Interview: Intel CTO on future technology

IT Week: For the last 30 years the interface for computers has been the keyboard. As Intel's chief technology officer, do you think something better will arrive?

Pat Gelsinger: Fundamentally, I don't think we'll get away from the keyboard, we'll just augment it in suitable ways. We didn't demonstrate this [at the Intel Developer Forum] but by using an array microphone you can track the user as they speak and dramatically reduce the signal-to-noise ratio, which really improves speech recognition. Add in camera tracking and you can do visual recognition and lip reading. You can mimic the way that the human ear focuses sound by using six directional microphones. We are now getting to the point where we have the computational power to do all this and make an interface perfect for voice input.

Hard drives are currently outstripping Moore's Law [that densities will double every 18 months] in storage capacity growth but will we see changes in how and where hard disks are used?

At the moment there seems to be no end in sight for the potential of hard drives to scale up. Hard drives will continue to get bigger and better. But in some circumstances they are unsuitable. For example, to get better mobile computing we need to eliminate the hard drive in laptops. Drop your laptop and the two things that will break are the screen and hard drive. Moving to solid-state memory would solve the latter problem. While there's no solid-state memory on the immediate horizon that will get rid of hard drives, it could be used soon. I've just got my first 1GB USB Flash key fob, for instance. Now 1GB isn't quite enough for me, but 10GB - that might work.

Low battery life is the bane of the mobile computer user's life. Are we any closer to developing an all-day battery?

It's becoming clear that there isn't going to be a miracle in battery technology. The capacity of batteries has been improving at a steady four to six percent a year and that doesn't look like changing. We therefore have to keep driving down energy consumption. I get five hours use out of my laptop, which is pretty good.

So, do you predict that any forthcoming technologies will make a substantial improvement to useful battery life?

With Dothan [the next-generation Centrino chip] we'll be getting into the six- or seven-hour range. All-day computing is a realistic step from that, particularly if you start having low-power operating modes, like the small secondary screen on a laptop lid we demonstrated, which can handle email and other functions. As things improve you'll see the laptop become more of a personal productivity platform that could be left on all day and used as needed.

For nearly a decade there has been talk about the coming quantum computing revolution, yet it seems no nearer. What is it that is causing the delay?

It will probably be talked about for another decade too. It certainly won't be relevant for another decade. That doesn't mean it's bad research but it's far from coming to the commercial sphere. We need to challenge some of the assumptions about quantum computing and what it will be used for. We're now building 3bit quantum computers with possibly a 5bit one under construction. That's fine, but I'm already building 64bit computers. When we get quantum to 14bit then it can be used for encryption, which is one of the key applications for it. The country that gets a quantum network first will have a real competitive advantage.

ABOUT PAT GELSINGER

Pat Gelsinger is Intel's senior vice president and chief technology officer, working in the company's Corporate Technology Group.

Gelsinger joined the company in 1979 and was the chief architect of the i486 processor.

He later managed the Pentium Pro group and helped define Intel's ProShare video conferencing and internet comms product line.

Like this story? Spread the news by clicking below:

 del.icio.us     Digg this     reddit!