Brian's Brain

One in a series of posts…

Back in late March, for subsequent publication in late May, I wrote regarding Intel's Atom microprocessor:

Perhaps the biggest question on the company’s road map for the remainder of the decade is the degree to which Atom will cannibalize Intel products in a manner that is fiscally unattractive to Intel, instead of broadening the overall x86 market at the expense of competitors, such as ARM, as Intel hopes.

That was, as it seemingly has turned out (and if I do say so myself), quite a prescient observation. A month back during the company's most recent earnings briefing, in fact, Intel's CEO Paul Otellini got verbally 'roughed up' by numerous analysts as to the degree to which Atom's market entry would negatively impact corporate revenue and profit margins that are currently bolstered by profitable Centrino and Celeron CPU sales. Otellini's tepid response:

[Atom] is less than a third of the performance of our Centrino (high-end mobile processor). You’re dealing with something that most of us wouldn’t use.

did little to quell investors' concerns, although in this editor's opinion, he was largely spot-on. The PC market in the United States, for example, is saturated and as such, replacement-system sales will likely accompany a consumer expectation that the new system will outperform its predecessor in performance, storage capacity and other metrics. The bulk of Atom-fueled system sales here, therefore, will likely represent incremental per-household systems used for quick information access and other limited-function scenarios, much as I employ my Nokia N800 Internet Tablet today.

Such systems might have traditional (albeit shrunken-down) notebook PC form factors, a concept with which Asus has had tremendous success of late with its Eee PC line. I'd earlier mentioned MSI's Wind; Acer's price-just-cut Aspire One (reflective of just-launched and coming-soon competition from, respectively, Lenovo and Dell) also looks quite compelling. I confess that I've long been a fan of the thin-and-light subnotebook form factor, beginning with my first perusal of Radio Shack's TRS-80 Model 100 and accelerating from there with the first such system I owned, Dell's visionary but ultimately ill-fated 320SLi.

Conversely, Atom-centric systems might instead come in more pocketable dimensions, reminiscent of my N800 and Apple's iPod touch. Integrated cellular data and/or WiMAX connectivity is certainly possible as a supplement to the inevitable Wi-Fi. And if the manufacturer adds cellular voice service, too, you've got an Atom-based smartphone along the lines of my beloved T-Mobile Dash or Apple's iPhone.

To wit, as I've said before, I confess that I find Intel's "x86 everywhere" mantra (including the embedded verse) quite compelling, especially now that it's increasingly accompanying CPUs that don't come with the baggage of substantial performance and/or power consumption deficits versus ARM-, MIPS- and other alternative CPU architecture-based competitors. The ability to tap into an enormous, mature code and development tools base with little-to-mostly-no alteration is tremendous (right, software developers?). The same thought process applies equally well to Intel's upcoming Larrabee media processor versus existing GPUs from AMD/ATI, Nvidia and other suppliers, specifically with respect to beyond-graphics applications…but I'll save further discourse on this particular subject for a coming-soon dedicated post.

Atom's other key targets, perhaps obviously, are the emerging markets that formed the foundation of Intel chairman Craig Barrett's conference-launching keynote on Tuesday morning. Atom will enable hardware manufacturers to profitably hit price points which are currently unachievable with Celeron and Centrino CPUs and accompanying chipsets…price points necessary to stimulate broad adoption by potential customers who don't have the disposable incomes of their mature-market peers. For similar price-point-expectation reasons, along with the fact that Windows largely hasn't yet established a defensible 'embrace and extend' beachhead in emerging markets, open-source Linux and bundled application suites such as OpenOffice will be comparatively quite popular. And, as emerging market consumers' disposable incomes (and system capability expectations) grow, reflective of increasing societal prosperity, Intel and its partners will happily upgrade them to more robust hardware.

Speaking of upgrades (and price cuts), I confess that I don't have any sympathy for companies like Sony, who bemoan the "race to the bottom" that low-cost systems in laptop PC form factors potentially represent. The tech industry is an always-amazing case study of Darwinist evolutionary theory in action. As such, it's up to PC OEMs and their building-block hardware and software partners to continue stoking demand for mid-range and high-end system configurations, in parallel with cultivating new markets for low-end Atom-based gear in order to expand the overall ecosystem. If the suppliers don't continue justifying the need for traditional notebook PCs…if the technology treadmill stalls…then open-market consumers are completely within their rights to select lower-cost equipment that adequately meets their needs.

For those of you interested in test-driving Atom for potential development purposes, I'd earlier mentioned that mini-ITX reference boards are now shipping. Alternatively, you might want to go with a more assembled configuration such as MSI's $139.99 barebones system. Regardless, keep in mind that dual-core Atoms (with HyperThreading support, thereby effectively delivering virtual quad core capabilities for some instruction thread scenarios) are coming soon. Or perhaps you'd prefer an archaic, under-clocked AMD CPU, instead?