Via: How Do My Readers See Ya?
Having taken the latest in a series of stabs at assessing AMD’s chances at future success last week, it’s now time for me to turn my attention to Intel’s other primary x86 competitor, Via Technologies. Via has, I admit, long has inhabited a soft spot in my heart (personally, that is…professionally, I of course have no favourites), in spite of its longstanding reputation for buggy core logic chipsets…;-) (and in part, perhaps, coming from the same underdog mindset that justifies my longstanding fondness for the Chicago Cubs and Sacramento Kings). Many years ago, the company began touting a microprocessor message that while quite controversial at the time, has now become conventional thinking:
- CPUs are now fast enough to tackle most mainstream applications, therefore
- The industry should instead be directing more focus at reducing power consumption…
- …and cost.
Via obtained its CPU expertise by virtue of a mid-September 1999 acquisition of Centaur Technology, then a subsidiary of IDT. And Centaur’s head is Glenn Henry, who in a previous life was (among other things) CTO of Dell Computer. Back in 1991, when I co-launched Intel’s first flash memory specifically architected for PC BIOS storage, Dell was the first major OEM to embrace the chip and the concept (the fact that the company then obtained the lion’s share of its motherboards from Intel had nothing to do with this…ahem…), and Glenn graciously contributed a testimonial. So I guess you can say that he and I go back a long way….though I highly doubt he remembers me.
Aside from its general visionary stance on PC microprocessors, Via was also a company I kept a close, consistent eye on because it tailored its products more directly towards the bulk of EDN’s engineering audience than did either AMD or Intel. Let’s be frank; for both of the latter companies, the embedded market is fundamentally an after-thought, intended to keep fabs full during personal computer capacity glut periods, and conversely the first set of customers to go on the constrained-supply list once PC demand takes off again. Now look at Via. Company representatives will talk my ear off, if I let them, about embedded computing opportunities such as network appliances, robotics and PC-inclusive displays at every meeting opportunity. Chips come with extended life cycle guarantees, critical for system designs that aren’t obsolete by intention in less than a year. And, if you don’t want to roll your own board design, the company will gladly sell you a bag-o’chips in embedded-tuned mini-ITX, nano-ITX or pico-ITX board form (including the pico-ITX-based ARTiGO).
Speaking of mini-ITX, some of you might recall the two-part hands-on testing I did of Via- and Intel-based boards back in 2004:
- Hands-on project: And then there was one
- And then there was one: one more time
- Article Addendum: And then there was one
Along with a 2006 followup piece:
I won’t mince words; Via’s products weren’t stellar performers, versus the Intel-based alternatives. From an absolute standpoint, I wasn’t terribly concerned about this fact, since I was running full-blown Windows O/S and application suite builds and presumed that the Via-based designs would run in a more application-acceptable fashion with the stripped-down Windows, Linux and other operating system candidates more commonly found in embedded designs. But from a relative standpoint, Intel’s speed advantage, coupled with its much higher volume cost advantage (and nowadays, its much improved power consumption profile), should give both Via and its customers great pause.
Why? I’ll answer this question by means of analogy. Earlier this week, a few days after the publication of my AMD analysis, not-unexpected news that the company would be substantially cutting its headcount in the coming two quarters (more) came to light. Yesterday afternoon, two other EDN staffers and I were cyber-conversing about AMD’s situation and one of them made the valid point that to be a player in the x86 CPU space requires a substantial amount of R&D investment regardless of the consequent return. What applies to AMD applies equally to Via. And looking at Via’s penetration to date of the traditional PC space reveals some troubling data points. The company’s processors are not in some successful systems that you might anticipate they would be by virtue of Via’s earlier-documented focus areas: the OLPC (AMD), for example, or Asus’ Eee PC line (Intel). And what about the systems that are Via-powered?
- A line of Lindows/Linspire-based desktop PCs sold in Fry’s Electronics stores got lukewarm reviews, in spite of a sub-$200 price tag.
- A more recent, conceptually similar Everex-branded and gOS-based line sold by Wal-Mart was, after an initial impressive sales blip (likely fueled by rabid Linux enthusiasts) pulled from retail stores (due, I suspect, both to dwindling sales trends and ballooning nonstandard software suite-based technical support call and return costs) although they’re still available through the merchant’s website. On this point, I should note that Via sells a $60 developer’s board based on the gPC design that, from my perspective, appears to be right up my readers’ alleys…
- Everex also developed an Eee PC-targeted laptop design called the Cloudbook, which ended up coming to market substantially later and more expensive than originally planned, and was roundly panned, to boot.
- Everex-partnered and Via- and subscription-based Zonbu desktop and laptop designs didn’t fare very well from a review standpoint, either.
- Via finally cracked a tier-1 PC supplier earlier this week with the unveiling of HP’s Eee PC-targeted (note the trend?) Mini-Note 2133. But once again it got panned by reviewers (note the trend?).
Continue reading with ‘Via: The PC’s The Thing Wherein You’ll Catch The Processor King‘…
ugh. via commented:
