Intel & Otellini

Looking rested and fresh following Intel’s spectacular third quarter performance, President and Chief Operating Officer Paul Otellini addressed HBS students Monday on major technology trends, Intel’s changing business model, and the future of computing.

Empathizing with today’s tough job climate, Otellini commiserated with the students as he remembered his own graduation in 1974 from the University of California at Berkeley during another deep recession. He joined Intel, and after several years at the company he directed its marketing efforts to IBM. Otellini continued to gain experience, managing the Peripherals Component Operation before heading the Folsom Microcomputer Division. In 1988, Otellini was promoted to Operating Group Vice President and in 1989 he became Andy Grove’s (then President of Intel) technical assistant. Following several other promotions he was named President and Chief Operating Officer in 2002 and elected to Intel’s board of directors.

Otellini made some brief remarks and opened the session for questions.

He seemed proudest of the $300 million in savings Intel achieved last year through programs implemented two years prior to maximize efficiency and lower costs. These programs are expected to save $600 million this year and $1 billion in 2004. “These are real dollars, not just opportunity costs,” he remarked. In one example, the company was able to convince the only two silicon suppliers in the world to bid against each other for Intel’s business.

Wall Street has noticed also; Intel’s stock has almost doubled in 2003 (up 97.1% as of market close October 29). A student immediately asked him if he would still track those savings even if Intel didn’t sell the microprocessors that make it possible. “Yes,” he fired back, “for the ROI.”

Otellini identified two major trends that are altering the compute model: mobility and Linux. In the 1990s, networks were connected by a tangible plug. Today the norm is wireless. “We are the first generation where computing goes with us” rather than us going to work, and to the computer, every day. Computing today is becoming both seamless and ubiquitous, he said.

One example is the new 802.16 wireless standard for PDAs that gives radio towers a 30 to 50 kilometer range. Multiple towers could form a network that is similar to cellular, making cell to wireless transfers seamless. You could “wire up an entire state” that way, Otellini said.

Security for Wi-Fi was another student’s concern. Intel has been building encryption into handsets and increasing the “virtuality” of PC platforms, making them more secure. Virtuality is about addressing logical components, units, computers, or networks, rendering it more difficult for a hacker to find the absolute address for any part of the system. It also has a side benefit in that it allows components to be upgraded without requiring a full software overhaul.

Otellini acknowledged that it would never be perfect and that Intel gets hundreds of attacks each week and “some of them get through.” The student inquired if there would ever be a “Microsoft” for security, “even though it’s an oxymoron.” Much laughter followed.

The other “sea change” is Linux, an open source operating system which
allows microprocessor architectures to perform at the top 10 to 15 percent of current computing levels. More powerful PCs mean that applications run cheaper and faster, changing enterprise computing dramatically. Otellini estimated that 25 to 30 percent of the server market currently runs Linux.

Otellini was asked to describe his view of Microsoft and their collaboration together. “They are our biggest partner and we are both the largest consumers of each other’s R&D output,” said Otellini. “We have complimentary goals 80% of the time and compete the other 20%. I think that’s healthy.”

He then set the audience laughing saying, “Having said that, we are both very strong willed companies, and like any family, we have spats.”

One student asked how Intel sells into the various channels given that convergence appears to be taking hold. Otellini explained that originally each channel was approached separately. However, executives started to notice that the various groups were selling the same platforms to the same companies anyway, so in mid-October they went back to one unified sales force, even though they are “ahead of design processing.”

Another student inquired, “How do you balance the need for non-PC expertise with the desire to promote from within?” Otellini replied that he rotates middle to senior managers out of their areas to broaden their exposure, but, “I still like to bring in outside expertise” such as consumer marketing that Intel lacks.

Globalization ranked high in questioners’ priorities.

Asked whether Intel intended to keep its headquarters in the U.S. for the decade, the firm answer was “Yes. We are still an American company…we still think like and are responsible as an American company.” 60 percent of Intel’s employees currently reside in the U.S.

Offshore manufacturing was the subject of the next query, to which Otellini replied that Intel started manufacturing in Tijuana about 34 years ago just after the company was founded. Today, 65 percent of Intel’s revenues are from overseas. They operate two types of offshore operations: manufacturing plants that perform everything from basic assembly to high level engineering, and marketing operations. Opening a division in another country may sometimes be dictated by the local tax structure. For example, the EU had planned to tax Intel’s chips at a 35 percent rate so the company is opening a plant in Ireland. Similarly, Intel may choose to open a sales and marketing division in China because it is likely to be necessary in the future to avoid stiff tax rates.

The cost savings from manufacturing overseas are likely to be temporary, said Otellini. For example, Israel used to cost ten cents on the dollar per engineering man month but it has now risen to $1.10 on the dollar.

Shanghi, China now costs about half of what it costs to produce chips in the U.S., rather than one-tenth. “If you look at [overseas manufacturing] for low cost it’s a mistake,” said Otellini.

Asked if he foresaw a change in the value chain as a result of falling prices, Otellini acknowledged that Intel, Microsoft and Dell make money while “everyone else seems to lose money.” As pricing pressure continues in China, India and other emerging markets, computer manufacturers aren’t going to be selling systems for $499, but at $199 – $299 price points. The key, noted Otellini, is integration because it drives cost/functionality down. Intel enjoyed strong sales in emerging markets in the recent third quarter.

Interestingly, Otellini claims he never focuses on achieving a percent goal for gross margin because in high technology, margins become squeezed over time. Instead he tracks absolute dollars.

Citing limitations in the supply of LCD screens and the lack of R&D investment by both memory manufacturers and foundries, Otellini said that his greatest fear was a lack of capacity in an improving economy.

Finally, Otellini was asked about the next revolution – what would it be? “Well, the evolution is easy – convergence,” he said, but declined to make a prediction about The Next Really Big Thing. Advances that are here or just on the horizon include adding radio capability to Intel’s chips, increasing the degrees of virtualization in the enterprise or consumer environment, and continuing to advance silicon technology out until 2010 or 2012 – “but then you run into the laws of physics.”

Eventually everything may be run by organic computers, “but I’ll let you guys figure that one out. I’ll be retired by then,” said Otellini.