VIA Nano and Intel's Atom benchmarked head-to-head

Netbooks based on VIA's Nano mobile processor aren't nearly as common as those based on Intel's Atom, but based on the benchmarking that's been going on recently, that's a shame, since the Nano appears to be much faster than the Atom 230. PC Perspective, Eee PC News, and Hot Hardware all ran some tests recently, and a 1.8GHz Nano L2100 with Chrome9 graphics was usually able to outperform a 1.6GHz Atom 230 with GMA950 graphics at everything from MP3 ripping to 3D benchmarking. Of course, that's not without a tradeoff -- the Nano was a bit more power-hungry, and the Atom's memory and graphics systems were occasionally faster than the Nano's. Still, it seems like the Nano has more raw horsepower than the Atom -- and it's pin-compatible with VIA's popular C7M, so hopefully we'll be seeing machines like HP's Mini-Note make the jump relatively soon.

Read - PC Perspective results
Read - Hot Hardware results
Read - Eee PC News results

[andrew @ Jul 29th 2008 1:41PM]

And the 200mhz clock speed difference was not mentioned? Should try and compare like with like. That could make all the difference.

[phanbouy @ Jul 29th 2008 1:49PM]

from the article: "While the VIA Nano L2100 processor we tested had a 200MHz, or 12.5%, higher clock frequency than the 1.6GHz Atom 230, it typically outperformed the Atom by more than 15% to 20% in the applications we tested."

[Kax02 @ Jul 29th 2008 1:54PM]

In multiple ways actually.

VIA Nano™
L2200
1.6GHz
800MHz
NanoBGA2
65nm
100mW

As opposed to the 500mW rating carried by the 2100 at 1.8GHz.

Maybe it was supposed to be a demo of flagship versus flagship though in which case I believe the current top speed model is the 1.6.

[Mam00th @ Jul 29th 2008 2:00PM]

Performance is about a lot more then just clock speed although I would bet that the Nano would still top the Atom if it was at 1,8ghz

[Pochi @ Jul 29th 2008 2:35PM]

Power consumption is the key with these. And Intel is still king.

[IvaT @ Jul 29th 2008 2:37PM]

No mention that the Nano uses 5 TIMES as much power as the Atom. If the clock speeds were identical and the Nano still bested the Atom by say 5%, I'd take the Atom in a heartbeat solely because it used 5x less power. It's a no-brainer, except for Intel haters...

[phanbouy @ Jul 29th 2008 2:43PM]

@pochi

yep, the VIA used about 40% more power at load, but their power consumption at idle was about identical. if these super-duper power saving features VIA's touting are all that, we might find real life battery drain for average use to be on par (what should've really been tested)

[phanbouy @ Jul 29th 2008 2:48PM]

@IvaT where is this "5 times more power" coming from? RTFA or give a source

from the article: "Intel's part has an obvious advantage, consuming 19 fewer watts than the Nano under load, but remember the Nano is 1.8GHz, 65nm part, while the Atom is a 1.6GHz, 45nm part. If / when VIA is able to transition the Nano to a more advanced 45nm process, its power consumption should be reduced significantly."

19 / 45 = 42%. math is guud.

[Nimblesquirrel @ Jul 29th 2008 6:10PM]

Compare like with like?
Ok, lets compare the Atom with it's Eee predecessor, the Celeron-M.
At 1.6GHz, the Atom runs at neary twice the speed of the 900MHz Celeron-M. Does the Atom offer twice the performance? Nope. There are plenty of benchmarks out there that show that the Atom only offers marginal performance increases over the Celeron-M. If the Celeron-M was clocked to 1.6GHz, it would spank the Atom. Power consuption and heat dissapation are a different story, but you wated to compare 'like with like'.
With all the different processor technologies out there, clock speed really is no means of comparison.

[David @ Jul 29th 2008 8:07PM]

@phanboy - You are right, I was being generous when I said 5x. It's really closer to 19x. Maybe you should stick to things you know about.

"The VIA system uses 17.4 more watts while under load than the Intel system. That means if they are under load for one second, the VIA system has used 17.4 more watt-seconds than the Intel. On the other hand, the Intel only uses 0.9 more watts under load than the VIA uses when idle. So, if the VIA works for one second, then drops back to idle while the Intel is still working to complete the task, it would take 18.33 more seconds before their total power usage becomes the same (17.4 + (-0.9 * 18.33) == 0).

