Silicon Graphics Inc., the perma-struggling workstation-turned-server-maker, filed for bankruptcy protection today, and was immediately bought by Rackable Systems. Rackable has signed to take on all of SGI’s liabilities. The deal is expected to complete in 60 days. The combined company will target the hyper-competitive market for x86 boxes for cluster and high performance computing, internet and cloud services. “This combination gives us the potential for significant operational synergies, a strong balance sheet, and positions the combined company for long-term growth and profitability,” said Rackable CEO Mark Barrenechea.
Well, as the title says
Nope.
http://www.rackable.com/news/pressrelease.aspx?prid=672
:).
Doesn’t look that way, sadly. Both SGI and Rackable show the news prominently on their respective front web pages.
A sad end for a once great company.
To me SGI was always high end high grade work stations. The kind you could never afford but always wanted!
Its such a shame to lose a brand as powerful as this. I can only hope the new owners revive the brand and retarget it at its workstation hertiage (not itanium servers which were rubbish)
It would be like losing SUN or DELL today(i dont think you could every reeeeealy lose IBM ;-))
I’m still surprised that Apple had no interest.
$25 million is not a lot for a company that has $28 billion in the bank.
They could have learned a lot and had a real understanding for two other markets.
SGI also had some 500+ million in debt that the buying company had to take on. also it would make no sence to Apple, not even a little bit. It barely makes sense here. how sad though how the mighty have fallen, SGI and soon posibly Sun, it just makes me sad really…
Back in 2001, I thought Sun would be around for a while…. a long while.
Wow, was I wrong.
Makes me wonder if my current stock picks aren’t just as retarded.
Then again, I did call the Be -> Palm deal. Too bad Palm was a bunch of retards.
“I’m still surprised that Apple had no interest.”
I think Apple wouldn’t have much reason to buy SGI, after all the most important/interresting part of the SGI business at this point is probably their NUMA/interconnect offering.
This is much more the field of a company like Rackable.
SGI reminds me of Corel in that it tried a litany of stupid things and as a result it drained so much money that when a good idea did come along but required significant money to cover the transition period (from the old to the new model) they lacked what they needed.
A prime example was their foray into Windows workstations, their purchasing of Cray and probably the worse of all – selling off the MIPS unit when had it remained focused would have bought it much needed revenue by selling its products to network device manufacturers etc. It was the lack of focus on who they are, what they are, and a greater vision for the company to work towards.
Yes, there was a transition to cheaper workstations that were good enough, but this didn’t happen over night, SGI could have gradually lowered the price, increased volume, and they would have been in a situation where although they would be slightly more expensive – it would still be an option for someone to purchase. Servers the same thing, gradually bring down the price, improve IRIX (making it easier to use) and get it to the point where it would be an possible option rather than a pipe dream.
Yer, selling off MIPS was very stupid, and like many companies that make knee-jerk reactions when the writing is on the wall, they throw the baby out with the bathwater. Moving from MIPS to Itanium was even more idiotic, and not only was it the wrong loss-making platform but it also destroyed their credibility quite a bit.
Moving into a high-performance computing niche was obviously the way to go because that’s how they ended up surviving this long, but they sold the very thing that would have kept them in business.
True, and hey, even if their high end failed – they could turn into a fabless chip design company, fold the existing engineers into MIPS, change the name, and focus on low cost, low power CPU’s and chipsets for the embedded market. They would have a plan B if their plan A failed.
Imagine today had they still owned it – they could be jumping on the Netbook bandwagon pitching low powered alternative to Atom, even if they only got 10% of the netbook market (35million) it would still be 3.5million units selling with MIPS based processors – imagine the money they could have made!
Btw, you talked about MIPS based machines; they actually owned the MIPS organisation itself; the organisation that designed the chips. A gold mine had they used enough commonsense to see their MIPS as the saving grace if their workstation and high end side needed to be put out to pasture.
For everyone that remembers big, expensive Unix workstations from the 80s and 90s this is the end of an era. Sadly, SGI got its lunch eaten by more powerful x86 PCs with the advent of them running the x86 ‘Unix’ everyone had been waiting for – Linux. They got into Linux too late but it kept them going for a while rather than prolonging what was already dead. Sun also got their workstation business eaten and the end-game is in sight for them as well. They’ve had a bad quarter, trust me. It’s been a tough period that is finishing a lot of once great giants.
It really is the end of an era. Is Unix dead? Almost. There’s certainly nothing new happening, that’s for sure.
Yup, it’s the end of an era. Boy did I enjoy using IRIX.
SGI died because of the following reasons:
1) Everyone who made the company great either left or was laid off.
2) With the exception of a few updates, a new SGI MIPS machine was basically using hardware and software designs from 1996. The latest MIPS R16000 is basically an R10000 (1995) fabbed on a smaller die running at the higher clockrate. Irix has stagnated since 1998.
4) Hey! I have an idea! Lets abandon development on the most advanced UNIX on the planet to work on Linux, that way no one will need to spend $20000 on those expensive UNIX machines!
5) Hey! I have an idea! Let’s sell all of our patents to Microsoft so our current CEO will get a nice bonus when he defects!
6) Hey! I have an idea! Lets change our world renowned logo to something a 5-year old made with an old typewriter!
7) Hey! I have an idea! Lets divert all of our engineering teams to develop really expensive proprietary PCs to Run NT, that get the same performance from machines 1/4th their price!
8) Hey! I have an idea! Lets treat SGI hobbyists like dirt! Who cares that most of them are IT professionals who would normally lobby their company or school to buy our products!
9) Hey! Lets hire Mr. Krabs as the CEO, Spongebob Squarepants for Marketing and Patrick Star for R&D!!!
Several years back I was working for Tier 1 ISP provider. Doing some performance measurements across the cloud. Once in a lab I came across a couple of dozen of SGI stations, sitting on the shelf, doing nothing. Been told when company was looking for Unix stations to put on the cloud to measure traffic, somehow SGI convinced them to purchase their machines, on the basis of performance. What was the problem – software support. SGI implemention of Unix was sub par. Many standard interfaces didn’t work properly. Many proprietory interfaces, poorly documented. Ended up getting Sparkstations.
I already had some programs I wrote and runing on Solaris. Out of curiosity I got one SGI station and tried to port the code on it. No luck. Many timer interfaces didn’t even compile. Looked at documentation, realized SGI did lots of proprietory stuff. Gave up on the whole thing, decided not to waste my time.