18 Comments

Summary:

We had an e-mail tip-off earlier this week that Isaiah was coming, but had to keep it under our hats to be sure. I’d say an official press release from VIA confirms it as the Isaiah CPU is now announced as the VIA Nano family. Following […]

ViananofamilyWe had an e-mail tip-off earlier this week that Isaiah was coming, but had to keep it under our hats to be sure. I’d say an official press release from VIA confirms it as the Isaiah CPU is now announced as the VIA Nano family. Following a similar pattern to Intel offerings, Nano availability comes in multiple flavors, ranging in clock speeds from 1.0 GHz to 1.8 GHz, all with an 800 MHz front side bus. The 64-bit, out-of-order processing along with several powers-saving features (support for C6 sleep state and Adaptive PowerSaver Technology to name a few) give the Nano a nice performance per watt ratio.Although LAPTOP Magazine indicated commentary that the Isaiah wouldn’t be used in sub-notebooks, it’s clear from the press release that there’s a socket for newly named Nano chips based on this quote: “Initially to be launched in two skus, the VIA Nano L-series processors for mainstream desktop and mobile PC systems and the ultra low voltage U-series for small form factor desktop and ultra mobile devices such as mini-note.”Bear in mind that the ultra low voltage U-series chips top out at 1.3+ GHz with a maximum TDP of 8W, while the faster clock-cycle L Nanos are the 1.6- and 1.8GHz chips that can use two to three times more power, hence they’re likely to be in standard, not sub-notebooks or UMPCs. So, now that the wait for more information is over, what do you think of the Nano, especially in light of the Intel Atom?

  1. Yes I can’t wait for some real world testing now. I hope this years Computex really will be the second coming of the ultra portable market that it appears to be.

    Share
  2. They need to get silicon out ASAP, otherwise Intel are going to take their market.

    The Nano looks to be better than the Atom at the moment with regard to performance, but with the Dual Core Atom likely to be available before the end of the year VIA only has a small window of opportunity to get established before Intel catches up.

    Share
  3. The earlier Isaiah specs that floated around trouncing the Atom were impressive. But now I don’t recall what speed that (now-) Nano chip was. Could the ones for cheap subnotes be no better than, say, a Celeron (which also whips the Atom)?

    Share
  4. I don’t know how Intel’s TDP rating compares with VIA’s, but with 5W TDP rating, the 1GHz U2300 has to compete with and beat U2100 1.06GHz Core 2 Solo which has 5.5W TDP rating.

    The VIA Nano U2300 does operate on faster FSB (800MHz) vs C2Solo’s 533MHz, But I don’t know how much L2 cache it packs or much else. It could be tough, and I think it’s TDP is barely low enough for UMPCs where Atom will live…

    Share
  5. Mike, I think it’s far too soon to say. We need to see actual devices with these CPUs to test effectively.

    Shogmaster: the max TDP of Intel Atom is 2.6W. Good point on the two ULV chips.

    Share
  6. very disappointing if you ask me. where is the super chip running at 3.5w that we were all promised? what i see here is a standard 1GHz 5w CPU, why wouldnt i just go Intel? i can pretty much guarantee Core Solo will outperform the Isaiah & at the same TDP. theres even a chance that the dual-core Atom will outperform the Isaiah & probably come in around 3-3.5w.

    all those people were right about VIA constantly blowing it with subpar products released far too late, they just did it again.

    the only advantage i can see now is pricing.

    Share
  7. Atom’s 2.6W TDP won’t mean much when it’s so freakin’ slow. Most will be begging for higher TDP just to get some performance back.

    I know for fact that 1GHz Isaiah/Nano will destroy the 1.8GHz Atom in perfromance, with bunch of the benchmarks being out so far and having experienced Atom UMPCs and MIDs @ CES (most clocked @ 1.6GHz and even slower than C7Ms at the same speed).

    The good thing is that the Nano is using the same exact package as the C7M, so OEMs could switch to Nano in their existing UMPC products without any fuss (OQO, I’m looking at you) and get a nice minimum of 2X performance boost for their effort. But I don’t think Nano will end up in the smallest form factor devices that will harbor Atom because of it’s sub 3W TDP.

    But there will be overlap of Nanos and Atoms in certain form factors (7″ screen devices like Q1 Ultra for instance), and there’s where Nano could be the real processor of choice since Nano’s 60 square mm would be much cheaper than Any of the 65nm Core 2 Solos (which are just Core 2 Duos with second core disabled, making the die size as big as the Core 2 Duo counterparts)…. As long as Nano performs like a Core 2 Solo at similar clockspeeds.

    Share
  8. “theres even a chance that the dual-core Atom will outperform the Isaiah & probably come in around 3-3.5w.”

    Fat chance on both counts. 1.6GHz Atom couldn’t even outperform a 900MHz Dothan Celeron in the recent benchmarks. It would take more than a dual core for Atom to perform near a 1GHz Nano. And with twice the core, you are talking twice the TDP. 5.2W for 1.6GHz dual core Atom.

    From my rough estimates, 1GHz Nano could be close in performance to 1.06Ghz Core 2 Solo, with similar TDP and much better cost due to much smaller die.

    The real question is how fast Intel will counter with a 45nm Core 2 Solo, which would bring the die size near the neighborhood of Nano’s and lower the TDP as well.

    Share
  9. Shogmaster is full of BS, just by the way you word it it’s obvious you have no undertsanding of what your talking about. besides, you havent used any of the Atom devices you claim you have. your just speculating based on what youve read that other bloggers have posted but admitted they were prebuilds.

    besides anybody who makes a statement like this has no credibility whatsoever, twice the cores doesnt mean twice the everything (not even performance).

    “And with twice the core, you are talking twice the TDP. 5.2W for 1.6GHz dual core Atom.”

