CNET held a writing contest for a free pass to PC Forum.

The Question: What IT innovation have the experts underplayed...or even completely missed?

The Winning Answer by Peter Glaskowsky: The next five years will see major innovations in standard high-volume computing platforms such as virtualization, partitioning, and the universal deployment of multiprocessor configurations. All of these advances represent significant improvements, but none will be as pervasive and dramatic as the introduction of hardware-based security and reliability technology. Microsoft has delayed the debut of its Next-Generation Secure Computing Base technology, but this merely reflects the magnitude of the effort involved in the initiative. By the end of the decade, NGSCB will completely change the way we think about PC hardware and software. It will be literally impossible for unauthorized software to infect PC operating systems or applications. Even legitimate configuration changes will be strictly controlled-- and some changes will even be beyond the authority of the system's users and administrators. The same technology will support unbreakable digital rights management, effective piracy prevention, and-- depending on how it's used-- allow improved personal privacy or undetectable remote monitoring. Hardware will keep these controls from being bypassed under most circumstances, forcing users and IT staff to come to grips with these new limits on their customary freedoms. This new technology will shake up the computer industry far more than anyone suspects today.

WDYT? Got a better answer?

Here's a transcript of my speech-- at least, the way I originally wrote it. :-)

I'd like to thank the organizers for the chance to speak here this morning. I think I'm not really what they wanted, though. The CNet contest was meant to attract input from outside the usual circle of industry analysts and executives, to see what the rest of the world thinks is important to IT today.

Instead, they got me-- a former industry analyst, now an executive in the semiconductor industry. So let's concentrate on the fact that I haven't been HERE before, and I don't think most of you attended Microprocessor Forum while I was running that event, so perhaps I can still manage to say something you'll think is original.

Anyway, I'm here today to talk about a truly revolutionary new technology that will eventually change how we interact with our computers, and with each other. I speak, of course, of the ability to send Pizza Hut orders from within Everquest 2.

But seriously... I'd like to talk for a few minutes about the future of virtualization and security from a hardware perspective. This isn't what our company is working on; with any luck, I'll be back here in two years to tell you about that-- but I think virtualization is going to be very important for all kinds of computers, including PCs, servers, and consumer electronics.

Most of you are probably familiar with the Java sandbox, a virtual machine environment that isolates untrusted Java programs from the resources of the physical machine they're running on. It's called a sandbox because it's a safe place to play. The sandbox provides reasonably effective security, but Java programs running in the sandbox necessarily have severely limited capabilities, and the isolation mechanism-- basically, software emulation of a virtual machine on a physical machine-- causes performance issues.

We're about to see Intel and AMD introduce microprocessors with hardware support for isolated virtual machines. Software developers such as VMware and Microsoft will use these chips to put multiple independent operating systems on the same machine at the same time, with true hardware isolation that is much stronger and faster than the software mechanisms we see in VMware today.

So far, we've mostly heard about how hardware virtualization will give us the ability to run Windows and Linux at the same time, and have special system-management code running in the background so that IT managers can diagnose and repair problems even if the primary operating system has been corrupted.

But this idea goes much further. Each application in your machine-- or each suite of applications, like Microsoft Office-- can have its own separate virtual environment with its own operating system. Software developers can distribute their own special flavor of OS with each application, if that's the best way to deliver a consistent, reliable user experience.

This creates a situation more like video-game consoles, where each game is in total control of the machine and can operate without interference from user-installed or malfunctioning software. Data sharing between applications can then be accomplished using the familiar networking model rather than communicating through shared memory or local disk space.

If an application doesn't want to share, it won't have to. This creates the risk that applications will keep too tight a grip on the user's data, imposing unwanted digital rights management rules on text documents, home movies, and other kinds of files we expect to be able to share.

Unlike some people, I'm not too concerned about this risk, because I know market forces will prevent any such solution from succeeding. Hardware security will give us more freedom than it will take away. After all, Intel and Microsoft won't be able to see what's going on in these secure partitions any more than a virus will-- that's inherent in the concept. Hardware security for your computer is like door locks for your house, but potentially even more effective-- and honest people shouldn't complain about other people locking their doors.

Before I go, I'd like to add that I've been accompanied here by my girlfriend, Joelle Mellon, who is a librarian and schoolteacher in Baltimore. I told her about yesterday afternoon's discussions on educational policy, and she says that if we're going to turn our focus from our industry to hers, she'd like us to fix her computers first!

contributed by Peter N. Glaskowsky user1053 on Mar 22 8:32pm