AMD's Hester touts PC, graphics convergence

Company's chief discusses hardware advances driven by increasingly powerful software, the home media center, and more

AMD had a great year in 2006, using its success with the Opteron server processor to win greater respect from users -- and troubled rival Intel. But tumbling chip prices and a new crop of Intel products have put new pressure on the company. On Tuesday, AMD's Chief Technology Officer Phil Hester talked with IDG News Service during the International Consumer Electronics Show in Las Vegas. He spoke about trends in new technologies he was seeing on the show floor, the progress of AMD's acquisition of ATI, and the new technologies that are needed to deliver the long-promised goal of sharing Internet and television content between devices.

IDGNS: I was talking to your colleagues at ATI about the TV Wonder [a digital cable tuner] and your own initiatives with AMD Live, and it seems like you're both working on similar approaches. Is there overlap between the platforms in areas like content security or Vista?

Hester: My perception is that video content owners have learned a lot from the audio content owners, and there's a lot of concern and focus on the whole issue of end-to-end content protection, particularly when it comes to bringing HD (high-definition) content. Standard-definition content people are pretty comfortable, but the issue is with premium content. We'll see a big improvement around DRM (digital rights management) with Vista. Combined with a PC, it will make the set-top box a more secure environment, so now with Vista, you've got a platform for premium cable. It has a natural ability to do DRM and use 3D graphics for the user interface.

Essentially every client you see now is going to have a decent level of 3D graphics capability. If you went back to the early PC platforms in the mid-1980s with the 286 generation, you had an open socket next to it for the 287 math coprocessor, and at that time, a very small fraction of the population ever bought that math coprocessor because the applications didn't demand it. The same thing happened with the 386 processor, although you had 20 percent of people using it. And then the 486 integrated that as a standard part of every microprocessor, driven by the fact that the operating system and the applications demanded it.

I think the same thing now is happening with 3D graphics. If you look at the things driving that, whether it's business information rendering or entertainment, by the 2009 time frame, just about every PC is going to have to have some level of 3D graphics. Not just the top end of the enthusiast market, but it is going to be a standard PC experience without having to incur the cost and efficiency hit of using a separate graphics card.

If you look at the [CES] show here, to me this is kind of a proof point -- there's a convergence of traditional consumer media streams in the PC. You go out and look at all the 1080p HD TVs and the interactive multimedia technologies that are present and all the capability of some of the gaming platforms, you recognize that the PC was never really architected to deal with that. So over time, GPUs -- graphics processing units -- have been added to the PC to deal with these new types of data. But if you think about these new types of data becoming just as important as the standard data in the PC, you'll see you have got to rethink processing elements, taking the best of the information processing ability of the x86, 64-bit architecture and marrying that with the graphics capability of GPUs.

So in this new world that you're seeing, which is the merger of data streams, you really need a new class of microprocessor. We've been talking about that as an accelerated processing unit, an APU. The class of processors that effectively merges the ability to deal effectively with both types of data streams are fusion processors that are optimized to deal with both traditional x86 64-bit instruction sets and also this new multimedia, graphics-rich set of information that will exist, largely driven by the transition around Vista and in the gaming world, with essentially every game now being designed with 3D graphics.

IDGNS: When those chips are ready to be used in consumer electronics, what other technology changes need to happen to enable the vision of marrying content on the TV and PC?

Hester: We don't have a religious view about what is the media center of the home. I think people will tell you there are three centers of the home, depending on your personal views. One is certainly the PC, another one could be your set-top box, the third could be your gaming platform, and more and more over time, it's also the portable device that you carry with you. So we're really not focused about telling people which one of those three host platforms would be best for acting as a central device. We're more concerned about providing a set of applications and services that allow all those devices to interoperate as easily as possible.

IDGNS: Next year when we're here at CES, will we see hardware coming together as well?

Hester:  Yes, you'll see elements of it coming together. The first fusion CPUs will be available in prototype in late 2008 and in production in early 2009. And we've already, through discussions with our customers, reached agreement that the best place to initially focus this is the notebook space. That is where you get the most benefit of the efficiencies from a power standpoint and also from a physical area standpoint."

IDGNS: Looking around the show today, how do you see the progress report in moving toward the TV-PC convergence?

Hester: To me this is the proving point that it's happened. If you go back four or five years when you had Comdex and CES, the computer folks were pretty much persona non grata at the consumer show. But now if you walk the floor here, you'll see it's kind of an equal mix of consumer technologies that are incorporating PC sorts of things and PCs that are incorporating consumer technology. A lot of these 1080p HD panels now have computer capability that's analogous to what was previously in a high-generation graphics card, and you look at what's happening in some of the things like the ATI TV Wonder TV adapter, now you've got things that function like a consumer electronics TV that are living in a PC and are getting that PC to do what you or I would have considered the job of a set-top box. And then you look at the capability of some of these cell phones with the Imageon chipset, and they can shoot pretty good quality real-time video and do digital signal processing and image stabilization, and they're doing some of the things that a PC could certainly do, but you'd never have thought a cell phone would be capable of.

IDGNS: Is there anything else that has to change to make this really gain traction in the marketplace, like better price points? Because vendors have been promising some of these capabilities for a year or two now....

Hester: Yeah, I think simplicity. People want a turnkey appliance that doesn't have a thousand-page user manual, so we have to do this stuff under the covers. Certainly price point's important, but if you look around at cameras and the dropping price point of a gigabyte of memory, you can get, I think, a very powerful system today at a price point that is very affordable in most developed nations. In developing nations, you'll see a focus on cost, but there is an emerging class of device that will be better suited to our initiative called 50 by 15, where we want to get 50 percent of the world net-connected by 2015. And in a lot of those places, to be honest, the PC is overkill. In many cases, they want to get information on their cell phones. In certain cases the subsidized business model could be the best way to do it if you'll tolerate a 30 or 60 second advertisement when you turn on your PC. So you'll see the emergence, I think, in a lot of countries, of different ways of delivering information technology, so a different purchase model can address their needs.

Copyright © 2007 IDG Communications, Inc.

How to choose a low-code development platform