A Stool Bed Time Story About Cpu's

[DogBoy]

I as going to post this as a reply to another thread but I guess I got carried away scribbling and scribbling away pecking at the keys furiously.

 

Then I realize this little story would make a nice bed time story for Bearz out there in Stool Land.

 

It illustrates how American business has changed over the last 30 years and why companies today are not very profitable and probably never will be.

 

And how you'd be better off owning a convenience store than investing in most US companies today. The convenience store is at least as sound business model and a lot easier to understand.

 

This could easily be made into a whole book describing the downfall of American tech as a whole.

 

Sometimes cheaper is not better, it's just cheaper.

 

And once everyone and everything is cheap then noone can make any dough and everyone goes broke and works for next to nuthin which is where we are headed.

 

-------------------------------------------------------------------

 

"At 3GHz or so they have hit the ceiling due to heat and size "

 

And tell me why I would need a 3Ghz CPU to download some internet pages ?

 

And as for gaming who has the time ? Some pimple-faced 15 year olds ? Not me I can tell you that for sure.

 

For an old timer (programmer for 23 years) this whole thing that happenned in the CPU biz is totally hysterical.

 

Let me tell you what I know about the mainframe CPU "game" and how it was played to maximize profts for big corps.

 

Today the CPU chip game has been totally botched to the point where people are getting some really neat stuff for practically nuthin.

 

IBM saw this problem back in the early 70's --- that tech development of mainframe speeds could easily outstrip the marketing people's ability to make $$$ off of it.

 

What did they (IBM) do ? I'll tell you and it certainly wasn't to make faster and faster machines for no good reason at all.

 

They kept new developments "under the cover" for years and only fed customers new developments as they had squeezed as much profit out of the old product lines as possible.

 

Don't beleive me ? Then you have no idea how corporate America thinks (or used to anyway) at the highest levels and probably never will.

 

More than one IBM engineer told me that IBM could market machines 10 times faster any time they wanted to. It was literally one phone call away.

 

Gene Amdahl, the break-away engineer who developed the IBM 360 which was the most successful mainframe system ever designed, knew this all too well and played this CPU metering power game even better than IBM.

 

He (Amdahl) knew that to do otherwise would lead to a MAJOR ass-whoopin from Big Blue. IBM would have ruthlessly

driven him out of business and his creditors would have fled in terror as mainframe prices plumetted much like CPU prices have done today.

 

Then there was a Japanese company called Fujitsu that tried to "mess with" Big Blue my making really fast machines at bargain prices.

 

IBM caught them breaking into some development lab and stealing information on IBM products amd took them to court and tied them up with this for years and years as court costs mounted into the millions.

 

IBM spread all kinds of rumors about how Fujistsu mainframes might crash and may be no fix available.

 

IBM made it perfectly clear that if anyone bought a Fujitsu mainframe they might as well erase the phone number of their IBM system engineer from their address book.

 

Corporations were fearful of buying Fujitsu mainframes because they'd anger Big Blue and not get the same level of support for the MVS operating system (like Windows for Mainframes).

 

The jist of it is that IBM was able to keep any competitors that got out of line (ie. made machines that were too cheap and too fast) bottled up with law suits and make potential customers fearful.

 

A few small firms like Amdahl were allowed to exist by Big Blue because they were well-behaved.

 

Amdahl often even screwed customers worse than Big Blue.

 

The Amdahl V7 had a switch on it which would make it run at a higher speed (it would be like going from 700 Mhz to 1.2 Ghz) but the catch was that whenever a customer flipped the switch into HIGH SPEED that hade to pay xxx $$$ per MINUTE for every minute they were in the high speed position.

 

It would be like Intel having a switch that when flipped to 1.2 Ghz would cost you like 50 cents per minute.

 

The main reason companies bought Amdahl processors was to do real cpu-intensive tasks like solid object modeling (which I worked on) and engineering stuff like that.

 

Amdahl made faster CPU's but the open secret at the time was that Big Blue had CPUs sitting in the lab that were 5 times faster than anything Gene Amdahl could hope to make.

 

All Big Blue had to do was give the order and every mainframe on the planet would have been 5 times faster.

 

Unlike Intel today, they never had to.

 

Intel is a real pitiful company in my view. Anybody who'd invest in that business has got to have holes in their head. It has no future at all and not much of a past either. You'd be better off buying a convenience store.

 

Anyway that was a little stoolie bed time story about the Good Ole daze of corporate America.

 

Someone should write a book about how Intel let the "CPU cat" out of the bag and screwed up a whole industry tuning it into something no more profitable than running the local convenience store.

 

At least convenience stores are a lot easier to understand than CPU's.

 

These days that kind of stuff only goes on in the medical biz and a few other areas.

 

Instead of holding back new products and maximizing profits like in the old daze companies just outright lie about their earnings or throw up smoke clouds so big noone can really tell what they're dong.

