SGI: Discussion

I wonder if anyone has any thoughts on this comment from Greg Matter over at Sun? (Permalink: http://blogs.sun.com/roller/page/Gregp? ... e_confused )

By 2010 microprocessors will seem like really old ideas. Motherboards will end up in museum collections. And the whole ecology that we have around so-called industry standard systems will collapse as it becomes increasingly obvious that the only place that computer design actually happens is by those who are designing chips. Everything downstream is just sheet metal. The apparent diversity of computer manufactures is a shattered illusion. In 2010, if you can't craft silicon, you can't add value to computer systems. You'd be about as innovative as a company in the 90's who couldn't design a printed circuit board.

clavileno wrote: I wonder if anyone has any thoughts on this comment from Greg Matter over at Sun? (Permalink: http://blogs.sun.com/roller/page/Gregp? ... e_confused )

By 2010 microprocessors will seem like really old ideas. Motherboards will end up in museum collections. And the whole ecology that we have around so-called industry standard systems will collapse as it becomes increasingly obvious that the only place that computer design actually happens is by those who are designing chips. Everything downstream is just sheet metal. The apparent diversity of computer manufactures is a shattered illusion. In 2010, if you can't craft silicon, you can't add value to computer systems. You'd be about as innovative as a company in the 90's who couldn't design a printed circuit board.

a typical blog - just drool.

foetz wrote: a typical blog - just drool.

Seems that way...

clavileno wrote: I wonder if anyone has any thoughts on this comment from Greg Matter over at Sun?

This guy needs to read some good sci-fi and pull his head out of his ass.

Well, MIPS never did any silicon. They have 3rd party foundries.

The only reason why SGI bought MIPS to begin with was because in the early 90s MIPS was about to go bankrupt and SGI needed to ensure they had their processor supplier around. Maybe it was the adquisition of MIPS that doomed SGI....

even if we had all paperwork directly from sgi's financial dept. it would still be hard to say what exactly made problems and what was good.

all these 'what happened' or 'who didn't clear the toilet in bishop's room' threads are so useless...

dunno about 'useless', they're a pretty good vent for frustration about the protracted demise of all that what was cool, fun and impressive about SGI. till the mid/late nineties.

Whether such threads actually make any impression on reality is another matter entirely.

Like an autopsy, I think there are lessons that can be learned from poking around in the corpse. It's not really about blame at this point, just wisdom. This rise-excess-fall cycle is ubiquitous in human endeavors. For this crowd (relatively young and moving into or towards lead roles in society), it may help avert future dissaster.

When I was a pre-teen, my favorite comic book was the X-men. It was right after issue 100 that it caught my eye and was about to begin its sustained surge of popularity. The lesson that it impressed upon me it, in the simple terms for that day "The worst thing that can happen to anything is for it to become popular." Strange when you think about it, but almost always true---sucks the quality right out of it. Now almost 25 years later, I'm still trying to figure out exactly why that is. The closest I've come to an answer is that it attracts the wrong kind of people.

The other (related) eternal puzzle is catagorized in my mind as "Why did the Beatles break up?".

So I now add: "Where did SGI go wrong?"

and on a more positive note: "How did Apple turn it around?"

Well, let's start with something different on this thread. I want to talk about the power of the marketing, and how you can make look better what in fact is the same, almost the same, or plainly worst.

Please notice that I'm not intending to offense to nobody, but:

Anyone noticed that:

1) When SGI uses Intel processors, we say "-their new hardware is just not the same as before!"
2) When SGI uses ATI graphic pipelines we say: "-oh nahhh... these crap graphic hardware from these canadian company... after all... what could know the canadians about how to make good drivers?"
3) When SGI used to have on their site the yin-yan logo, with the legend: "Linux And Irix: The Best Of Both Worlds", we said: "-oh geez!; please, somebody there can remove these nasty logo thing with Irix and Linux on the same sentence, right now?"
4) When SGI uses SATA drives we say: "oh yup; those cheap and unreliable shitty disks for gammers... oh well, they are not the same as SCSI!"

...And anyone noticed also that:

1) When Apple uses Intel processors, we say "-wishes do come true!"
2) When Apple uses ATI graphic pipelines we say: "-Oh!!! ...That means it’s screaming fast. Up to 256MB GDDR3 graphics memory!!!"
3) When Apple uses to have on their site the yin-yan logo, with the legend: "UIniversal Binary: Software just works", we said: "-oh geez!; it is a miracle... they managed to run Mac apps inside the intel-OSX!!! Geniouses!!!"
4) When Apple uses SATA drives we say: "-bloody hell!!! ...they will provide an amazing 250GB 7200-rpm Serial ATA hard disk drive!!! Cool!!!"


