Backtrack about a week. Intel rolls out its suit against VIA Technologies and S3 Graphics Inc. on September 7, 2001 in the United States District Court of Delaware. It alleges that in creating the P4X266 chipset, VIA and S3 have infringed on five Intel patents issued between 1997 and 2000. These patents cover bus technology--the way in which a microprocessor interfaces with the rest of the world. Considering that the P4X266 will bring double data rate (DDR) memory to the Pentium 4, bus architecture would seem to be the issue du jour here.

But wait, there's more. VIA's suit, filed five days later, has nothing to do with the CPU bus. It strikes at the heart of the Pentium 4, alleging that Intel has infringed on a patent (#6253311) which is jointly owned by VIA and Centaur Technology. This patent deals with the different formats in which numeric data may be stored in a microprocessor. If this claim is valid, then the bus issue pales by comparison. (VIA, through its acquisition of Cyrix, is also involved in the design and manufacture of CPUs.)

However, here's the wrinkle in VIA's suit: Intel was granted a patent (#5889984), not named in either of the two legal actions thus far, that precedes VIA's by nearly two years and which appears to differ, on cursory view, by only marginal elements of language. Naturally, when talking to the two combatants--er, companies--they speak on the advice of their respective counsels, so, each is totally assured of the merits of its own case and is dubious of the merits of the other's. This is how lawyers earn $500/hour and hone their skills in rationalization.

Of the two suits, VIA's is the more critical. If Intel has crossed the line in creating the internals of the Pentium 4, the payoff will be monumental--both in actual reparations and in the increased cost of new CPUs to cover damages. Even if VIA's suit is proven valid, that doesn't affect Intel's infringement claim, as the two cover entirely different fields of endeavor. On its part, the Intel suit, if proven correct, will at worst cancel--or at least delay--and incrementally increase the cost of VIA's P4X266 chipset.

The most likely result of these legal actions, if they follow the historical pattern of court battles between the two, is that Intel and VIA will leverage their claims against each other into some sort of mutually complimentary settlement that lets them hide their daggers behind their backs and beam toothy smiles of contentment at the cameras. You and I, on the other hand, will have no reason to feel content. During the process, we will have been denied access to less expensive systems and, after all is said and done, what we will get will be either cobbled Intel first-generation single data rate (SDR) SDRAM boxes or pricier VIA-based DDR SDRAM systems.

Although not the perfect analogy, we had a similar situation back when the "the phone company" did its almighty best to keep other parties' instruments from being connected to its lines. There were some valid concerns at the time but, by in large, it was all meant to maintain a monopoly. In the case of Intel and VIA, the court doesn't need to settle just these claims (or those that are ongoing from past actions) but rather to finally be definitive about where a microprocessor stops and the rest of the world begins. Without that specific judgment, we're doomed to go through this techno litigation loop time and time again.

What's your position on the Intel-VIA lawsuits? E-mail Bill or post your thoughts in our Talkback forum below.

For more coverage on Intel and its commitment to Rambus, see Readers livid, Rambus vulnerable, commentary from TechUpdate's Editorial Director, David Berlind.