How fast are SGIs really

[jtk77]

you guys missing the point.
If SGI hit mass consumer market, they would put out cheap little no power boxes running their OS using WORD type products.
Wouldn;t be the O2 and such you are thinking of now.

They would have had to mass produce cheap boxes , the key was they were a few years ahead of MS at the time, but MS and intel suddenly produced cheaper boxes and went into full blow mass -market production and sales .. MS buying all office products and then putting them in the hands of business with boxes that ran Win95 then 98 and thus making a ton of cash..billions.
With the right thinking SGI could have done that as well .
but they were too narrow minded and focusing on the HIGH level systems.
You HAVE to have a consumer level somwhere in order to thrive. Yopu can eep the high level for the prestege and R and D factors.
MS knew this.
Apple figured it out.
Sonys got it down..
Nvidia...
tons of other conpanys figured this formula out and theyare profitable now.
SGI stayed in there little un-expandable niche of high-end graphics.

[gmask]

Consumers never had any reason or interest to work on SGI -let alone UNIX. -too expensive, too much command line, too expensive, no useful perhipherals, uh...too expensive.

Didn't SGI have a processor that was used in a playstation or n64 at one time? That's about as consumer level as it gets without the complexities of their OS.

Truly though their highend workstations were never meant to be consumer level computers. Even for more demanding buyers once the PC became a viable platform they could not possibly compete and they could not maintain and edge when it came to competing on the wintel workstation clone market... I think mainly because by that time the market for highend display cards made their offerings seem less significiant considering the extra cost than just a few years before.

So they are staying afloat with their massively scalable super computers.. this seems to be to have allways been their mainstay for R&D and their continueing clientel of the scientific and governmental concerns.

Would the consumer market appreciate an operating system that rarely crashes probably but not if they could buuy something less expensive like windows.. Consumers have been groomed to expect a certain amount of planned obsolescence to everything they buy including software.. so investing in a platform that will never crash and due to the cost will require that you use the system for ten years is foreign and impractical.

[jtk77]

who knows? now sure, then well..it was up for grabs.

[ottscay]

Actually, I think that SGI should have made a gaming console. Not on their own of course, that's a marketing nightmare (see Microsoft), but they could have teemed up with Nintendo or Sega to make an advanced gaming console and mass produced them...

Oh well.

Scott

[jtk77]

Really , they should have explored any consumer base.
Companys that stay proffestional only and don't make at least half a million per hardware unit conssitantly do not make it, or at least make it successfully unless they stay small.
If you look at history its true.

[nuabudikae]

The Nintendo 64 had a custom MIPS processor that was scaled back a bit versus the SGI offering in their workstations at the time. Like someone else said, that is as close as they ever came to the consumer world.

I don't think SGI could have ever successfully created a consumer level computer, complete with consumer level software and a price joe six pack would be willing to pay for. Consumers have consistenly shown that they want bare minimum functionality for rock bottom prices. That is why computers under 800.00 these days (with monitor and printer) are so popular. If consumers get more than the minimum, then they are happy but they are pretty adament about sticking to a certain price segment.

Those that are willing to go to the prosumer level in the price structure typically look to Apple or the X86 Workstation vendors to meet their needs. These folks typically know what they are looking at and understand the value of a SCSI hard drive and a good video card and are will to pay accordingly for it.

Up next is the high end pro user which is typically where the SGI marketing began. They would appeal to the Hollywood film buffs who had money to burn. At the time the SGI was the defacto standard for high end work and i think few in the company truly understood how "their" industry was going to be radically transformed in the next 6-18months.

Perhaps SGI should have paid a little bit better attention to where their X OpenGL hardware engineers went. It was none of than nVidia. CEO/Founder Jen-Hsun Huang had the vision to realize that the next big wave in computer graphics would be consumers clammering for real time 3d graphics. He geared his company accordingly and turned it into arguably the worlds most powerful 3d hardware entity.