This means that to use the same amount of power over the long term, the VIA has to be able to complete tasks in ~5% of the time that the Intel takes. Put another way, the Intel would take 19 times as long to finish any task.

But, their own graphs show that the Intel system only takes about 2 times as long (at most) to finish tasks. So, there's no way that the VIA uses less energy over the long haul, but that's what they claim.

Their problem is that they are comparing the total power used by the VIA system only while "active" on the task, and not looking at the power that the VIA uses on idle while the Intel is still completing the task. This is like saying that as soon as you finish that task, you immediately power off (or drop to standby)."

[loosely_coupled @ Jul 30th 2008 10:21PM]

Where is the Celeron M ULV comparison? That thing wipes the floor with Atom. If they made it at 45nm, it would be great!

[Mam00th @ Jul 29th 2008 1:48PM]

There's no doubt the Nano is more powerful then the Atom, but the major advantage of the atom is its low power consumption. Although I would bet the Nano is able to crank more performance per watt.

[phanbouy @ Jul 29th 2008 1:50PM]

intel's advantage is 45nm vs 65nm fab. once that changes (hopefully) competition will be even better

[Mam00th @ Jul 29th 2008 2:06PM]

@phanbouy

Although you are right that the manufacturing process play a good part in term of power consumption, the design is the key. The Atom simpler design consume a lot less then the Nano superscalar out-of-order design.

Another advantage of the Atom is also the cost. Like you mention their 45 nm process help them keep the cost down as there are more die per wafer but there is also the 300mm wafer and the small size of the chip that helps keep the production cost down.

[retro77 @ Jul 29th 2008 1:48PM]

Apple and Oranges.

[From My Cube @ Jul 29th 2008 1:50PM]

what are you saying here? Steve Jobs hates florida orange juice?

[TRLK @ Jul 29th 2008 1:54PM]

Yes! Finally! I wondered when an Apple reference would come up

[Chris @ Jul 29th 2008 2:17PM]

@From My Cube: I thought he only liked Apple juice.

[retro77 @ Jul 29th 2008 2:25PM]

you kids fail...miserably

[Bandit5317 @ Jul 29th 2008 8:07PM]

Guys, I honestly don't think he meant that as an apple reference.

[broli @ Jul 29th 2008 1:51PM]

Sure performance might be better but any word on the other obvious thing...eehm THE PRICE. Let's also not forget that VIA partnered with nvidia lately. AND nvidia has the Tegra platform which beats atom graphics wise any time of the day with less power.

[Mike @ Jul 29th 2008 1:52PM]

I'm still waiting for their C8 processor, which sounded amazing (esp about the "playing Crysis" part)

[Ladislaio @ Jul 29th 2008 3:23PM]

That would be the nano running at 1.8 ghz. You know, the thing that is in these netbooks that were benchmarked(but with a nvidia Geforce 8600 for graphics to play crysis)
http://www.engadget.com/2008/06/05/video-mini-itx-2-0-with-via-nano-really-does-play-crysis/

[TRLK @ Jul 29th 2008 1:53PM]

That was... well unexpected I suppose. I always thought VIA was the epitome of crap CPUs along with Transmeta Crusoe. Seems like the tables have turned...

[Ninjakamster (PS360 FTW!) @ Jul 29th 2008 2:12PM]

Same here, had a negative view of Via with their horrific performance on their C7 processors.

I would love to buy the HP Mini-Note, but with a more powerful processor, the Nano looks to be the ticket!

[Minilap @ Jul 29th 2008 2:04PM]

Wake me when either of these companies release the dual core version at 2.0Ghz.

[a12ctic @ Jul 29th 2008 2:16PM]

Why do you need a dual core on your sub notebook -_-

Don't you think that might be just a little too much?

[YpoCaramel @ Jul 29th 2008 4:43PM]

Now that's just random. Want us to wake you up when netbooks have 1TB of RAM too?

