SGI: A "Much Simpler" Company
Thursday September 5, 10:09 am ET
BusinessWeek Online
Technology: NEWSMAKER Q&A
Silicon Graphics (NYSE:SGI - News) was one of Silicon Valley's computing pioneers. It invented much of the visualization and graphics technology used today by the Defense Dept., Hollywood, and the medical industries for manipulating vast amounts of complex data and for working in graphics-rich computer environments.
But the Mountain View [Calif.] tech company has slowed down in recent years, making what analysts judged to be some ill-timed and unwise bets on graphics technologies as well as a few very expensive acquisitions that didn't pay off. SGI's chronic problem has been that it tends to create new standards in high-powered computing, only to see those businesses deteriorate in value as competitors swoop in and make the same technologies at a fraction of the price.
Now, SGI has new management, led by CEO Bob Bishop, a company veteran with 16 years in sales and marketing. But even after several rounds of spending cutbacks, SGI remains on squishy ground. True, its $1.10-a-share range is a big improvement from the 31 cents it was fetching after September 11. But that's little comfort to investors who bought at more than $40 a share in late 1995. Even after an impressive stabilization effort by management over the past year, many investors still believe SGI won't be able to pay its nearly $230 million in debt that's due in two years.
Not to fear, says SGI, a turnaround strategy is in the works. It continues to pare expenses and hopes that a few smart bets in its traditional markets can bolster revenues. Recently, BusinessWeek Online Reporter David Shook spoke with SGI Chief Financial Officer Jeff Zellmer and Senior Vice-President for Marketing Jan Silverman. Here are edited excerpts from their conversation:
Q: The investment community seems to think that the company may make it. What should investors know about SGI today?
Zellmer: I think investors were betting last summer, as opposed to now, that we wouldn't make it. Then, our stock hit an all-time low of 31 cents. We had some serious problems. But the company has strengthened itself. We've reduced operating expenses by 40% year-over-year, gross margins have increased 10 percentage points. We've been operating-cash-flow positive in each of the last three quarters, and our unrestricted cash has gone from $120 million a year ago to almost $220 million.
The issue now is debt. It's about $230 million -- due September, 2004 -- that's trading for roughly 55 cents on the dollar. We're working with advisers to devise the optimal strategy, and we hope to better articulate the strategy over the next several quarters. A potential option is to buy some debt back at a steep discount. We could do a debt-for-equity swap, but we're not high on that option now because of where the stock price is.
As we strengthen the company, we may do some form of refinancing -- extending the term or perhaps raising the interest rate. There are a number of things we can do to make the company more attractive to investors. And we have some time.
Q: Given that you don't have the financial capability to pay off the principal on that debt today, and it's due in two years, is bankruptcy an option?
Zellmer: I'm not planning for bankruptcy at all. We've done everything we needed to do to solve the immediate liquidity issues. And we did that by doing some strategic transactions with Microsoft (NasdaqNM:MSFT - News) and NEC that unlocked significant value in the company that was either not on the balance sheet or not recognized in the stock market.
Last summer [2001], our entire company was valued at about $100 million in the stock market when NEC agreed to buy 60% of our SGI Japan sales and services unit for $95 million. Similarly, Microsoft bought a portion of our intellectual property for about $62.5 million.
We also closed a plant in Switzerland. We collapsed two separate sales forces for government and commercial sales into one unit. We've pared our workforce down to about 4,200, from 6,800 in March, 2001. We didn't do toxic financing or sell stock real cheap or sell assets at bargain-basement prices. We raised money by demonstrating and unlocking real value in this company that the market wasn't recognizing.
Q: What are some of the initiatives under way that are crucial to the turnaround?
Silverman: The computer industry right now is riding a data explosion. Total online data storage in the world is roughly 8 exabytes [or a billion gigabytes]. In two years, that number will jump to 50 exabytes. No area of tech is growing faster. Our systems are designed to handle huge amounts of data. It might be a seismic analysis of an oil field, new enhancements in the handling of MRIs or CAT scans for radiologists, oil-and-gas exploration, or the digitization of film for Hollywood studios.
To give you a sense of how fast the data-storage business is growing, last year we had a client in the energy business that did a seismic analysis of an oil field that resulted in 2 terabytes of data. This year, a similar survey might yield 24 terabytes as the engineers increase the resolution of the analysis. So the need for our computing is there.
Q: Why SGI and not IBM or EMC?
Silverman: We've introduced a storage solution that's ideal for many kinds of image-processing applications. The federal government uses lots of computers that need to access large satellite-image feeds very fast -- as if it's local. We've developed a special storage-area network [SAN] that's fundamentally an array of computer disks. Instead of attaching directly from a fiber channel to a computer, the machines attach to a little control computer that acts as a traffic cop -- it basically directs traffic through a fiber-channel switch.
The net effect is that any computer that attaches through this array can get the data much faster than if it were a standard network. All the data flows as one big data store, yet every computer that attaches to it accesses at a very fast rate. It's really a unique architecture. Nobody else in the business is doing this. Also, the data is only copied once from one location to the next in our architecture, not twice [once at each end of the file transfer], which is common in most networks.
Q: And you have a partner?
Silverman: We've partnered with Hitachi, which just bought IBM's disk-drive business and is giving EMC a run for its money in the storage marketplace. Hitachi is very aggressive. They wanted a high-end storage-area network, so they've licensed software technology from us.
Q: You also have a big push in Hollywood to help studios digitize films. Please explain that.
Silverman: No matter how film is stored, it deteriorates. So we're in the business now of digitally mastering film. Not only because they decay but because Hollywood now needs to generate multiple-resolution copies of each new film. When a studio shoots a film, it puts it in video format for VHS, a higher-resolution format for DVD, and then there's HDTV, which is now available in some areas. I can buy The Matrix on DVD, and it looks great. But when I see it broadcast on DirecTV on high-definition broadcast, it's awesome.
An even higher resolution is to film a movie completely in digital video, as George Lucas did with the latest Star Wars film [Episode 2: Attack of the Clones]. While standard HDTV is 1,000 lines of resolution, high-definition film video is about 2,000 lines of resolution.
So we're saying: Let's use our big computers to digitize film and then dump it into that SAN. Put the 4 terabyte master film file on the SAN, and then have editors do the touch-up effects, fix the color, add the Dolby digital soundtrack, or do other types of editing. Each person does what he needs to do, then puts the data file back on the SAN, allowing the next person to manipulate it.
Our SAN is structured to allow people to operate without stepping on each others' toes. Then, when the production is finally finished, the film can be spit out on whatever resolution format is needed. That's the market we're after.
Q: So the basis here is developing a way to make graphics-rich data travel faster across a network for any number of industries -- from film production to energy exploration to government entities such as NASA or the Defense Dept. And you feel this will make a successful turnaround possible?
Silverman: Today, the company is much simpler. It's focused on three core competencies: advanced visualization, complex data management, and high-performance computing. We think that's a good strategy. We've been around for 20 years, and we plan on being here for a long time.
biz.yahoo.com |
