Slashdot Mirror
SGI Warns That Bankruptcy Might Be Year-End Option
tbcpp writes "OS News reports: "SGI issued its most ominous regulatory filing to date, warning that a bad 2006 could force the former high-flyer into bankruptcy. In order to improve its business, SGI will consider measures ranging from axing or selling off product lines to pursuing 'a strategic partner or acquirer.' The hardware maker will basically look at anything and everything to remain a going concern.""
They backed the wrong horse (Itanium) and don't appear to have a Plan B. We have some nice parting gifts and the home version of The Silicon Valley Company Game.
SGI's heyday was when most people thought of them as The Purple Computer Company; the Jurassic Park Era. And yes, their lack of a brand identity and strategy was part of their undoing.
"I'd rather be a lightning rod than a seismometer." -Ken Kesey
This would be a real shame as SGI has talent for engineering great systems. The Altix is a really nice architecture, the idea being you start with a 4 CPU node and can scale to a very large system with a single system image, high availability, easy scalability of memory, cpu, storage and interconnect, and has nice management tools for partitioning, etc. Unfortunately the price of entry is a bit high, and I think that perhaps going with IA64 rather than the budding Opteron was a misstep at the time.
I also feel they lost a lot of momentum by dabbling in various unpopular markets like high end NT workstations, expensive specialty graphics workstations (given this was a core market for them earlier, but high power graphics became commoditized) and didn't really strongly launch into the linux server market and make a big presence in time. If they had pushed a cheaper starting system for a scalable single system box they may have done better, but who knows.
The fat lady has been following them around for a long time.
SGI faced the innovator's dilemma big-time; it was tricky to cannabalize their $2 billion workstation business for a $300 million graphics card market. And to move from being a full-system vendor to being a graphics card vendor. And even with all the management and business-issue problems, I noticed three problems their engineering effortsg never overcame:
- trouble with quality and shipping on time (see IMPACT)
- couldn't match/switch from 3-4-year development cycles of the workstation business to 6-month product cycles of the PC graphics card business
- engineers were loath to give up control of the chipset/box/OS in order to settle for just controlling the graphics subsystem. They tried to be a full-system player in a PC world. Given that Compaq couldn't really do it (something that was at least semi-obvious at the time), its not a surprise they, coming from the workstation space, couldn't do it with their integrated NT workstations.
- The engineers were delivering product that was differentiated but not in the areas that the biggest customers cared the most about. The benefits of UMA (unified memory architecture) graphics just weren't in sync with what the market most wanted: the fastest 3D at the cheapest price. And in the classic workstation space, polygon-pushing was what was most needed. Half their business was CAD workstations and in the end they lost that to Sun/HP/IBM who didn't have the sexy texture mapping stuff but could render polygons "good enough".
SGI also benefitted from many years from the other workstation vendors under-investing in 3D graphics. When that era ended, even the workstation business they were in got a heck of a lot more competitive.
Anyway, that's what comes to mind when I remember back to SGI in the mid-90s. In hindsight, I don't know of any silver bullets that would have gotten them out of the situation; it was death by a thousand cuts. At the time, I wondered if a merger with Apple would have made sense but it wasn't clear that the disfunctionality of the two organizations at the time would have melded into something better. (11072394) Maybe a damn good CEO could have helped them carve out a more defensible role in the industry; that's the only thing that got Apple through as far as I'm concerned.
They were responsible for the OpenGL spec itself, had a ton of influence on directions taken in the CG market generally, and instead sang endlessly about something called "Virtual Reality" while the rest of the world realised that unless it could be affordably domesticated, there would be no market for it. While NVIDIA and ATI said "Hey, mind if I check out this 'gaming' thing while you're out?" they were selling Caves with Dolby and a few O2's to CEOs of mining companies and a few UNI's once or twice a year.
I know, I worked in one. SGI reps would come over with "THE FUTURE" written all over their face even when we were openly replacing their boxes with white PC's running GeForce cards.
Snobbery or stupidity (they often converge), it is utterly their fault.
But their stock performance was very, very mediocre, and I wondered why so many smart people could do things that don't shine in terms of corporate results.
In the late 1980s I saw the same thing happen with a Hyundai. Motoring magazines reported on a really nice sporty little car they'd prototyped. It was really cute two door, a unique looking convertible that would have sold like hotcakes. Then as it got closer & closer to release it gained full rear seats instead of being a 2+2 layout. Then it got a bigger trunk for more luggage, a fatter roofline for more rear-seat passenger room. The "radical" front styling was softened, then it was given another two doors. In the end it was just another small four door hyundai, and when released was received so poorly it never made it out of Asia.
A press statement from Hyundai stated something along the lines of "market anticipation failed to convert to sales" when it was canned. That's because the beancounters, the conservative marketers massages the product into something virtually the antithesis of the original product the market built up its anticipation about.
Seems a common theme in the big companies, where something good is created but because of a lack of forceful "no. don't touch" from smart management everyone gets to poke their fingers in and change things, making Yet Another Lowest Common Denominator Product.
