Sign up for our newsletter and get the latest HPC news and analysis.

Addison Snell remembers SGI

Addison Snell, now of Tabor Research, but once of SGI, remembers his time at SGI and offers his view of what went wrong:

SGI logoDo you know what the saddest part is of Rackable’s attempt to acquire SGI’s assets out of bankruptcy for a paltry $25 million in cash? There would be very little effect on the HPC market as a result. Thanks to a cascading series of problems — bad marketing, operational misfires, bad technology bets, and ordinary bad luck — SGI didn’t have much left to lose.

It’s an interesting post in that it walks through most of the milestones we remember of the recent (post-graphics giant) SGI with an insider’s take.

The rest is not as interesting, and it’s more recent, so you remember it. SGI started working on clusters. Bishop was out, replaced by Dennis McKenna, who guided the company through its first bankruptcy, then Bo Ewald. Under Ewald, SGI purchased the remains of Linux Networx and integrated the software into its own cluster stack, but it was too little too late. Customers had been hurt too many times, and they weren’t coming back.

The last five years have been a grind for SGI, and many of us – especially us former employees with purple blood – silently rooted for the company to succeed. “Not dead yet” was the constant diagnosis. Until this month. Now SGI is gone.

I’m not willing to write off the company yet. Maybe I’m just clinging to happy memories of my own SGI-filled days of happiness (my own Crimson as a master’s student, a Personal Iris connected to my own 64-processor nCube, my Indy, Octane, and Indigo2, and a horrible PowerChallenge Array that wouldn’t run MPI between cabinets). But at this point the partners have at least privately professed positions that would keep the core of SGI’s innovation — large scale shared memory in hardware — alive in the market. Of course implementation is a different thing.

Comments

  1. You are sooooo wrong. The PowerChallenge Array that would run MPI between cabinets. You just were not guaranteed the same results each time you ran it. You could get one of four outcomes: the right answer, the wrong answer, a core dump, or a machine crash.

    I remember that PowerChallenge Array all too well.

  2. I had not heard of this before so I found this really interesting to read, and have an understanding of what is going on, I found it strange that the company is gone but you yourself have said that you are no writing the company off just yet.

Resource Links: