End of the Line for Silicon Graphics


For some, it may be the end of an era -- for others, simply the inevitable outcome for a technology past its prime.

The storied life of Silicon Graphics came to an end this week with the announcement that Rackable Systems is buying the company for a paltry $25 million. This is the same Silicon Graphics that cut a swath through high-end computer systems in the 1990s with innovative technologies that many believed were destined to trickle down through all classes of computing.


That expectation never materialized, however, as the industry instead embraced a trickle-up model -- low-cost commodity hardware combined with advanced networking into a framework that was at once more flexible and scalable than high-end systems and could be purchased and expanded over time rather than in a single, costly acquisition.

Silicon Graphics' problems were highlighted in the Chapter 11 filing the company made the morning the buyout was offered: assets of $390 million with debt topping $526 million and an impending default on about $160 million in loans from Morgan Stanley. The company also reported losses of $49 million in the second quarter alone, with the stock price -- once a Wall Street darling -- scraping the bottom at a mere $0.25 per share.

Rackable no doubt plans to use Silicon Graphics' technology to boost its fortunes in the high-end enterprise market. Although it was late to the commodity hardware game, Silicon Graphics did manage to build a respectable portfolio for the x86 market, not to mention clustered storage and even cloud computing technologies.

Still, Rackable has its own financial troubles to worry about. The company has posted a string of quarterly losses of late, and has cancelled a stock repurchase plan worth about $40 million.

The bigger question in all this, however, is what it portends for the future of high-end specialty hardware. As Maureen O'Gara points out here, Silicon Graphics' fall comes amid swirling rumors of the impending demise of another former high flyer: Sun Microsystems. It would be ironic indeed of both companies that were founded in the same year, 1982, were to check out together as well.

In all likelihood, high-end specialty hardware will continue to have fans in scientific research and large transactional operations, but those opportunities are likely to become rarer as technology advances. There's a reason why more and more organizations are turning to massively dense commodity environments: they provide the necessary computing power for less cost. As fewer high-end installations come online, costs will only increase to compensate both for the flagging economies of scale and the tremendous amount of research needed to keep on the cutting edge.

It would be a shame if development on advanced systems were to come to a halt entirely because that research still has enormous potential to benefit all users.