Jump to content

Putting computer prices into perspective


deanmass

Recommended Posts

  • Members

Hey Dean,

 

I remember getting my first PC back somewhere around 1989. I think it cost about $900 and it was a black screen and green lettering. Then around 2 years later I got a 386 with Windows 3 (or something) for $1200. That was a "killer" deal because not only did I get a monitor and printer with that set up but Windows was in color! From there I went to a G4, presently a G5 and who knows whats next...

 

I think the $2500 in 1984 is probably closer to $6000 or even $8gs today with the dollar being what it is.

Link to comment
Share on other sites

  • Moderators

And yet, with all this horsepower, for the most part the computers of today aren't that much faster (in terms of doing something) than the computers of yesteryear.

 

Software bloat, inefficiency, and complexity always expand to take up all available computing power. :idk:

 

Terry D.

Link to comment
Share on other sites

  • Members

I have an original 1984 Mac, with a third party SCSI card and 10 Meg external drive. Still works. And, any one on this forum could still run it, since the OS behaves very similar to XP and OS X.

 

I came across some plans where you rip out the guts and turn it into an aquarium, but I just don't have the heart to do that.

 

js

Link to comment
Share on other sites

  • CMS Author

 

And yet, with all this horsepower, for the most part the computers of today aren't that much faster (in terms of doing something) than the computers of yesteryear.

 

Myabe the yesteryear of five years ago, but I recall reading a story in Mix when Mix was a real magazine, about a deconvolution algorithm run against a recording of a poem that Jim Morrison recited to someone over the telephone, from a phone booth. They wanted to get rid of the phone booth ambience and use it on a record. The computer that processed it (it was probably a mainframe or a good sized minicomputer - I don't remember what) ran all night to calculate the new file that was less than two minutes long.

Link to comment
Share on other sites

  • Members

I have an original 1984 Mac, with a third party SCSI card and 10 Meg external drive. Still works. And, any one on this forum could still run it, since the OS behaves very similar to XP and OS X.


I came across some plans where you rip out the guts and turn it into an aquarium, but I just don't have the heart to do that.


js

I found a Classic II at a pawn shop and did that; put the light in a 5.25" floppy drive enclosure.

 

DSCN0193.JPG

Link to comment
Share on other sites

  • Members

And yet, with all this horsepower, for the most part the computers of today aren't that much faster (in terms of doing something) than the computers of yesteryear.


Software bloat, inefficiency, and complexity always expand to take up all available computing power.
:idk:

Terry D.

 

Huh? That's nonsense. I can do things on my laptop that would choke my desktop from 7 years ago. I can use many more plug-ins at the same time than I use to.

Link to comment
Share on other sites

  • Members

My first computer was an Amiga 500 with the Ram expansion and an external floppy drive...The whole thing set me back about $1200.

 

It was better than any mac or PC at the time...and, I probably got more done on it than any other machine I have had with the possible exception of my SE Midi Rig...lack of ram can keep you focused:)

Link to comment
Share on other sites

  • Moderators

Huh? That's nonsense. I can do things on my laptop that would choke my desktop from 7 years ago. I can use many more plug-ins at the same time than I use to.

 

I think audio apps are an exception, because manufacturers actually make an effort to optimize their code.

 

In stark contrast, try any Microsoft Office product, a prime example of bloated code. I can show you some (large) documents that take more than 5 min to open running at 100% CPU. WTF?

 

There's no question that computers today have a huge amount of horsepower in terms of megaflops or whatever metric you want to use. The problem is that most software is rushed out with no concern for how efficient it is.

 

Even DAW software (like your plug example) should be much, much faster. Using Audition 3, I can have about 5-7 inserts of Drumagog working in real time before I run out of CPU. I'm running a dual core 3.0GHz processor with 4 GB of RAM, 800MHz DDR2.

 

That's HUGE horsepower even compared to some old mainframe supercomputers I've used. It could do a lot more, if the code were written tighter. :idk:

 

What I'm saying is that when processing power doubles, application speed nowhere close to doubles. And that's without even bringing I/O bottlenecks into it. Which, btw, why do my brand new SATA hard drives rotate at 7200 RPM when my 10 year old SCSI drives in my MX2424 rotate at 10,000 RPM? :confused:

 

Only going to get worse now that we've hit a speed limit on processors and have to resort to dual and quad core parallel processing. It's really, really hard to write code that makes meaningful use of parallel processors, not to mention the problems with multiple processors sharing the same physical memory.

 

Terry D.

Link to comment
Share on other sites

  • Members

My first computer was a Micron with Seagate Barracuda drives that cost $600 and had 2G. Now my USB flash drive has 8G and I purchased it for probably under $35.

 

The irony in part is what Terry is mentioning about bloat. Obviously, the applications are doing much more, but they don't run nearly as fast as the blazing computational speeds of the computer might suggest.

 

Where you can get this back sometimes is if you upgrade your CPU to the max amount some time later. You can get some pleasantly fast speeds in your apps then since the computer has considerably more horsepower than the apps were initially written for. At least, that's been my experience.

Link to comment
Share on other sites

  • Members

There are a number of sectors that optimize - gaming tends to be a prime example - there you can see some pretty crazy tricks that are counter to classic structured programming (the trade off can be readability, portability, flexibility, etc) concepts

 

basically, the more specialized the app and more focused the use model, the more we can optimize for that use model (on the minus side, come off that use model, and performance can tank)

As we have more general purpose apps, it can be harder to optimize on the application level

 

in a way, a look at computational efficiency as sort of area-under-the curve type view could give us some better accuracy

 

on the algorithmic level, we tend to get better and better

 

one problem on the modern app level is that the personal/"desktop" computer actually is doing a crapload and it's stuff we've come to expect (GUI and keeping that updated. multi-thread management, etc) and a lot of it is to make it palatable for the user

we often are using computer as "information appliance" as opposed to linear data canon -- the funny thing there is we are often asking computers to do things humans are good at, but computers aren't particularly good at (so sometimes we are insensitive to how hard some of it is)

 

and some of those problems we try to crack aren't particularly well defined --so we have crcks in heuristic attempts more than algorithmic inefficiency.

 

It can be kind of a trick of how we are interpreting things (someone on HC had a great sig a while back.."On the cellular level I'm actually quite busy") and the modern Personal computer can be subject to that sort of interpretation too...it's kind of a matter of what the requirements and acceptance expectations the project have on them -- and what group of problems are actually being addressed

 

I neat term I think takes into account slick management of an engineering problem is "elegant".

 

Now, there is still the problem of non-linear Big O (asymptotic time complexity) -- and we can be insensitive that too...even basic ideas, like sorting or route optimization, grow in non-linear fashion, so as we pump more data around, the time complexity can grow much faster

and that's a raw nature-of-the-beast problem.

Link to comment
Share on other sites

  • Members

I found a book from the 80s about buying a computer. List price for a keyboard was about 800. Memory chips were like 100 per k.

 

Found the box for some 30 pin SIMMs, I think it was 8 or 16MB, in my friends attic last month and the price sticker was about $380.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...