 

Thru the genius of marketing schemes and legal strategies American businesses used to be real profitable now they just pretend they are.

 

Good Night Bearz !!!!

[DrStool]

Dogie- You are one of the people who makes this website what it is.

 

G'night buddy!

 

Doc

[mksloth]

I suspect you're being tongue-in-cheek; if you're not, your point about cheaper being just cheaper is a little flawed. BTW, I respect your insights in general. IBM, in its dealings with wannabe competitors, did run afoul of antitrust laws. Since you're focusing on hardware speeds, consider, if you will, the rate of speed change in the IBM mainframe days against the latter period. It should be obvious that IBM was throttling not just the market, but also innovation. I'm sure you remember the PDP-11 and the VAX- things that helped make much of the platforms you and I are using right now. IBM had diddley squat to do with that, and actually did its level best to curtail these innovations.

 

But, getting back to hardware, it was hardware advancements and corresponding price decreases that brought about the PDP-11 and the early forms of ARPAnet. So your thesis that cheaper is just cheaper doesn't quite work. At least, not in the general case. As for the PC and Intel, without getting into whether they're going to ultimately survive or not, they've done an admirable job of emulating IBM in its hegemony days. The only real competitor I can see to Intel is AMD, and they're bleeding from practically every orifice. All the other competitors have been put to pasture with tactics not too dissimilar from those of IBM's. And I won't rehash the history of IBM once its beloved mainframe monopoly was found to be a dwindling game. Suffice it to say that IBM is now trying to make numbers with the help of its "rent-a-researcher" program...

 

What is going on right now is similar to what's been played before. The game is changing from raw speed. And everybody is scrambling about trying to figure out what's next. The previous inflection point was the realization that "IBM mainframe compatible" wasn't necessary. At that point too, there was a similar scrambling about.

[DogBoy]

Yes, Tongue in cheek for sure.

 

I was NOT advocating with the corrupt practices of the past or the corrupt practices of the present.

 

Just discussing the differences.

 

Companies used to lie about and manipulate their products and potential compeitors.

 

At some point hey must have found out it was alot easier to just lie about and manipulate earnings.

 

There is an argument to be made that from the point of view of shareholders and employees the old way was better because the companies REALLY did make nice profits (albeit by screwing competitors and customers) and employees had some job security as well.

 

Old corruption, new corruption ?

 

I dunno.

[DogBoy]

As for the point you made about DEC CPU's they (DEC) were in a different ball game entirely.

 

The difference between DEC and IBM was like the difference between Microsoft and Apple.

 

IBM sold to the fortune 500 corporate market whereas DEC went after market niches.

 

By 1980 most business software was written for the IBM 360/370 machines. Most add-on software as well. It was far and away the largest software base.

 

For a company to switch to DEC from IBM woud be like switching from Windows to Linux or Apple.

 

It would require a very large training expense not to mention the purchase of new software and hardware (in the case of Apple).

 

To 99.9% of companies today it would be unthinkable.

 

Just as it was to companies in the 70's and 80's to switch from IBM to DEC.

 

So any comparison of price-performance of DEC and IBM mainframes of that era is not really relevant.

 

And besides DEC was probably doing the same thing as IBM (maybe to a lesser extent) as far as "metering" new technology to customers.

 

By the way the same sort of thing went on in the auto business untilt eh Japanese came along.

 

New innovations in the car biz were suppressed for many years in order to maximize profits. This is amply documented.

 

Somewhere aong the line (early 90's) some smart Harvard MBA must have got the brilliant idea that it would easier to manipulate the stock price than to manipulate the "real market" (customers and competitors).

 

Plus by then most industries had gone global and it would be nearly impossible to pull of a manipluation on that kind of scale.

 

But Wall Street was still a very "clubby" place as it still is today and manipulating the share price directly was easy.

 

Anyway this would make for a very fascinating book.

[mksloth]

The problem with metering innovation is that it typically ends - in a bad way. My point about DEC was to refute the notion that cheaper is only cheaper. The fact that DEC came in and carved out a none-too-shabby niche for itself in an era when "nobody got fired for buying IBM" clearly supports this. Granted, DEC boxes got into the market not in traditional business uses as did the big iron from IBM. However, they did grow to capture a significant amount of the business marketplace. Part of the problem of metering one's innovations is that the company concerned cannot satisfy all its potential customers. This is because the company has typically painted itself into a one-size-fits-all situation that doesn't allow it to use shelved innovations to solve problems for customers for whom its existing products would be a sledgehammer. This is what DEC and others exploited pretty well - at least in the beginning.

 

In some ways the mainframe software market is probably still pretty large - I don't have all the numbers in front of me. However, the cheaper and quicker motto was what opened new markets.