...Oh Steve; what a magnificent genious you are, and what a magnificent genious are your marketing chieffs!!! ...You can make almost the same thing that the rest, but it will look different, and people will think that they are "Thinking Different" ! ...Heheheh... we gotta love you, mediatic Guru!

I think Apple turned it around because:

1. They got back their colourful, egomaniacal leader.

2. They had a crufty old OS that needed to be thoroughly revamped. Which they did, and the revamped product offered real benefits over the old one.

3. They realised that the old 'applets' that came with their OS, were... old. So they started building/shipping applets that fit with current desktop usage: photos/images/video/music.

4. Jobs understands that design _really_ matters. The beige corporate square boxes with uninspiring model numbers were replaced with sleek sexy stuff designed by Mr. Ives. The 'hallowed' gray finder/Mac OS look 'n feel is now a gaudy eyecandy fest.

5. They diversified into high volume hardware after first grokking on-line distribution of digital music.

6. They made sure that the platform remained a viable one for 3rd party ISV's.

Those that know a little about SGI history will see how SGI could/should have done the same. Here's my take:

1. Jim Clark was the soul of a company he ran in order to make huge piles of cash by creating cool stuff. Then in came the suits to turn SGI into a respectable 'proper' company. Clark didn't take well to losing control and left to make more money elsewhere, SGI never recovered.

2. IRIX needs to get proper support for a bunch of commodity hardware, and a bunch of other system/network software. It'll never happen.

3. A major desktop revamp was on the cards as late as 2001/2002. It never happened.

4. Having met Bob Bishop a couple of years ago, I wasn't impressed by his awareness of things relating to 'coolness' and 'design' - something SGI was always famous for. Mr. Bishop simply isn't a superstar CEO. Jobs is, Clark was. Can't see Bishop building a way-over-the-top sailing ship. If I'm not mistaken, the man has a house in Switserland - that exciting land of the Cookoo clock.

5. SGI should have grokked PeeCee based desktop 3D way back in the days of the Voodoo cards. A bunch of SGI engineers left and now NVidia is doing the high volume desktop 3D stuff.

6. Yeah right.

Well Diego,

that is really easy to explain, for Apple it's an upgrade but for SGI Though I wish there will be some new Apples with nvidia gfx inside, that would make a really nice package.

Matthias

As to the original post asking to comment on: "chips is where it's at, mobo's are obsolete."

I'm not sure I understand what this greg chap is on about in his blog. As I understand things, chips need interconnects - no? Mobo's are simply surfaces that support and implement interconnects - no?

If so, if you've got chips you'll have mobo's. Though it's unlikely that the classic PC _architecture_ will still be around in 2010.

On the other hand, wasn't the whole point of the Origin 3xxx architecture to make a system with upgradable interconnects? Even so, somewhere along the line, a 'chip' will have to be stuck onto some kind of connector-to-the-rest-of-the-system thing, and that's a Mobo in my book.

squeen wrote: Like an autopsy, I think there are lessons that can be learned from poking around in the corpse. It's not really about blame at this point, just wisdom. This rise-excess-fall cycle is ubiquitous in human endevous. For this crowd (relatively young and moving into or towards lead roles in society), it may help avert future dissaster.

sure but we don't have the corpse

The other (related) eternal puzzle is catagorized in my mind as "Why did the Beatles break up?".

GeneratriX wrote: ...And anyone noticed also that:

1) When Apple uses Intel processors, we say "-wishes do come true!"
2) When Apple uses ATI graphic pipelines we say: "-Oh!!! ...That means it’s screaming fast. Up to 256MB GDDR3 graphics memory!!!"
3) When Apple uses to have on their site the yin-yan logo, with the legend: "UIniversal Binary: Software just works", we said: "-oh geez!; it is a miracle... they managed to run Mac apps inside the intel-OSX!!! Geniouses!!!"
4) When Apple uses SATA drives we say: "-bloody hell!!! ...they will provide an amazing 250GB 7200-rpm Serial ATA hard disk drive!!! Cool!!!"

I didn't really mean to compare Apples to oranages , but since I'm talking to SGI shortly about my concerns with the Prism, I'd like to address the points Diego brought up (and get some responses).

Brombear is excatly right.