    Share
  10. Actually, I am glad to see that they are launching with ULV options (they would have been very stupid without one). I think the “L” is for larger budget laptops or desktops.

    I am quite excited about U2300 (1.0GHz) and U2500 (1.2GHz). I want to see HP 2133 with U2500. That should perform better than one with C7-M @ 1.6GHz. From the benchmark numbers, Nano @ 1.0GHz should be comparable to (or even slightly better than) C7-M @ 1.6GHz.

    As for the competitions, from the benchmark numbers, I think Diamondville @ 1.6GHz seems to be comparable to Celeron @ 900MHz. I think Nano @ 1.0GHz should be comparable in terms of the ALU performance. I don’t think Via is going to beat Intel in terms of TDP anytime soon, but considering the contribution of CPU to the total system power consumption (and what kind of batteries the systems come with), Via might be able to win quite a few design wins with Nano.

    Share
  11. Quoting Jkr
    ——————————————–
    Shogmaster is full of BS, just by the way you word it it’s obvious you have no undertsanding of what your talking about. besides, you havent used any of the Atom devices you claim you have. your just speculating based on what youve read that other bloggers have posted but admitted they were prebuilds.

    besides anybody who makes a statement like this has no credibility whatsoever, twice the cores doesnt mean twice the everything (not even performance).

    “And with twice the core, you are talking twice the TDP. 5.2W for 1.6GHz dual core Atom.”
    ———————————————-

    http://en.wikipedia.org/wiki/Silverthorne_(CPU)

    “Diamondville

    On March 2, 2008, Intel announced the new processor (code-named Diamondville) to be used in Classmate PC/Netbook. It is used in Intel’s low-cost Mini-ITX motherboards (code-named “Little Falls”).[14][15][16][17] It will supersede Conroe L by using Diamondville as single-core Silverthorne core (4W TDP) or dual-core Silverthorne core (8W TDP) running at 2.2Ghz each.”

    Share
  12. Oh, btw, the Atom processors that will be used for most of the ULCPCs are not Silverthorne (“Z”) ones, but Diamondville (“N”) ones. I think Diamondville has slightly higher TDPs than Silverthorne. (~2W vs. ~4W).

    Share
  13. Yes. It’s a different variant of Atom, but the point I was making to Jkr was that the only dual core Atom planned is the Diamondville core, and that indeed, dual core version has twice the TDP of the single core.

    But then again, this has always been the case with Intel’s dual cores. Twice the core, twice the TDP. He can check it out for himself at Intel’s website:

    http://processorfinder.intel.com/Default.aspx

    Share
  14. So what exactly are these chips going to be used for? Would any be of any competition to intel’s dual core ulv processors?

    Share
  15. Do Atom and Nano have offerings at the same TDPs? If I remember Atom’s TDPs correctly, the lowest Nano TDP is still higher than Atom’s highest TDP. If that’s the case, the two products do not really directly compete against each other. Having said that, I also have to note 5W-8W is squarely within OQO’s target TDP. So the idea of a U-series Nano in a device which fits in your pocket isn’t out of the question.

    Since Nano is out-of-order and Atom is in-order, I would expect Nano to perform better cycle for cycle. However, for a given TDP, Intel may be able to clock Nano faster. The results of independent, fair, competitive benchmarks may be really interesting.

    Also, TDP only starts to tell the power story. We have very little independent information about the power consumption of either one of these chips. That will be interesting too.

    Share
  16. Atom’s lowest TDP of 0.65W is for the Z500 clocked @ 800MHz, but the Z510 clocked @ 1.1GHz jumps to 2W TDP. Weird…

    http://arstechnica.com/news.ars/post/20080402-new-atom-cpu-marks-intels-baby-steps-into-embedded-world.html

    Anyways, with 1.6GHz not being able to outperform 900MHz Dothan Celeron M, I really don’t think you’d want 800MHz or even 1.1GHz Atom anywhere near your XP install, let alone Vista.

    As for the Nano and Atom not meant to be competing in the same space, you should tell that to the OEMs. They’ve already decided that those two will be competing in the same space. Many Atom announced devices are far larger than their 2W TDP would suggest. Many are planned for 7″ UMPCs and 9~10″ EeePC competitors. That’s exactly where VIA want Nano to compete in.

    Share
  17. Oh dear god! Will someone PLEASE make the brands easy to distinguish! Why are there Silverthorne and Diamondback (or wtf) for Atom?! I thought seeing Atom = Atom, 1 CPU. Now there are frikkin flavors?! Why should we have to keep track of this madness?! Why couldn’t they have done Atom-S and Atom-D to make it easier?!

    Share
  18. You can think of “Atom” as “Pentium”, “Core”, “Core 2″. It’s the family name. A family can have many variations.

    Atom started with “A” series, which I think was a very quick hack of an existing Celeron family. They announced the “Z” series and the “N” series. The “Z” series is Silverthorne, which was designed first for power and then for cost. The “N” series is Diamondville, which was diesigned fist for cost and then for power. That’s why the “Z” series has < 2W TDP and the "N" series has 4W TDP (and slightly bigger and cheaper, I assume). However, both the "Z" and the "N" series are based on a same core/ALU design. It's just that they were optimized differently.

    Intel’s marketing has been playing a game. They introduced the “Z” series first (I don’t think they officially announced the “N” series) which was smaller and consumed less. But I think all the low-cost netbooks (MSI, Dell, etc.) that are coming out are indeed using the “N” series (mostly @ 1.6 GHz).

    I think Nano @ 1.0 GHz (5W) should be comparable to or even better than the “N” series Atom @ 1.6GHz (4W). The prices (to the manufacturers) matter, so that and the manufacturing capacity would matter.

    Share

Comments have been disabled for this post