We all though Apple was dead and out too. And now look at them. It sounds like it is house cleaning time and a change of direction. The OpenGL standard is nice but really out dated. If a company created a real time ray tracer (RTRT) they could pomel the raster graphics erra. But this is just my opinion.
It's amazing how quickly people forget. Ricky "Microsoft Mole" Belluzzo is the reason SGI got it's head-shot to begin with. Remember Rick? Yeah, he was the guy who, while working at HP back in the mid-90's, made the announcement that HP would be "dumping HPUX in favor of Windows NT" without any warning or approval, forcing HP to do the world's larged backpedal ever seen.
He then went to sabotage SGI with the SAME STUPID GAMBIT, before finally going "home" to Microsoft.
If they'd had their dramatic culture shift towards Linux and back towards Openness only a year or two earlier, it would have made a big difference to them.
Nope.
What killed SGI was the standard Big Computer squeeze, just like it killed Cray, DEC, Tandem, DG, etc, etc. The commodity hardware improved enough to eat their lunch, and there just aren't enough of the super high-end customers to keep them in business.
SGI could have survived by returning to their roots as a graphics hardware maker. Instead of ATI and Nvidia, we'd have SGI and a handful of also-rans, but SGI's management thought that making graphics boards for PCs was beneath them.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
Microsoft always has to leave some room for improvement, so that people will buy future releases. I'm sure they still regret the whole year 2000 "lets make good software" fiasco. Companies are going to keep running those 2k products until their hardware fails. That why they decided to delay Longhorn/Vista so many years.
They certainly can afford it these days,
The cost of merging with SGI would be way beyond the cash outlay. SGI brings with it an enormous burden of management distraction: government contracts, security clearances, legal issues, etc, etc.
and benefit from all the UNIX goodies that SGI has produced over the years.
Everything Apple could possibly want from SGI can be had without buying the company.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
SGI *was* cool. How many of you has actually help the company by really *buying* some stuff from them?
Apple *was* cool and *is* cool. I, like many many many of you, own an iPod.
Just room for one big poseur up on top, eh?
If you think that bringing the company back from the brink of collapse, to its current state of record-breaking growth makes him a poseur, then we're not speaking the same language.
-jcr
The only title of honor that a tyrant can grant is "Enemy of the State."
Another thing you should take-away from the experience is the importance of diversity. If you took that money and split it among five companies, one of them being an amazing success like Apple (which has gone up 8x in five years), and the other four being dismal failures that went bankrupt, you still would have made 60% returns on your entire portfolio.
By investing in only one company, you really put yourself at a disadvantage.
my blog
That's true for the software, but not the OpenGL trademark.
And what of subsequent revisions of the spec? Are they always guaranteed to be open?
"Nine times out of ten, starting a fire is not the best way to solve the problem." - my wife
" I just think its cool being able to purchase computers that used to cost tens of thousands of dollars for a couple hundred bucks."
That's cause you haven't actually done it.
I have a few bigass suns in my barn I picked up, 4 yrs old for like 2 cents on the dollar. Lots of cpus, ram and disk.
Like I said, they're sitting in the barn now. My not so recent IBM 1U servers are way faster and use a tiny fraction of the power.
The Suns would be ok space heaters if they wern't so damn noisy.
Need Mercedes parts ?
Engineers did not flee to nvidia, sgi sued them for patent infringement and won (settled out of court), they got cash from nvidia and then transfered engineers to nvidia. They started to use the nvidia cards in thier PCIG's (the settlement assured them the top 10-20% performing cards), then canceled the project. Now they use ATI cards in thier onyx 3xx systems. It is true that they seem to only want to sell systems to the high end users, while ignoring developers and entry level systems. Several high level people have left over this issue and more are fleeing. They have canceled production of thier Origin 3000 systems, which still have a very strong aftermarket demand (my company cannot keep enough of them in stock...). We have sent them several deals for new systems, but the customers feel that because they have bought remanufactured systems from our company, that they get the run-around and eventually give up and buy more used systems from us, when they really wanted the new technology (we get them the Altix systems on the used market). These were deals in the several hundred thousand dollar range. We even tried to become a reseller for them, but again because we sell the refurbuished systems our "business model will never allow us to become resellers". Quote from Tom Wall. That is fine, we make a HUGE margin on the refurb's, and undersell sgi EVERY time. We were in business before sgi and happily (or sadly) it looks like we will be in business long after they are gone.
Idiots should have got into the GFX card market.
Uh... people who worked for SGI DID. It ain't their fault that management's a bunch of asshats.
Where the smegging hell do you think ATI and NVidia got their talent, hmmmmmmm?
And, contrary to the Slashdot headline, SGI is one of the companies that is kicking in some of that $10 billion.
Breakfast served all day!
Except that you don't make money selling PCs. Everybody sells PCs. The margin on a PC is a few dollars. You need a huge volume to even consider making a profit.
You make money by selling real computers. You do sell less of them, but you do make a profit and you get to sell a bunch of services.
SGI was making workstations and servers. The market for workstations died. Instead of focusing on high end machines and dropping the workstation market, they redesigned their workstations as funky coloured PCs. Now they die.
Beluzzo did his job well. He was richly rewarded.
May contain traces of nut.
Made from the freshest electrons.