His vision reminds me of Doug Walkers with Alias/wavefront. They continue to push the R&D envelope while at the same time are exploring and will soon be revolutionizing a number of things 3d. He is positioning his company to tap into the next 3d wave. Lets never forget the primary objective of any successful business is to make MONEY.

My .02 of rambling...

john

[kiwi]

Are we all forgetting that Silicon Graphics DID make a run at WIntel-based PC's?

They were technically, for the most part, very sound designs.

They used good old-fashioned quality engineering which, unfortunately, required some expensive components and significant R&D overhead.

With all they did right in their efforts to mass-produce the most kickass and desireable PC workstation in the world, the 320 and 540 Visual Workstations still only rated as an "OK" value with most pro's.

Their "Cobalt" 3D hardware Pipeline was only barely able to keep up with the newly-killer NVIDIA and venerable Wildcat boards, and DirectX implimentation was non-existent (gaming suffered a bit on these boxes)
ands fell further behind in the price/performance arena quickly...

Using NVidia chips after that was the only move they could have made -there was no time or resources to engineer a completely new 3D hardware pipeline for WINNT -heck, they still havn't done it.....

-but, ultimately, it was NVIDIA (and to a lesser extent, the OTS (off the shelf) PC performance explosion) that did the most to put SGI's PC's out of business.......

I'm, glad they pulled-out .....gracefully as they did.

KIWI

[nuabudikae]

Are we all forgetting that Silicon Graphics DID make a run at WIntel-based PC's? <<

GRIN, yeh i had forgotten about those. Shame on me. Those boxes always struck me as being astetically unappealing compared to the jazzy Octane or even the cute O2.

[1000101-]

They did actually buy out intergraphs workstation division and also put out the 330 and 550 series (both powered by nvidia)
Its just the company changed directions so often and never focused in any one given direction
result

DETh

[jtk77]

I have a Zx10 SGI..had it for like ...3 years now? Almost..freaken thing still sings..though..
I will have to ditch my Wildcat 4210 card soon for an FX if the Maya 5 render thing is as cool as it sounds.

[bundee3d]

Well, IMHO the introduction of the 320 and 540 was not hampered by cobalt's performance. I think price and upgrade ability was the main issue. Things like, you couldn't turn Cobalt off and to max out the 320s ram at 1024 would have cost over $5,000.00 Cobalt may not have been the fastest at doing 3D but it offered very good 3D and 2D acceleration in a stable environment. It is my understanding that in 2D it actually out performed the Wildcat. I base this on some benchmarks I read around the time the 320 and 540 were released.

I really think developing Pentium based workstations was a good idea. I think the real problem lied in SGI's attempt to apply the same business model to wintel that was successful on IRIX. Sell expensive hardware then sell an expensive support contract. Of course this didn't mesh with PC users provided there were cheaper alternatives. As far as things like gaming and Direct X, it was kind of a problem for me but you have to remember who SGI was designing these machines for.

Just in general I think SGI's overall business model was too rigid for the modern workstation and they lost their edge as the market begin to change.

'cel

[mike_kennedy]

I have worked on SGI, Apple and PC's ranging from the SGI 320 and 540 to Intergraph to biege boxes.
I have found that the biege boxes were the best value by far because even though they are not built to last the way the high priced workstations were, they were far cheaper and far more upgradable meaning I recieved a far better return on investment.
If a biege PC costs $2000 and the high priced work station costs twice to three times that, then I am limited by the high priced workstationin terms of how current I can stay speed wise. But if I stick with biege boxes I can stay much more current, and often do so cheaper.
I can't see how SGI could even bother with the low end workstations at this point. They missed the boat completely.
Irix was nice to work on, but it was always slower for what I was doing....
ANd I always buy with my own money and there for an SGI is pointless. My clients never see my boxes and no one cares about SGI anymore, they just want the best quality for the best price.
Cheers
MK