L.A. Financier Reemerges as Key Player in Cray Deal.

When Tera Computer Co. announced last week it was acquiring Cray Research from Silicon Graphics Inc., it released a six-page announcement without Terren Peizer's name in it.

But the deal would have been unthinkable without him.

The former Drexel Burnham Lambert junk-bond star largely has shunned the limelight since testifying against former boss Michael Milken a decade ago.

But his West Los Angeles financing and consulting firm, Intellect Capital Group, has a growing portfolio and reputation for viewing opportunities that others might not perceive -- in this case, the supercomputing industry.

"Terren Peizer came along, stepped back and said there might be a way to jump-start this industry," said Louis Mazzucchelli, an analyst at Gerard Klauer Mattison. "And he has the financial wherewithal to pull off this sort of thing."

At first glance, the merging of Cray, the most recognizable name in supercomputing, with a small, unprofitable competitor of sorts may seem surprising. But Peizer is betting his years of success in investment banking that there is unrealized potential.

Last June, his firm took a 10 percent stake in Tera for an undisclosed amount and set about shoring up its finances. The Seattle company hasn't turned a profit since at least 1995, when it went public. In its most recently reported quarter ended last Sept. 30, there was a net loss of almost $14 million. And since peaking above $17 in 1997, its stock traded as low as $2.88 last year.

The day before the announcement, Tera stock jumped from just under $8 to $11. On the day of the announcement, the company's share price hit a 52-week high of $11.50 before closing at $9.88. "Clearly, the minors were out there (that some sort of deal was imminent)," said Mazzucchelli.

Quick recapitalization

Within a month of getting involved with Tera, Peizer led an investor group that raised $30 million to recapitalize. In August, he was appointed chairman.

Soon after, SGI made it known it was interested in divesting itself of Cray, which it had bought in 1996 but never managed to align with the rest of its core business. At the time, Tera wasn't considered a prospective buyer.

"When we heard that Cray was out there, (Tera President) Jim Rottsolk said to me in jest, 'How about we go after Cray?' And I said in all seriousness that I'd be more than happy to finance the transaction if we could make it work," Peizer said.

That took some doing. SGI didn't take Tera's initial contacts very seriously, Peizer said. And even though Cray had problems, its business dwarfed that of Tera.

After finally acknowledging that its suitor was serious, SGI remained skeptical.

"The biggest concerns (on the part of SGI) were getting government approval and making sure we could close the deal," Peizer said. "When there are two companies, and the smaller guy is buying the bigger guy, there is a perceived risk."

But Tera had a couple of advantages. The government approved of the Seattle company's bid, seeing it as a way of resuscitating Cray, which had been faltering as of late after years of dominance. Supercomputers perform large simulations such as weather fluctuations and atomic tests, and Cray's customers include weather forecasting centers, government agencies and universities.

Tera was able to further assuage SGI concerns about its financial strength by raising another $30 million in December, bolstering its prospects.

"The $60 million that was raised over the last six months went a long way toward convincing SGI," Peizer said.

Fending off competitors

But hurdles remained. Among other things, Tera wasn't the only one interested in Cray. SGI had several competing offers. Last fall, there were published reports that L.A.-based Gores Technology Group, a little-known technology acquisitions firm, was close to buying Cray.

But Tera's work on new supercomputer architecture has been greeted with enthusiasm by industry insiders, and its perseverance in the industry was seen as a plus to SGI as well.

"They are one of the few companies that want to be in the vector supercomputer market," said SGI General Counsel Sandra Escher. "We think they were a good fit."

The money helped.

"We outbid the competition," Peizer said. "There were much more well-capitalized bidders. But cash is cash, and if you can demonstrate that you can pay the same cash as anyone else, it helps to be aggressive. We were going to do whatever needed to be done. There was zero chance I Was going to lose this deal."

Although terms weren't disclosed, the deal's value reportedly is under $100 million, or a pittance compared with the $740 million SGI paid for Cray in 1996. The reduced price reflects not only SGI's willingness to cut costs by getting rid of an asset it didn't really want, but also the recent slackness in the industry.

Peizer is convinced there is money to be made. Combining Tera's cutting-edge architecture with Cray's processing prowess will enable the new company, to be called Cray Inc., to reduce costs and expand its base.

"Supercomputing is a $2 billion (annual) market that is salivating for a new system," Peizer said.

Mazzucchelli puts the market closer to $1.5 billion, but agrees that opportunities exist for the new company to expand it. "It's a nice confluence of events," he said. "It's a good way to jump start Tera's plans."

As for Peizer, he would prefer the focus of the deal remain on Tern and its prospects. But he admits that Intellect Capital may find itself subject to greater scrutiny than before. The company has seven companies under its portfolio at an undisclosed value, and expects to announce several bigger deals later this year.

"We've been operating below the radar screen, which is the way we like it," Peizer said. "Unfortunately, this is going to bring us more to the forefront."