Bishop takes the heat as SGI mounts recovery effort

This is how SGI chief executive Bob Bishop jokes about attacks he took from investors and Wall Street for streamlining the company during the past year.

Once known as Silicon Graphics, SGI morphed during the year from a company selling supercomputers, large servers, multimedia software and embedded chips used in consumer devices into one catering to just a few markets: technical and scientific computing and digital content creation.

Along the way, Bishop shed four divisions, including the well-known Cray Computer operation, and advanced SGI's aggressive push into Linux. But the changes meant a continuous string of one-time charges and uncertainty about SGI's future. The company's stock price plummeted, and Bishop had the misfortune of delivering quarter after quarter of losses.

Bishop took the hot seat Aug. 23, 1999. With his first year at SGI's helm behind him, new products to market and the start of a new fiscal year, Bishop is optimistic, feisty and even psychic. He told CNET News.com's Joe Wilcox that he expects a return to profitability in fiscal 2001.

CNET News.com: You took it on the chin from Wall Street over the streamlining.
Bishop: I decided to take all the hits. I knew when I came to this job I'd have to take the hits, and I took the hits.

All of those steps had consequences: a number of one-shot charges in the books as we do this. The financial community had to look closely at what was going on, because it affected our tax position, affected our revenue position, head count, cost of goods, gross margins, every number.

I spent a year trying to explain this. I'm actually happy that in our fiscal 2000, which ended June 30, we really got through it all. I'm just delighted we're off to such a fresh start in fiscal 2001, and we have the entire product portfolio to the market.

Those steps you took involved cutting loose four divisions. What can you tell me about letting Cray go?
I divested four pieces to get simplicity, clarity and full alignment. That's why I did it.

We started by divesting Cray, so we're out of the vector (computing) business. We could not marry...a vector and parallel (computing) architecture. There's a religious difference, and we rightly concluded this marriage doesn't work, so we divested the vector piece and kept 800 Cray people and the lion's share of the technology. We chose the pieces to integrate with the parallel architecture.

We're quite happy with what we did. And now the 800 people from Cray that stayed with us are highly motivated SGI people. There's no religious difference anymore.

You also cut loose your MIPS chip business and some share value with it. Why?
We're 100 percent committed to the MIPS-Irix product line. We had built an embedded chip business, embedded into consumer products. It was highly successful; there were approximately 50 million chips per year being shipped into consumer products.

For example, Sony PlayStation...and...Nintendo 64--they all run on MIPS. Laser printers, automobiles, and all different products. But it did not fulfill our mission, our sweet spot, focused on the technical computing user from a system point of view. So we divested that.

Keep in mind the designers for the high-end MIPS are still in our company. And we have a road map to take MIPS from 400 MHz today out through 1 GHz in the near future.

Where are you now after the divestitures?
The consequences on the remaining part of the company--about 6,800 employees--is a highly focused, streamlined operation focused on a single mission. And the sweet spot of that single mission is the technical communities.

You are now focusing on technical and scientific computing and creative content, such as film production. But haven't you backed yourself into a niche market?
Our revenue is only about $2.5 billion, but we estimate that market to be $25 billion. We only have 10 percent of that market and lots of room to grow.

You've had a string of unprofitable quarters. Do you expect a turnaround during fiscal 2001?
We've given guidance to Wall Street at our last call on July 20 that we would be profitable for the full fiscal year 2001 and that we would be growing the revenues somewhere between 15 and 20 percent. That's the guidance we gave Wall Street, and we're on target to do that.

So you are 100 percent confident the rough times are over?
We did the turn. Emotionally, the company's turned. And it's getting runs on the board. Getting runs on the board is pretty important to Wall Street. You can't change their game. That's their game. We're back. We're back quite strong. The products are great, and the motivation in the company's quite high right now.

What's one of SGI's biggest failings?
We've sown the seeds of innovation but failed to harvest them. Someone else has come along with stronger marketing capability or putting in lower-cost systems and found a way to harvest our work.

I'm anxious to learn how to harvest more of what we sow. We are truly respected as the central innovator in this space, but we truly have not harvested volume, nor profits, from this effort.

In July, you unveiled the Origin 3000 server and pinned much of the company's future on the product line. How is it doing?
We had $100 million in advance orders and feel strong about what we have to offer. These new products are mind-boggling. These new servers have got attention. Now we've got to deliver these new products in volume, and we are dealing with a number of supply constraints in the industry, which are affecting everybody at this moment.

Component shortages have hit everyone hard. How are you dealing with the problem?
It's amazing how this problem is permeating the industry. We're clawing our way through it. It's a day-by-day struggle with our vendors, but we've got most of our componentry pinned down. We don't have as much upside as we want. Because we are at a point of takeoff and our booking stream is quite strong, we would like to be able to deliver more rapidly.

I understand you expect Linux to account for half of SGI's revenue within five years. How does your Linux strategy interact with your proprietary technology?
We offer a full line of Linux products, but only on Intel (processors). We don't offer the crossover. We're not putting Linux on MIPS.

When Intel has its Itanium processor on the market, we will soon have a full range of IA-64 Linux products. The architecture is already done. The subsystems are already done. It draws upon what we've done on the MIPS-Irix side, so we don't have to redevelop this.

So we're ready to go, and we think Intel will have the chip ready by the end of the year. Then our systems will follow probably in the March time frame next year.

Linux is gaining momentum, but don't you think your revenue plans are aggressive?
No one can truly predict the potential for Linux, but our scope of it is that it's, firstly, tightly linked to the Internet world. I think today it already occupies probably 25 percent of the servers in the Internet space.

We're very focused on broadband Internet. Our servers we're looking to deploy in a broadband Internet space, where scalability and throughput become imperative. We're not talking about alphanumeric data, but visual content, interactive content, immersed content and continuous changing content. That's the world of broadband Internet.

So you're talking about leveraging your technology used in creative content areas, such as film editing and digital film creation?
Yes. And not just old-generation movies, the archive movies. We're talking about new-generation movies that have an interactive content component, where the viewer is immersed and has some dialog with the characters and plot.

We haven't yet seen the Internet next-generation, new genre content. But if you had to compare it to anything, it would be the video game environment extended over the Internet with multiple players. That kind of content is very sophisticated, involving high-speed access to big databases. We're ready to deliver that with our new architecture.

What kind of business opportunities do you foresee in this market?
We think media will be sold and transacted, and instead of e-commerce we really should be talking about media commerce. That transition is under way because of the emergence of broadband Internet. We're gearing ourselves for broadband Internet. We see our new architecture fitting into that space, and Linux will be the gateway to broadband Internet.

Do you plan to do more with marketing? Say, promoting your technology's use to make successful movies?
We have come to a kind of rebirth, the relaunching of the company. We have committed some pretty significant funds to global promotion. For example, it's on our platform that "The Perfect Storm" special effects were developed. It's on our platform that the whole "Space Cowboys" simulations and special effects were developed.

I can go on discussing this, whether you're talking "Toy Story" or "Jurassic Park." It's our stuff, and we have not taken advantage of this as we should have.

For example, "The Perfect Storm" is a global phenomenon. It's a cult. And the animation and special effects are just phenomenal. We should be riding along, and our promotion should match what we've done for those customers.

If you had one thing you would like to communicate to your customers, competitors and Wall Street, what would that be?
We're back.