VIA Nano (Isaiah) is officially released

Ah, but Intel measures TDP as average power usage, while VIA's numbers are an absolute max. According to VIA, a 2Ghz Nano should pull between 2-5W under normal usage. Not bad considering the Nano is manufactured on a 65nm process compared to Atom's 45nm.

That's also just the CPUs. Via has a lot of low-power platform experience. I personally cannot wait to see two systems side-by-side, running benches while having their power usage measured.

I would really like to build myself a Mini-ITX HTPC based on a 45nm dual-core version of Isaiah next year. :)
 
a Mini-ITX HTPC based on the VIA Nano would be sweet.

I can't wait to get my hands on one of these
 
Rabid Badger said:
Ah, but Intel measures TDP as average power usage,
Click to expand...
Um, no. But points for about the 5th or 6th redefinition of TDP that I've read. :p TDP is a measure of the cooling solution needed to run the processors as tested with "commercially available software." It isn't a maximum theoretical power the chip can use, but even AMD no longer uses that definition in the data sheets, despite what Ars states below.

Intel measures "average power" as average power: http://arstechnica.com/news.ars/post/20080529-via-takes-the-wraps-off-isaiah-meet-the-nano.html

The Atom chips use in the 0.16W-0.22W range on average, "measured CPU power whilst running BAPCo MobileMark'05 Office Productivity suite on Microsoft Windows XP for a period of 90min at 50°C" according to Intel in the Ars link above. That's not bad really since it's pretty representative of typical use.

The Isaiah does look like it will be a better choice in the miniITX form factor though since it runs faster than Atom, while still using a relatively low amount of power. But in MID and UMPC designs, Atom should have edge in smaller form factors and longer battery life.
 
pxc said:
Um, no. But points for about the 5th or 6th redefinition of TDP that I've read. :p TDP is a measure of the cooling solution needed to run the processors as tested with "commercially available software." It isn't a maximum theoretical power the chip can use, but even AMD no longer uses that definition in the data sheets, despite what Ars states below.
Click to expand...

I sincerely doubt VIA is using the same definition considering their website uses the terms "TDP" and "Maximum Power" interchangeably.

As I said, I'll wait until I see both platforms plugged into power meters to make my judgment.
 
Nano may be a nice addition. I remember reading the preview 6 months ago; it was a nice balance between the old single-pipeline C7 and the multi-pipe designs of Intel and AMD, with the resulting design closer to a Pentium III with an improved SSE unit than anything else. I'd like to see the performance numbers and actual power usage.

But with dual-core Atom coming out later this year at the same power point, plus low-voltage Core2 Duos coming in at under 25w TDP, Intel may give Via a run for their money.

EDIT: here's some basic benchmarks between the Atom and Nano. Floating-point performance is exactly the same, as the architectures would suggest, but Isaiah has the edge in integer processing due to out-of-order execution.

http://laptopcom.blogspot.com/2008/04/first-benchmark-atom-cpu-vs-via-isaiah.html

But then, at the same clock speeds (1.8 GHz) the Atom uses ~4w and the Isaiah uses 25w. Now I know why they gave a "performance per watt" teaser comparing Nano with the Celeron M, and not the Atom: the Atom would tear the Isaiah a new asshole in performance/watt.

Mind you, the performance/watt will be closer if you go with the Isaiah U series, but Intel has a similar line of ultra-low-power Atom that sip less than a watt TDP. Basically, Isaiah has the same problems I guessed it would have: too inefficient to play with the ultra low power chips, and too low performance to play with the big boys on full-size entry-level notebooks (Celeron M).
 
Rabid Badger said:
As I said, I'll wait until I see both platforms plugged into power meters to make my judgment.
Click to expand...
Yeah, but 0.16W-0.22W vs 2W-5W that each manufacturer declares as "average power usage" isn't going to be bridged by what happens at maximum power. That also isn't likely to be close at all, no matter how it's defined: with 100% loads the top 25W TDP Nano will use much more power than the top 2.5W TDP mobile Atom. But the top Nano will have better performance too, and likely be priced much cheaper than the top Atom.

It's not a big deal really. Some of the lower power Nano chips will definitely be available in UMPCs and SFF desktops if you want to buy one. I am looking forward to Atom powered UMPCs though. If the rumors of dual core Atom chips going into mass production in July are true, I might hold off for one of those. Nano would be much more attractive at 1/2 the official TDPs.
 
I wonder how the Isaiah would do in the new ASUS EEEpc's that will have the Atom chip?
Yu know.... if they had the same socket.:p
 
You must log in or register to reply here.
Back
Top