By Tony Waltham
Thursday, June 14 2007 11:22 AM
URL: http://www.zdnetasia.com/news/hardware/0,39042972,62021044,00.htm

Moore's Law, which says that solid-state transistor density doubles every 18 months or so--resulting in faster processors and denser memory as well as lower costs--is good for at least another 10 years, according to Intel's regional vice president for sales and marketing, John Antone.

In Bangkok for Intel's Centrino Duo and Centrino Pro platform introduction last month, Antone explained in an interview how the next one billion computer users would be enabled--one village at a time--with a combination of cheaper hardware, network access and programs that engaged governments and shaped their educational curriculum in particular.

Speaking of a reinvigorated Moore's Law, named after a prediction made by Intel co-founder Gordon Moore over 40 years ago, he said that Intel's latest 45-nanometer transistor had a gate that used a new element called hafnium that significantly improved the electrical characteristics by cutting down leakage current.

Historically, transistor gates have been grown out of silicon dioxide by exposing silicon to oxygen, which was where glass came from, he said, adding that glass was an insulator but that it had been getting harder and harder to grow really fine glass.

"So in these 45 nanometer transistors we actually deposited hafnium...It's more like a metal, so the switching capabilities are significantly better than the capacitance associated with silicon dioxide."

Antone said that Gordon Moore himself had called this "the biggest breakthrough in processor technology in 40 years".

"We've demonstrated the 45 nanometer, which will begin production in the second half of this year, we've talked about and shown pictures of 35 nanometers. I believe the we've put a road map out for 10 years that goes down below 10 nanometers," he said.

"When we start falling behind with lithography, then we'll start running into challenges with Moore's Law. But we can see our way through at least 10 years...if you can see it, you can build it," he said.

"The MIPS are going to go up, the MIPS are not going to stop, in the graphics engine or in the CPU engine, and memory...whether it is DRAM, XRAM or Flash memory, it's going to be under Moore's Law--so pick your point in time and we can do a lot more with 10-times all these things for the same price," he added.

Speaking of the incredible changes the industry had seen over the past four decades, he said the only thing that paralleled the amount of change would be the environmental and economic changes in Thailand.

Most of Southeast Asia had been like Thailand, he said. "The Philippines probably has been a bit more spotty. Vietnam is definitely on the rise...incredible, where are we going to stop?

"India is nothing short of chaotic, but the growth and the change is still incredibly dramatic and that's a billion people. China is not monolithic. After you get past the coastal cities, if you keep going either north or inland, there are hundreds of millions of people that don't enjoy the same lifestyle.

"And yet the Chinese are committed to paving their way and having networks out to all those sorts of places, so we really are in the process of taking a billion people into the modern world--or five billion people into the modern world--and the technology has got a hand in making all that happen, and that's exciting to talk about as well."

Antone spoke of how, with programs such as Intel's Teach to the Future and International Science and Engineering Fair, the company had been working on "bits and pieces of programs for 8 or 10 years" and he said that Intel had now collected them together and they had become a lot more formalized. "Governments have given us a lot of feedback on how we've formalized the program," he said.

Observing that there had been a lot of interest in Nicholas Negroponte's "One Laptop per Child" US$100-dollar notebook computer project, he pointed out that inexpensive devices were only part of the problem."You can give a device away, but if nobody uses it, you haven't solved any problem, right?"

While the device itself needs to get cheaper, Antone said that we needed to have networks available as well, and the network access had to be cheap. "Teachers need to not be afraid of the device, because once you get a kid on the Internet, he'll sort of self-teach, and go his own way. And there is nothing scarier to an educator than a kid going his own way.

"So it needs to be incorporated in the educational system in a constructive way so that teachers embrace it and use it versus saying 'this is bad, I'm not even going to give the kid a chance. I'm going to lock it in the closet because, the minute he has it, he's going to go out of bounds, I lose control, I don't like it."'

The Intel general manager for Asia Pacific said that there were ways to do

Tuning in to the next billion...

this by integrating it into the class in a "synergistic" way. "It doesn't have to be scary, but that's where our Teach to the Future program was meant to say 'let's show you how you can incorporate this and make it a big part of your class. You don't have to be scared of it, but you need to use it and it needs to be part of the curriculum'," he said.

The notion of developing a $100 device or subsidizing or giving away a $100 device was just one element of the problem, but not a solution to the problem in its entirety, he said, adding that Intel could envisage such devices now costing around twice that price, or slightly more.

He was speaking in advance of Intel's announcement made last week at Computex to build inexpensive notebook PCs with Asustek Computer, but he did outline some of the likely specifications of these devices.

"We do see our way toward, probably, $200 devices... $250 down to $200 devices, which we think will get adolescent kids onto the Internet in a meaningful way. At that cost level, where you start making trade-offs is in the size of the screen, the size of the keyboard, and the type of memory you have in there," he explained.

"So, at those price levels, it's a seven-inch screen...A seven-inch screen is pretty small for anyone whose been out there, but for a kid who's never been anywhere, it's a good starting device. A device that doesn't have a hard disk--you'd throw it out the window in about 10 minutes--but for a kid who's never been out there, it's a good place to start. These are some of the trade-offs that you have to make to get down to those cost levels," he said, while also noting that the corollary of Moore's Law was that every year hardware got cheaper.

"When you and I think of the next billion people, we think of it sort of as a number on a spreadsheet, but that kid on that next village doesn't relate to himself as being part of the next billion. He just sees himself as being the next village that isn't quite connected yet and he just wants to know when someone's going to get around to helping him out.

"So it's less about a billion and more about tens of millions of villages, one village at a time, that need to be enfranchised," he said, adding that the device itself was probably the easiest of all of those elements to get right because Moore's Law was going to take care of that over time.

Intel believes that WiMax could help speed up network deployment because the technology would be that much more affordable to deploy step-by-step. "It's how many next villages can we get to next week. And it's more and more next villages every week," Antone said.

"There's definitely a feel-good element about, if you think about a horizon or a frontier, seeing that frontier slowly push open and doing things to move it open faster...it's a good problem to work on," he said.

Intel was in the center of Moore's Law to bring the costs down and the company was taking the influence it had and "turning that into government relationships and telecommunications relationships. And then taking those over time sort of to herd people in a common direction. That's the other thing that we do really well, I think," he explained.

"As we picked Wi-Fi four years ago, we can corral people in the same direction and make a substantial change in the world. WiMAX, we're going to do the same thing. Our World Ahead program and getting people to think beyond the device and talk about the whole solution. I think we're going to make good progress there," Antone added.

Another area was taking the PC and embedding it in cars, phones and televisions, he said, noting that Intel had driven all kinds of work to get standards and to make devices interoperate, which he likened to plumbing, and which was hard work.

"There's a lot of the next billion (computer users) in Thailand, especially in the Northeast," he observed, noting that "the one final thing that does become a bit of a challenge to really grow broad" was language. With the basic Internet, "kids would get on and they would probably learn English because of it, but to get the existing population in, to get the culture, we need local content," he said.

"That's the final piece in making the Internet robust. It's got to get integrated into the local culture, the local language," he said. Local language content would be critical to making the Internet "the central part of a village society up in the Northeast, or way out on the west coast of Vietnam, or up in Laos, or in Sri Lanka".

"To get kids on, you need the network, you need the low-cost device, you need the education system and a government that's willing to integrate it. And so, when we're talking about the next thing, we want to talk about all these elements."