atom, especially when Via's new chipset is made available. I certainly hope they get more design wins and we see more widespread availability in the US.

Looks like the 8W 1.3GHz Nano will be roughly performance-equivalent to the 1.6GHz 4W Atom. It would be interesting to see a review of that chip coupled to a VX800 chipset (single chip, 5W TDP, HD video acceleration for H264,VC1,DivX, etc) on a mini-ITX motherboard.

The Intel chipset is what is letting down the Atom, especially for mobile low-power purposes like the EeePC. Intel need to get their act together here. However I don't think Intel really want to do too much with the Atom in this field, they want it to be used in MIDs with the feature-bereft Paulsbo chipset.

This article makes me love the 2.26GHz Pentium M in my HTPC even more. :)

... VIA... did good..?

/puts on ice skates.

I am an unabashed VIA fanboy, as well as a low-power enthusiast. When I saw the other reviews hit the web earlier today, I was really hoping I'd see something on tech-report. I must say, you have not disappointed me with this review.

Considering that I use a P-M in my main desktop rig (SD11G5 from Shuttle), its addition to the comparison was SO INCREDIBLY AWESOME. TR, by far, rocks socks.

also, please do look at the performance with a PCI-E video card. When I checked out the same thing a few years ago (http://www.techreport.com/forums/viewtopic.php?f=22&t=40381), there was light-years difference in performance with a simple PCI video card. I can't imagine what a recent-gen PCI-E video card (passively cooled, of course) would be like.

In addition to the Nano and Atom processors, we've included a Pentium M in our comparison for the sake of reference. Right now, as they read these words, marketing types from both Intel and Via are hyperventilating into paper bags, as are small contingents of pedantic fanboys from each side. Our thinking was this: we needed a familiar x86 processor as a performance and power consumption reference, and a Pentium M at 2GHz seemed like the best candidate we had on hand. We tested the Pentium M 760 on a desktop-class motherboard, just as we did the Atom and Nano, though the board is a little older.

Bravo for having the guts to do so (although given that the Atom and Nano are more contemporary, something like the D201GLY2A[1] motherboard/Celeron 220 [Yonah, i.e. 2006's Core Duo generation] CPU combination, which sells for about as much as the Atom board, might have been more appropriate -- in fact, while Intel mostly marketed it as microATX, in actuality it conforms to mini-ITX).

[1] http://www.provantage.com/intel-boxd201gly2a~7ITEM0F9.htm

With that said, competence is more important in tasks like these than raw performance. I conducted my own casual, seat-of-the-pants evaluation by surfing around in IE7 on both of these systems, and neither one of them felt slow to me. Both were snappy in pulling up and rendering web pages, and neither one was appreciably slower than the high-end quad-core desktop system I use every day. Both are adequate, even if the Nano is measurably faster.

Bravo for this observation as well. Common sense like this is lacking in innumerable hardware reviews, when it really ought to be more prevalent.

Intriguingly, the Pentium M 760's peak power consumption turns out to be lower than the Nano L2100's. FATALITY. How does it fare against a Celeron 220, which is still widely available? I suspect, given how the Atom fares, a better comparison would be against the C7 rather than the Nano, since the C7's power consumption is comparable to Atom's (rather than being several times higher as with the Nano).

Nitpick: The Nano's L2 latency is four cycles, and its L2 latency is about 24. The former's probably the L1 latency.

Will there be a mobile version of Isaiah different from the current Nano? Because if not, please consider comparing it against an Atom N270 (currently the only netbook-oriented Atom that's in all those 8.9" and 10" netbooks) with the 945GSE chipset (with its 6 W TDP compared to the 945GC's 22 W) as well.

I really want Intel to validate something like a G31 successor (G41[2]?) for Atom use, or failing this, OEMs to have the guts to put them together themselves. Atom uses the same bus protocol as today's Core 2 processors; why not a chipset that is two full process nodes smaller (and therefore that much more efficient)?

[2] http://www.fudzilla.com/index.php?option=com_content&task=view&id=8...

[H]ardocp did a recent review as well and came to the same conclusion: the Nano outperforms the current Atom noticably. They didn't talk about pricing and availability as much though. Apparent;y Intel was right when they said 'You don't want an Atom for a desktop' or at least the current single core Atom with 945G. The proper chipset should make it more interesting.

Having said all that the main attraction of these solutions remains the min-ITX form factor. The Intel board cost is rediculously low but one could also look at desktop dual cores like the Celeron C2Ds and an IGP mobo as a budget solution that has some grunt and better all around capabilities.

The power draw on the Atom is the really impressive bit.

Cost is going to win this one for Intel; Via's EPIA line has always been expensive and it doesn't look like the Nano is going to help. Sigh.

Makes one wonder what a Pentium M could do if moved to 45nm from its 90nm process. Nano's power use isn't that amazing compared to its performance.