[Minilap @ Jul 29th 2008 2:29PM]

Because even on a sub notebook I still multi-task, so why not? As long I have the option to turn off the second core to save some battery I don't see how anyone not want a dual core for an extra $20, or even less.

[a12ctic @ Jul 29th 2008 2:32PM]

Because it will cost more and not add massively to the experience. Single core chips multitask fine, the dual/quads help mostly for encoding (and some of those rare multithreaded applications) neither of which I would use my sub notebook for. These aren't supposed to be your main computer.

[Minilap @ Jul 29th 2008 2:34PM]

How was that random? Intel was supposed to release the dual core in Q3 anyways. Plus you guys do realize that these processors are not just for mobile devices. They are used in desktops as well if you are looking for a cheap low end but small and silent PC, and dual core will really make a BIG difference in that case.

[Tired_ @ Jul 29th 2008 2:37PM]

That's nice. We've been hearing about how awesome Isaiah/Nano will be for months now, but the best processor is the one you can actually use. Plan on selling these anytime soon, Via?

[OneLove @ Jul 29th 2008 3:04PM]

I like cheese.

[Kevin C. Tofel @ Jul 29th 2008 3:19PM]

"so hopefully we'll be seeing machines like HP's Mini-Note make the jump relatively soon."

I don't want to see that at all. The Nano's TDP maxes out at 25W, the current C7-M inside it hits 8W and the Intel Atom is 2.6W. Putting a Nano in the HP Mini-note (all things being equal) would drop the useful run-time to maybe 90 minutes on the 3-cell battery and 3 hours on the 6-cell battery. This CPU isn't meant for netbooks...

[Sarig @ Jul 29th 2008 3:43PM]

Okay, so that was the Atom "Diamondville", for the netbooks. Now how does the Atom "Silverthorne" that's smaller, and aimed at UMPC's perform?

[decapitor @ Jul 29th 2008 3:46PM]

All these new mobile processors call for TRIANGLE GRAPHS DAMMIT! One axis is power consumption one axis is price and the other is performance. Plunk them down and see which is best for whichever applications you need them for. And yes there is a good 30 Rock quote for this but seriously I call for triangle graphs.

[snorp @ Jul 29th 2008 4:02PM]

The VIA video drivers are absolute trash, though, so it doesn't really matter if the hardware is marginally better.

[Name @ Jul 29th 2008 4:25PM]

Good news if you like the OQO since I think the Nano is speculated as being a part of their next iteration.

Regarding the power consumption difference, I get the impression from the articles that the power under load discrepancy is not all that significant if you consider that the Nano will be under load less time since it is a faster processor:

"The Nano uses more power at load than the Intel Atom CPU does but it is also a much faster platform overall...if you need a low power part that performs well then the Nano is for you but if need the absolute lowest power consumption and can give a little back in performance then the Atom will fit your design nicely."

However, the real test is in a mobile device on battery power. I sure would like to see the battery life comparison.

[Corsair @ Jul 29th 2008 5:12PM]

Many of you do mention that Nano consumes more power but thats not the case. When they do compare, they fail to mention that NANO does not have a Southbridge chipset but ATOM does... and this chipset consume enough power... resulting in very small wattage differences between these platforms!

[Steve Paine @ Jul 29th 2008 7:52PM]

The differences at the netbook level are going to be so small that disk, software and mobo design are going to be the most important elements.

*The CPU wars are over. Design is king*

[bebop @ Jul 30th 2008 2:13AM]

The review measures total system power. Indeed small difference in the CPU can be ecliped by larger differences in the chipset, but the important thing to evaluate is platform power consumption: which bit of silicon does what is not important. On this metric, Intel was better than VIA, by a wide margin.

[thedesolate1 @ Jul 29th 2008 11:46PM]

what the hell kind of benchmark tests are those? why dont they put the AMD Turion benchmarks instead???

[thedesolate1 @ Jul 29th 2008 11:49PM]

Instead of Athlon X2... and core 2 duo for notebooks instead of the desktop version...

[bebop @ Jul 30th 2008 2:13AM]

Sorry for the long comment, but that PC perspective review was annoying as hell.

Its a clear win for the Atom, trying to fudge the issue by calculating the "energy consumed" over the time the job takes to complete is a selective criteria at best, bogus at worst. Who only has the PC running only when they have a 100% encoding job, turning it off the second it finishes?