1) SGI brought in faster processor (great!) but dumped a better (i.e. more robust, more parallel/scabale friendly) OS (IRIX). So we downgraded to Linux. We're hoping now that some of the functionality can be restored (e.g. real-time). Also, very little effort made to help the developer port over IRIX apps--RapidTransit=no thanks, wrong "broken" Motif version (2.2), no SGI widgets, header conflicts in shared memory arenas/semaphores, extension functions missing from OpenGL libs. No RGB image lib. The list gets longer the more we try.
2) ATI graphics cards are faster vertex processor than InfiniteReality, but in all other areas inferior (e.g. color depth, multisample, pixel read, image processing, memory size, overlays, etc...). They do however support a few more modern OpenGL extension like Shaders. Unfortunately, as I recently discovered, the drivers are wicked broke and barely work. What's more, the interface to OpenGL (i.e. SGI's openGL libs and headers) have adopted the PC cluster fuck API (extension wrangler my ass!) Shameful for a system with *known* graphics configurations!! Again, feels like a downgrade.
3) who cares? --- Taosim rules --- whatever...
4) The research I've done on SATA drives shows thay have slower seek times, less bandwidth and are less reliable than SCSI. The only thing they have going for them is that there are cheap vs. size. But if you are paying $200,000 for a supercomputer, do you really care is SGI saved $300 when they resell you drives? Hell no! ANOTHER downgrade. Whos is architecting these systems for SGI---Linux geeks fresh out of school?

Apple is a consumer product. iLife-style and all that. For hitting their mark I applaud Apple. The products are real works of art---quality in interface design, mid-level performance at an everyman cost. but SGI is trying to sell HPC. The big $$'s means you had better put something *extra* in those machines that the average high-schooler can't get off ebay. Cutting edge technology---because for real science and engineering ALL computers are way, way, WAY too damn slow. NUMA link is good, but the real-time and RASC are only marginal, and the graphics suck.

What's been eating me lately...a growing fear if you will, is an attitude that SGI has been giving me in regard to the Prism. I'm sure it due to limited resources, but I keep bumping into the attitude of either

1) It's SuSe Linux....there's nothing we can do about it.
-or-
2) It's ATI graphics...there's nothing we can do about it.

To which I find myself being forced (and I really do mean forced) into replying: "Then I'll have to find some other vendor that will provided an integrated system that performs." My only problem is figuring out where to go next for HPC---I just don't have the time to gin something together myself. That should really be SGI's targeted niche---folks with more money than time (or computer smarts ). And the systems they deliver should seem like magic---not a jumbo-size version of last year's PC trash.

One last comment: Could you see Apple saying "It's BSD unix...there's nothing we can do about it?"

\rant

By 2010 microprocessors will seem like really old ideas. Motherboards will end up in museum collections. And the whole ecology that we have around so-called industry standard systems will collapse as it becomes increasingly obvious that the only place that computer design actually happens is by those who are designing chips. Everything downstream is just sheet metal. The apparent diversity of computer manufactures is a shattered illusion. In 2010, if you can't craft silicon, you can't add value to computer systems. You'd be about as innovative as a company in the 90's who couldn't design a printed circuit board.

getting back to the original quote; this is more or less a true statement. to paraphrase John Mashey "all computer architectures have devolved into CPU, memory, and the attributes of the ways in which the are connected" at this point, yes you can't do much product differentiation at this level anymore. the falacy of the quote is that real customers that drive the computer industry don't buy these value add chips themselves, they buy them integrated into products the computer people, who can now provide differentiation by determining what those chip vendors produce in terms of 'innovation'. so nothing has really changed, only the scale at which differentiation happens.

of course this is all dwarfed by what product differentiation can be done in the realm of software; where a similar thing has occured. really; how much value do you add to an operating system before you have pleased 80% of the market? not much anymore, everyone has the same basic functionality whether it's IRIX, linux, windows, or macos; it's the applications that are ported to these OS's that sell them now, and their corresponding hardware platform. again nothing has really changed.

so is the original quote valid? yes, in a very general way. but then again at the same scale nothing has changed much since the 60's anyway (if you neglect eyecandy). yes even open source existed back then.

so *yawn* i blew a strut tower mount, bummer. now at least i have an excuse to install my new anti-sway bar end-links, and dual spring rate springs. so maybe my computers are slow, but my car is kickass. i'll blow away gammer on the road.

squeen wrote: I didn't really mean to compare Apples to oranages , but since I'm talking to SGI shortly about my concerns with the Prism, I'd like to address the points Diego brought up (and get some responses).

Brombear is excatly right.

1) SGI brought in faster processor (great!) but dumped a better (i.e. more robust, more parallel/scabale friendly) OS (IRIX). So we downgraded to Linux. We're hoping now that some of the functionality can be restored (e.g. real-time).
2) ATI graphics cards are faster vertex processor than InfiniteReality, but in all other areas inferior (e.g. color depth, multisample, pixel read, image processing, memory size, overlays, etc...). They do however support a few more modern OpenGL extension like Shaders. Unfortunately, as I recently discovered, the drivers are wicked broke and barely work. What's more, the interface to OpenGL (i.e. SGI's openGL libs and headers) have adopted the PC cluster fuck API (extension wrangler my ass!) Shameful for a system with *known* graphics configurations!! Again, feels like a downgrade.
3) who cares? --- Toasim rules --- whatever...
4) The research I've done on SATA drives shows thay have slower seek times, less bandwidth and are less reliable than SCSI. The only thing they have going for them is that there are cheap vs. size. But if you are paying $200,000 for a supercomputer, do you really care is SGI saved $300 when they resell you drives? Hell no! ANOTHER downgrade. Whos is archetechting these systems for SGI---Linux geeks fresh out of school?