 

Completely agree about the bozo from HBS (or wherever) that it was easier to game the stock than the marketplace. That said, companies still lie about their products vis-a-vis their competitors. Wasn't it IBM that coined FUD?

 

PS:

Just so you don't get the impression that I'm for DEC and against IBM. No - I used to hate DEC systems as I'd grown up on a diet of IBM 370 and Unix. And I have great respect for IBM's work.

[Hypertiger]

Right now due to a "relocation event" I am using my windows Me 233MHz 64 MB RAM machine and the difference is only fractions of a second from my AMD 2000 XP 256 MB RAM. Hell a 486 DX with 32 MB could cut the mustard... I've been telling people for years that processing speed is basically toped out at the 500-700MHz level. Anything after that is a waste of money... Minimum system requirement for Windows XP is a Pentium 233 128 MB RAM 6GB HD which is about 100 dollars US plus monitor is maybe 200 dollars, add an OS for 100 and thats 300 US dollars max. 3 GHz watercooled to 3.5 GHz? The world is ending, but just look at my framerates... Ha ha ha ha...

[Whadda I Do Whadda I Do]

Walmart is now carrying no Windozes and no Intel inside machines for around 200 bucks. Should make a nice upgrade for most.

 

AMD has been under Intel's thumb for so long, it's amazing that they are still alive. Everyone else has been crushed and recycled including Alpha.

 

But wait, AMD has only 1 huge fab plant albeit ultra modern and has been running in the top ten list of patents granted, for the last 2 or 3 years.

 

If innovation has anything to do with it I would expected a run up of their stock around springtime just before the market sails off the edge world and I can finally dump what I have and buy a pooper scooper.

 

Newisys might be of interest to the old timers as they are laying foundation for distribution of a superior line of frames and have not gone IPO yet.

[BAREister]

No wonder Buffet invests in companies that make things like razor blades, which will NEVER be made obsolete.

 

we're going to pay a frightful price for this obsession with tech and puters on Wall St, in our weapons systems, etc.

 

You've got types like Gene Inger running around making proclamations on hotlines to the effect that tech may be the salvation of our economy.

 

yeah, right.

 

HRFF hASS said this bear won't be over until "tech" is a FUR-letter word, and widely hated on Wall St.

 

Tech stocks can get halved yet again....and WILL thinx this observer.

 

Time will tell.

[ThorAss]

I think razor blades have been obsolete for years. Why spend the time and effort shaving in front of a merely. I whip out the cordless and shave in the car on the way to work or wherever. I'm sure there are ways to make my cordless obsolete as well. Just a matter of time.

 

Oh and beards, beards are another way.

[DogBoy]

Last year I bought a used Dell 17" Diamond Scan in mint condition at a sale held by my employer for $40.

 

I guess they don't plan on re-hiring any of the hundreds let go.

 

Used computer equipment never did have much resale value and now it's practically worthless.

 

The place I worked said it was cheaper for the company to off load these monitors (which were in new condition --- hardly used) to employees than have to pay a trash guy $$$ to pick 'em up.

 

Anybody who spends hard earned dollars on new computers had to be a real knuckle-head or else they're just rtying to impress their lady friends, I dunno.

 

And as for new innovations they're mostly useless frills.

 

Almost everything that can be done has been done with both hardware and software. There is absolutely no future in it whatsoever.

 

Ok maybe a LCD monitor would be nice to prevent my brain from gettin fried by the crt waves.

 

But vitamins are still cheaper so I'll just pop some vitamins and sit in front of my $40 CRT and running my AMD K6-2 Clone and watching my PUTFOLIO zoom.

 

I'd rather blow $1000 on a gourmet dinner than a stupid computer.

[Hypertiger]

the motherboard in my computer right now is from a garbage bin.

 

3-D Games are all thats left... And gaming rigs are a custom deal...

 

You all better jump on the Unreal 2003/Everquest II band wagon if you don't want to be left behind.

[DogBoy]

Hyper I work on computers all day for the last 23 years and am too pooped to play games when I come home.

 

It takes all my effort just to write these scribblings in the Stool site now and then.

 

To me computers are a living not a game.

[Hypertiger]

The escape from reality industry is very profitable; Gaming rigs are the real profit makers for me. 3-D First person shooters generally need systems that command the highest mark-up. I can build a system for upto 50% cheaper with very little performance loss and can be upgraded when the spare parts are dirt-cheap.

 

It boils down to producing the best frame rates, that is the computer industry. The rest is just smoke and mirrors to subsidize the game players in the frame rate wars?

 

It?s all a scam, you know it and I know it, right?

[Jorma]

Well this is old hat but most CPU's are running at about 2% of capacity. The wasted overhead is used to power the productivity and inflation stratistics. No more work is being done overall or at a faster rate but it makes for a hell of a good story.

 

People love stories.