No, you leave the machine running, and most people most of the time have the CPU running at idle or close to it. Atom platform consumes a little less power... and the mini-ITX test hides the difference on a large baseline... on a netbook drawing only 5-6W total, the 1-2W difference will translate into significant difference in battery life.

Second you are interested in what happens when you run a load, be it some encoding, or more typically, a movie or something. The atom N270 has a TDP of 2.5 W, so even 100% loaded its only going to produce less than 3W heat. The VIA Nano, in contrast, is going to spike to nearly 20W! Not only will the battery life be in the toilet, there are huge issues with heatsinking, fan noise and so forth. While its ok for a mini-ITX SFF PC, putting one into a Netbook like the hp 2133 doesn't make sense.

For years VIA had in the C7 CPU something that was comparable to a Pentium II mmx. Now, with the Nano, they seem to have taken a credible stab at replicating the Pentium M. Its a nice step up, but still years out of date.

All Intel has to do is put out a very slightly higher clocked Atom to completely spank the Nano. Or they could do as they are, and release the dual core N330. Still running under the Nano in both idle and load power, it will completely outclass it on any multithreaded apps.

[Corsair @ Jul 30th 2008 4:35AM]

You are ignoring what I have mentioned a few posts above! Atom Platform does not consume only 3watts but A LOT MORE because you will have to take into account the Southbridge chipset which consumes around 8-10watt.

Nano is integrating Southbridge functions, hence it does not need one. The Nano L2100 (1.8GHz) peaks at 25watt and the 1.6GHz model peaks at 17watt.

So, if we compare the two platforms (not processors alone, because it doesn't make any sense) at the same clock speed, the wattage difference is too small to worth mentioning and the rest of the benchmark should be identical.

[Corsair @ Jul 30th 2008 4:35AM]

You are ignoring what I have mentioned a few posts above! Atom Platform does not consume only 3watts but A LOT MORE because you will have to take into account the Southbridge chipset which consumes around 8-10watt.

Nano is integrating Southbridge functions, hence it does not need one. The Nano L2100 (1.8GHz) peaks at 25watt and the 1.6GHz model peaks at 17watt.

So, if we compare the two platforms (not processors alone, because it doesn't make any sense) at the same clock speed, the wattage difference is too small to worth mentioning and the rest of the benchmark should be identical.

[Corsair @ Jul 30th 2008 4:43AM]

You are ignoring what I have mentioned a few posts above! Atom Platform does not consume only 3watts but A LOT MORE because you will have to take into account the Southbridge chipset which consumes around 8-10watt.

Nano is integrating Southbridge functions, hence it does not need one. The Nano L2100 (1.8GHz) peaks at 25watt and the 1.6GHz model peaks at 17watt.

So, if we compare the two platforms (not processors alone, because it doesn't make any sense) at the same clock speed, the wattage difference is too small to worth mentioning and the rest of the benchmark should be identical.

[Hallx @ Jul 30th 2008 8:58AM]

Please people cut the crap and see through the FUD of power consumption under load. IMO, power consumption under load is not right criteria. From TFA by pcper:

"For our MP3 encoding test, the VIA Nano processor used a total of 37,323 watt-seconds (Joules) of energy while the Intel Atom processor used 38,290 watt-seconds (Joules) of energy. That is a difference of just 2.5% indicating that even though the Atom processor is slower, it's not that much less efficient than VIA's Nano."

hence in power consumption over a given task, NANO is better and idling with comparable power consumption. IMHO, NANO beats ATOM hands down.

[TJ @ Jul 31st 2008 11:55AM]

dude there's no fucking way those processors used 38,290 joules of energy. perhaps there's supposed to be a dot instead of a comma there since that much energy can throw an Eee PC 4.2km up into the air (~2.63 miles) if you neglect air resistance.

[fwupow @ Aug 3rd 2008 6:05PM]

Under certain benchmarks, chicken poo plugged into a socket 7 motherboard beat the Intel Atom processor.

I'm so tired of all the propaganda spinneroo shenanigans nowadays. There's no point in reading anything purporting to be news anymore.