Apple is a consumer product. iLife-style and all that. For hitting their mark I applaud Apple. The products are real works of art---quality in interface design, mid-level performance at an everyman cost. but SGI is trying to sell HPC. The big $$'s means you had better put something *extra* in those machines that the average high-schooler can't get off ebay. Cutting edge technology---because for real science and engineering ALL computers are way, way, WAY too damn slow. NUMA link is good, but the real-time and RASC are only marginal, and the graphics suck.

What's been eating me lately...a growing fear if you will, is an attitude that SGI has been giving me in regard to the Prism. I'm sure it due to limited resources, but I keep bumping into the attitude of either

1) It's SuSe Linux....there's nothing we can do about it.
-or-
2) It's ATI graphics...there's nothing we can do about it.

To which I find myself being forced (and I really do mean forced) into replying: "Then I'll have to find some other vendor that will provided an integrated system that performs." My only promblem is figuring out where to go next for HPC---I just don't have the time to gin something together myself. That should really be SGI's targeted niche---folks with more money than time (or computer smarts ). And the systems they deliver should seem like magic---not a jumbo-size version of last year's PC trash.

One last comment: Could you see Apple saying "It's BSD unix...there's nothing we can do about it?"

\rant

Hey Steve; I agree hundreed percent with you on every aspect of your post. But what I've tried to stablish now looks even clearer:

GeneratriX wrote: I want to talk about the power of the marketing, and how you can make look better what in fact is the same, almost the same, or plainly worst.

"...the same ...almost the same ...plainly worst."

I've only shown the case of Apple, to bring an example of how a succesful marketing managing, can help to clear from their inherent "taboo-registered-trademark" to things like the Intel processors (which by the way I've never liked, but I think they are a big company with mediocre products), and ATI graphics (which I consider decent graphics, but not too outstanding).

But of course! ...Apple is doing very well a lot of things, and their products are WAY COOL! ...As much cool as their department of industrial design, marketing, and management.

GeneratriX wrote: Hey Steve; I agree hundreed percent with you on every aspect of your post.

as i do.
this is by far the best post i've read in a looong time.
they should promote you

foetz wrote: this is by far the best post i've read in a looong time.
they should promote you

Who could promote me to what, Foetz? Geez! ...I'm asking them since months for a damn sponsorship to continue one of my key developments on the Global Development Program PLUS, and I've never received a penny as to at least put my programme on a production phase... and we are talking about something that should interest to a company with the word "Graphics" on their name...

Nahhh... don't be confussed, every cent to bring alive my software projects was originated in my own pocket and from my second job, or from the pocket of generous Nekochaners...

If they never wired transfer a single dollar to support my work, don't think that they will do just to put me to speak bullshit.

GeneratriX wrote: Who could promote me to what, Foetz?

Ouch!; What a stupid asshole I am, Foetz! ...I've not realized until just a few minutes ago that you probably talked about SQueen! ...Well, sorry both guys, not intention to offense to nobody, whatever...
Peace on the earth!
Cheers!

GeneratriX wrote:

GeneratriX wrote: Who could promote me to what, Foetz?

Ouch!; What a stupid asshole I am, Foetz! ...I've not realized until just a few minutes ago that you probably talked about SQueen! ...Well, sorry both guys, not intention to offense to nobody, whatever...
Peace on the earth!
Cheers!

hehe, yeah, indeed i meant squeen
but never mind feel free to promote yourself in your own company

peace as well

foetz wrote: hehe, yeah, indeed i meant squeen

That's what I suspected on my second reading! ...Sometimes seems that there are not enough gaps on the earth as to jump inside! ...Well, seems that I'm * a little bit * sensible about the plain idea to just crop my software line to multiply the time available for my second job, and I'm now finding relations with this matter even on the words that forms those little letter-noodles in my soup!

Oh well, I'm so grieved with these fact, that I act a lot more irritable than normal! ...You know, just not enough time to make everything at once, and I hate to be obligated to cut out one of my developments.

...You have my serious excuses!

foetz wrote: but never mind feel free to promote yourself in your own company

Okay; since today at 00:00 Hours I'll take the position of... ehhrrr... huhh ...mmmhhh so much to choose ...well, whatever! ...I'll just be the guy that works here in the office! ...Call me GeneratriX!

foetz wrote: peace as well

Heheh!; feel free to ignore those fireworks as procedent from my italian-ancestor's DNA!
Peace on Nekochan!
Saludos,
Diego