NVIDIA hopes to grow their Tegra business to eventually make up 50% of their revenue. By scoring a win with the Zune HD, possibly ending up in the future Nintendo handheld and Apple products, and countless other media, phone, and computing devices, it's no wonder why their expectations might be high. SoCs have always been very popular in the ultra-portable scene, and Tegra is among many leading the way for the future of this technology sector. With hardware accelerated flash, video, graphics and audio support, the capabilities of such SoCs has grown to the point of surpassing the form-factor of just smartphones, to encompass a vast array of devices extending all the way up to notebook-like devices, dubbed "smartbooks".
It's for this reason that ARM is becoming better positioned to take the computing world by storm in the near future. With their recent partnership with the newly formed GlobalFoundries manufacturing company, it's clear they intend on increasing the capabilities of their chips beyond the scope of what they're best known for today.
Right now there's a pretty large gulf between the major electronic devices that we use everyday that separates the two major forces in the computing industry. They are Intel and ARM, and both companies are seeking to cross this gulf at an increasing rate these days, extending their businesses and eeking into each other's markets. While ARM is best known for the popularity in cell phones, their expansion into the smartbook category thanks to Qualcomm's Snapdragon and NVIDIA's Tegra is just the beginning. Intel is doing similar things, extending their x86 standard used in larger PCs into MIDs, and soon smaller devices through their Atom product line. Intel has a long way to go if they hope to get Atom into the next big smartphone, and similarly ARM has a long way to go if they want their chips used in full-blown computers.
Arguably ARM has the bigger challenge. While Intel needs only to continue to shrink their Atom architecture into smaller and smaller sizes, ARM not only has to contend with performance considerations, but also software ones. Thanks to the Linux persuasion, a lot of the most popular applications we use are not necessarily tied to x86, but most of them are. Particularly Windows, the biggest application of them all. Windows CE hasn't received much attention from its parent company Microsoft, and lags far behind in the features x86 Windows users, and any other major OS user is accustomed to. That leaves Linux variations, but again we go back to the software support issue. It's not enough if an OS can browse the internet and play media fine, it has to be able to run the software your typical computer user is used to.
Both companies have to deal with the tight market penetration that the other has. ARM enjoys a much more open and versatile business structure that plays much more friendly with the demands of small device makers. Intel, by virtue of the fact that they are such a large company, isn't quite so nimble, and while their x86 standard plays such a role in the computing world that it should probably be stripped of licensing restrictions, Intel will still try to hang onto their IP with all their might. And while ARM gets all the software support, with all the money Intel is able to throw around, undoubtedly that won't be much of a problem for them, especially given their typical strong-arm tactics. This, too, will work in their favor when it comes to courting the device makers themselves. ARM, on the other hand, has no such money reserves, and while it's unlikely they'll ever be performance hounds, just getting into the general computer market with Intel opposing them will be a nearly impossible task.
ARM does have a few advantages though. Because they are so open, a lot of different companies can make designs with their instruction set, leading to diversity and intense competition, which should help to push performance. Their instruction set is generally evolutionary, meaning they'll likely benefit from a great deal of software back-compatibility, a trait they share with x86. It is for this same reason that many software developers (the smaller ones, at least) are eager to move away from the x86 standard. x86 gets a lot of blame for holding back innovation, and its the same reason why many don't think it has much chance in the ultra-mobile market. The biggest question Intel faces in getting their x86 into these smaller form factors is....why?
Why, indeed? An old standard that has grown bloated with age being positioned to enter a mobile market that favors slenderness and adaptability? And what about the other parts that make up an SoC? Do we really want what Intel considers graphics acceleration in even more of our devices? ARM & partners have the slick designs to get the job done in a space smaller than Intel has been able to manage. The only question is, do those slick designs scale upwards?
Before we can even answer that question, another question arises: Would Intel let them in, even if they do? The answer to that may be coming, and the opportunity for ARM to get their foot in the door is already here. Smartbooks may just be the beginning. Then it's "nettops", and bigger notebooks, and pico-ITX-based computers, and as the popularity continues to grow, so will the software support, and with the backing of many large electronics corporations, the expansion of their manufacturing and growth of their designs, who knows what the future will be? The bell for the first round between ARM and x86 is about to sound. Anyone wanna place their bets?
Now if they could just declare x86 as public domain, things will really get cooking.
