Desktops and laptops are about to be transformed by an array of emerging technologies, according to Dell technology strategist Brian Zucker. In a conversation with Computerworld's Robert L. Mitchell, Zucker tells how desktops and laptops are becoming - among other things - faster and smaller, and operating on wireless networks more frequently.

What emerging technologies will roll out in Dell desktops and laptops in the next 12 months?

The big thing this year is PCI Express, a serial interface similar to USB and 1394 that provides low pin count and higher performance. One port of PCI Express will yield 200 megabytes per second, which is twice what a PCI slot does today, and it takes up less than half the area.

The second half of this year is the launch of PCI Express products industrywide. The first thing you'll see is an addition or replacement of one PCI slot with a 1X PCI Express slot. That would be a great attachment point for gigabit ethernet or 1394b, which have traditionally saturated the PCI bus. The other thing you'll see is the replacement of the Accelerated Graphics Port socket with a 16X PCI Express socket ... to give you 4 gigabytes per second of bandwidth.

What impact will PCI Express have on your laptop designs?

The PCMCIA card is being replaced by PCI Express card technology (called Newcard), and that's a smaller form factor, almost half the size of PCMCIA. It has a PCI Express or USB 2.0 interface. You can design systems that are smaller.

On the other side, you'll see a lot of focus on pushing the desktop into a smaller form factor.

What other technologies will we see from Dell?

Wireless (LAN features) will become even more ubiquitous. Pretty much every notebook will end up having wireless integration. Ultrawideband technology is also very promising as a cable replacement, kind of like Bluetooth. You're talking 100 megabits per second at about 10 meters, and you can get higher rates, up to 480 mbps at 2 meters. It is a solution for audio and video. . . . The way UWB works is it extends across a large spectrum--3 GHz up to 10 GHz--and spreads the signal across the spectrum with a very low power level. It won't cause a lot of interference but has a significant benefit of bandwidth.

Why will UWB succeed where Bluetooth hasn't?

Bluetooth hasn't taken off because of cost. Things like keyboards and mice are so cheap that it's hard for people to stomach the cost of inserting wireless. Again, the cost will have to be worked out, but this opens up more functionality. I would say UWB is more of a 2005 play, but you'll hear about it a lot in 2004. I could see it being one of those technologies that might make headway in the consumer segment first.

What will we see in displays?

More and more flat panels. Contrast ratios have gotten better so that flat panels are as good as CRTs. There will be a crossover point in the near term where flat panels pass CRTs. The only thing holding that up is the cost.

Organic light-emitting diodes are a very promising technology, but there are significant hurdles.

What's in store for storage technologies?

The other technology that goes hand in hand with PCI Express is Serial ATA. (It's) a new interface for storage . . . devices that now transfer data at higher speeds than the older ATA-100 parallel interface provides. That helps with reducing the cabling nightmares inside the box and enabling better airflow and better performance.

We've not yet begun to offer Serial ATA drives on notebooks. On the corporate side, we're still shipping IDE drives for the most part, but we'll continue to increase Serial ATA drives we ship in the corporate space over time.

In 2005, you'll see the Blu-ray optical disk formats. The transition to Serial ATA helped with performance in enabling that to move forward. HP and Dell have joined the Blu-ray Disc consortium. This is a great follow-on with what we've done with DVDs. You now have a medium that's capable of storing 20 to 25GB of data on a single layer.

As processing power increases, how is Dell dealing with thermal issues?

Thermal challenges are big in both desktops and notebooks. People are demanding lower costs, so we've seen the adoption of desktop CPUs in notebooks, and the issue is cooling those. We're trying to target new designs that can cool 50 percent higher thermals. It's technology that's more of the same--traditional heat sinks with stacks.

In the future, things are going to be changing to where we're looking at refrigeration. Step 1 is we're at the limit of what we can do with forced air. Step 2 is liquid phase-change technology. Step 3, once you're talking about really high thermals several years from now, we'll start looking at refrigeration.... Active cooling solutions are under investigation at Dell and are being worked on in our labs.

What do you see coming in the areas of power management and battery technology?

Suppliers are doing more and more work with active power management, including stuff in graphics chips where part of the chips shut down when not in use.

The battery engineers at Dell tell me there are technologies dealing with the organic materials in lithium ion that will allow us to yield about 30 percent to 40 percent power-density improvements. But beyond that, we're looking at fuel cells and wondering if that's something that can be manufactured and can be accepted.

How far are we from integrating wireless LAN and WAN technologies on the client?

The Holy Grail is to get to what we call smart radios, or software-based radios, where you have an integrated cellular and 802.11 solution inside the box, and it switches automatically. That's what we hope the industry drives to in the future. It's a ways out, but that's what we're trying to direct our resources to.

Do you support industry efforts to embed security features into system hardware?

There is a Trusted Computing Group standard to have an identifier with the desktop so you can identify the PC or desktop. We've tended to go with identifying the individual, which we believe is more important, although both come into play in a total secured environment. Our first step was (offering a) smart card in our notebook line. Then you could look to extend that into a locked-down piece of hardware ... but we prefer the route of identifying the user.