# It is sure to be removed, so look at this ASAP

Back at SGI, the company did a pitful job explaining who it was, what it did, and how it did it, not to mention why it was important.
This is marketing, basic simple, get the message out marketing.
In the 90’s, SGI was a great company. We had at the time, great products.
And, as we naively believed that the world would beat a path to our door in order to get them, we sorta … kinda … forgot to tell people about what we did and why it was important.
Though I should point out, that occasionally, a flash of marketing brilliance would shine through.
This is one of those flashes. I’ll give you an anecdote after you view the video below. I expect that someone somewhere will complain about copyright, or song ownership and redistribution rights … so expect it to be gone soon.

This was, IMO, the clearest, most succinct expression of what we did and why it was valuable that I ever saw. I showed the video (on VHS) to customers, to friends. Everyone loved it. They got it.
SGI never did though. When it was released, it was during a sales meeting in 1996 or 1997. Apart from the VHS cassettes, our customers never saw this.
And the anecdote.
I protested to internal discussion groups that it wasn’t going to be seen by any decision maker, or by recommenders, as it was largely an internal use only video. I suggested (half in jest) of accidently getting it into the hands of some friendly media folks I knew.
I was slapped down.
Turns out they didn’t correctly license the music for replay, so they couldn’t distribute it.
Yes, thats right. A humorous, and yet informative commercial from 10+ years ago, making the case that we could help in all aspects of hard data collection, computing, storage, visualization, and decision making, was placed in cold storage because someone somewhere didn’t want to pay a few hundred or thousand extra dollars for the excerpt redistribution rights.
This little video said in 30 seconds what SGIs completely ineffective corporate marketing folks couldn’t or wouldn’t say in the preceding years.
If you sense anguish in my writing on this, you are correct.
How in heck do you run a great company with fantastic products, and then disable yourself telling the world about them?
Of course those were the days when SGI made their own products. And didn’t try to out-Dell Dell.
I expect this video to be gone in a week or less … hopefully not …
Let this be an object lesson in abject failure in getting a message out.

### 5 thoughts on “It is sure to be removed, so look at this ASAP”

1. IOAD,B (I’m Only A Dog, But) He says that he’s had a relationship with SGI since 1988, and He believes SGI’s biggest mistakes were not marketing mistakes. From memory, in roughly chronological order: the CEO from HP; the attempt to move into the consumer sphere; the low-end mips solution with integrated graphics; the early x86 solution with integrated graphics; most would argue the Cray acquisition; jumping onto the ia64 bandwagon-that-never-was. To be honest, the competing success of x86 left them with few choices; He’s not sure whether dropping mips was inevitable, and he’s heard the argument that by promoting OpenGL, they gave away their lunch. Their lifeline now seems to be large ia64 systems for the gummint, but there aren’t many others who want them. Most folks want x86_64 clusters, and it’s hard to differentiate yourself in that space. You’d think they’d specialize in HPC clusters, but even ICE requires external wiring for IB. A sad Woof to SGI. (Or what is that sound that dogs make when they cry?)

2. @Dog (or Yo Dog? 🙂 )
For HP, you mean “Rocket Rick”. I’ll go further. TJ, Ed McCracken, etc were basically the problem. It was downhill from there.
SGI’s mistakes were not technological, but marketing, market focus, not having enough marketing intelligence or believing the same to guide the product mix going forward. Not understanding how these little upstart chips from Intel could knock off the mighty mighty MIPS.
While recently cleaning up a machine to be retired (junked), I found emails of benchmarks on chemistry/informatics apps I had sent colleagues or posted to sgi.bad-attitude pointing out that the cost-benefit analysis was rapidly favoring the pesky upstart chips, and we really ought to be doing something about integrating them into our products.
Cray acquisition … I still believe that was asked of McCraken. We were busy terminating most of the Cray business we ran into when we up and bought them. I liked the majority of the Crayons I worked with, I found them intelligent and personable. It was the management of Cray, that largely took over SGI later on, that pretty much killed any hope we might ever had had. The J90 was supposed to die, but the product manager kept wanting to make them. The massive C932 had something like $100k circuit boards (our BOM costs). Made mistakes/RMA very very costly. Cray had, back then, many existentially challenging issues. SGI never finished digesting them, never made the hard decisions about what to pursue and what to kill. Consumer sphere … set top boxes? Ahead of their time. By at least a decade. Good idea, too early. The NT PC? Oddly, this was a good idea that many people hated. It had literally just started taking off in the sales offices when it was whacked. Low end MIPs solution with graphics? They didn’t push that hard enough. If you have an advantage, run with it, hard. IA64? Yeah, I was one of the (multiple) vocal opponents, daring to suggest that the IDC numbers were … uhm … I dunno … off a bit? Yeah … thats the ticket … off a bit. I remember begging management to have a plan “B” ready, just in case IDC was … you know … off. I remember them looking at me funny. Intel’s projections of uptake couldn’t be wrong. No way. Thats what I was told. If you happen to see any of them, you can tell them “I told you so” 🙁 but it won’t make me any happier to do so. SGI should have fully embraced x86, as it could have been a powerhouse there. Bring OpenGL on boards to the masses. Had they done this correctly back then, we wouldn’t likely be turning to nVidia/ATI for our OpenGL boards. Considering nVidia is a$5B company (was $10B recently as 2 years ago), and SGI is today a$6M company … this would have been a very smart move on SGIs part. That direction is foreclosed for them though. OpenGL was a way for them to grow if the elected to use it.
I agree that not too many folks want IA64 anymore. It has largely underwhelmed the market.
As to the point of x86_64 cluster differention, this is what I mean when I point out you can’t out-Dell Dell, and you really, really, shouldn’t try.
I do see your point in terms of engineering and management direction issues. I argued those as well in the past. Must have let them slip. Marketing wasn’t the only thing SGI was terrible at in those days. And other things have largely led to its demise.
But it is hard to beat demand. Pure unadulterated backlog should always win the argument.
And sometimes it does. But to get that demand you have to build products customers want, and tell them about it. The latter part is marketing. The former part is built atop marketing data, but are decisions made by the management team.
Ugh… I feel as if I am on sgi.b-a again. Must break free …

3. Um, don’t forget how upper levels smacked around potential customers’ sysadmins in the same timeframe. Our security concerns were “stupid” and unimportant. Yeah, the machines rocked, but they weren’t much good when firewalled off from most of the DATA.
Overall, a lesson in how it takes more than great engineers.

4. I didn’t see a huge amount of this … must have happened outside of my area.
I did see huge pushback from customers on deploying large numbers of machines because our machines required you sit there in front of each one with a CD, keyboard, monitor ….
Almost got us kicked out of Ford.
This I remember, because my management asked me to do something about it. And I did. Most of the customer base seemed to like it (autoinst), but we kept running into a few that insisted it could never be as good as jumpstart or other tools. Platform fanboi’s aside, the approach of working with customers, helping them solve their problems, was a good one.
I am sorry to hear that SGIs management ran roughshod over your security concerns. To this day we try to help customers prevent lossage, or at least contain it. We aren’t always listened to. I could go on and vent over the fallout from this, but suffice it to say that we are taking very pro-active security measures with our customers, working with them to craft a secure system, despite the best efforts to suppress such discussions.

5. This takes me back to running a network of about 8 different Unix variants on about a dozen hardware platforms for the TenDRA compiler group – our opinion back then was that SGI was a case of “nice hardware, shame about the operating system” – Irix was full of holes and gave Sendmail a run for its money in terms of the bug-of-the-month club (accounts with no password – ugh!).