Bound to happen sooner or later. In my opinion, they’d been hanging on by their finger nails for the past couple of years. But it’s finally come to pass. SiliconGraphics (SGI) is no more. SGI was sold on April 1st – a fitting day if ever there was one – for a mere $25 million. When once it was a nearly $4 Billion computing powerhouse, it shrank over the years like an ailing old woman, forever hobbling to fetch the water and feed the cats until she finally just succumbed to the labor of it all.
I was there for almost five years from 1993 to 1998. It was a wild ride and for part of those years it was some of the company’s best. I’m proud of our accomplishments but at the same time I can only shake my head in an “I knew this would happen” sort of sadness.
In its heyday, SGI was involved in some fantastic technology. They got into so many different areas, it was hard for me as a technologist to keep up and stay current, especially given my role in sales support as a technology cheerleader. It had become harder and harder to understand all the areas they wanted us to sell into and be worth a shit at any of them, much less all of them. I saw the writing on the wall as early as 1996 when we bought Cray Research. They were already failing and I suspected some arm twisting by the US government to keep their technology afloat. We bought them, cannibalized them and threw away lots of good people. Then sold it years later for a fraction. We got into Nintendo – big deal. Nothing good ever came of that. We got into telecommunications and sold into Travelocity – that was fun. We made some great deals, lost some great deals and for me personally, it was a tremendous growing experience as I spent the last few year in management.
The biggest deal for me was when they decided to go after the PC market with a Windows NT system. They wanted us all to learn Windows (we were Unix heads) and become experts. Then they made a custom system that wasn’t even pure Microsoft. They fucked up their one attempt at mainstream computing. When I learned it was going to be a custom “Windows” machine I lost my faith in the company. I bailed before they could lose the rest of their company.
I kept up with several people and followed their careers. I worked with former SGI people at subsequent companies, too. We knew that SGI was a tough company to work for and if you could make it there, you’d do well anywhere. We trusted each other. I went to work for Platform after SGI – huge numbers of Canadian SGI people had already gone there. When that petered out, I went to newScale, another place where SGI people drew me in. I stayed there un September of 2001 when the bottom dropped out of the tech market.
My last business contact with anyone in SGI was with Harry Eaddy, the first sales manager I worked with in 1993. In July of 2006, I heard he was looking for another engineer to help sell. We had lunch, we talked and I researched SGI carefully. I was excited about getting back into the company I loved so much in the 90s, but after a careful reading of their story, hardware, software and plans I decided to bail. They were down to selling Linux systems and a handful of graphics beasts to the core market they originally sold to in the early 90’s – engineering firms. That’s when I made the decision to come back to Korea. 
And the rest, as they say, in history.
On a side note, the old SGI campus in Moutain View, CA is now the home of Google.