“Speed Counts, Mobility Counts Even More”

Paul Otellini, President & COO, Intel Corporation is designated to take over as the fifth CEO of the company in 2005. He would inherit a company that has a legacy of success and paranoia. On his first visit to India, a country, which the current CEO Craig Barrett has, visited thrice over- Otellini calls it as a ‘learning trip’. In an interview with Cyber News Service, Otellini shared his views on what Intel is working at. According to him, while extending Moore’s Law, "Speed does count, but mobility counts even more". Excerpts from the
intearaction:

Intel wants to be a provider of building blocks to the Internet economy. Intel has invested in huge sums to get a foothold in the networking and communication segment. Yet, in these spaces, Intel is mostly a secondary or marginal player. It’s not really up there, so are you going to be a serious player in that specific space or is it just the kind of technology to suit your mobile platforms?

Paul: We focus on most of the companies, if you really look at them very carefully. We have found a number of architectures that are now beginning to sell. Some of the work is being done with Intel’s Centrino. As we build upon the technology, it is getting better and better and we are integrating technology into our platforms, chipsets, microprocessors. We want to be the world leader in Wi-Fi products.

In the wireless chip space, Intel has many aggressive competitors like Broadcom and Intersil, unlike in the desktop segment. Given that Intel’s performance over the last few quarters on the wireless chip has not been so positive, what are your strategies to tackle competition in these segments?

Paul: The fact is- everyone is aggressive in Wi-Fi. It is a pervasive technology. But to get my comments- there are 27000 access points being deployed and installed every day, 1 in every 3 seconds. And there has been a sharp reduction of cost to make it very cheap. As we look at the competitive horizon, our opportunity to use the knowledge and having married with the communication platform, gives us a long-term advantage over the discrete wireless members who don’t have computer knowledge.

What’s your forecast for semiconductors- Intel and the industry?

Paul: Semiconductors? I am not going to give you an Intel forecast! We never forecast numbers like that. Our first half this year is slightly lower than our first half last year. The industry, as per IDC, can look at most of the forecast as 6-10% growth in the IT industry. We should be in that range but I am not going to argue with the numbers.



Intel plans to get involved in the making of a wireless environment in a big way. There are reports that there is a $500 million plan for promoting Centrino deals with operators, hotels and hot-spot application developers. Can you confirm this and can you also elaborate on some of these plans?

Paul: It’s mostly $ 300 million, not $ 500 million. Yes. It’s a very large plan. Part of it is for a worldwide advertising campaign, part of it is for working on co-operative fields with airports, hotels, and airplane manufacturers worldwide, and validates the hotspots so that they work well with Centrino. A part of it would be used as venture capital to give investment to small companies that do work in the area of accelerating Wi-Fi rollouts. The sales team here are trying to boost the environment of Wi-Fi take-off. We are driving high performance network intelligence where we can look at every bit and look at the virus infection on it and put intelligence on it. That is what is taking Moore’s Law to network.

Speed (as in GHz) vs Mobility (as in connectivity)- the mantra has changed. Is there a Moore’s Law supporting the connectivity paradigm?

Paul: Moore’s Law is good here too. While Moore’s Law started with the sheer number of transistors packed in a chip, over time the progress was measured in terms of Ghz. So, if you go back to transistor count, mobility is very much a part of that. It is not about just driving the clock speed overtime but adding functionality: now mobility is more important to us. Intel develops more and more features that are architectural to take advantage of the transistors. Whereas features like hyper-threading on the desktop, doesn’t require a lot to be added on the transistors front, but does double up the performance of the chip. What I mean is that it is not only about transistors or speed. Architecture counts too.

IBM, Oracle, and Microsoft have pledged support for AMD’s 64-bit chip Opteron. How do you view that? What are your plans on taking the 64- bit on the desktop given that AMD and Apple are touting that they can deliver 64 bit computing to desktop users by the year-end?

Paul: We have been working on the 64-bit architecture for almost a decade. We are about to introduce our third generation silicon as the world’s fastest computer, as for transaction processing. It’s only going to get better and we will pick that up to be a factor of 10 over the next two years and the architecture is much better in terms of scalability. There’s a lot more to enterprise computing, supercomputing than 64 bits but its back and we are very happy with where it is and where it is going.

There is really no need to have 64 bit on the desktop- at least in the foreseeable future. No other application takes advantage of it. One exception would be workstations and we have Itanium there. At some point of time, applications and OS on desktop will require 64 bit addressability and memory prices would come down enough so that you can afford 64-bit at your desktop. And at that point of time, expect Intel to be the player.

Initially, Intel was pushing 802.11a instead of 802.11b whereas b is a cheaper technology. What is going to be Intel’s strategy for pushing 802.11g?

Paul: Our intention is to provide all the standards into the offering as fast as possible with high performance and low price.

What are the prospects for Intel’s communication chips business?

Paul: As I said, we are doing very well in growing the Wi-Fi industry. The design activity is going very well. The problem is the industry is not matured yet. The handsets with large cache memory with the intelligence inside based on the PCA architecture is increasing the logic inside the handset. Handset technology is growing and I expect it to be a larger player there.

Cyber News Service