MIME-Version: 1.0
Server: CERN/3.0
Date: Sunday, 01-Dec-96 19:13:15 GMT
Content-Type: text/html
Content-Length: 4857
Last-Modified: Friday, 03-May-96 19:03:40 GMT
The Real Deal of the $500 Network Computer
The Real Deal of the $500 Network Computer
by MULVA Inc.
...And just what is this madness???
With all of the hype surrounding the new network computer and platform
independent programming languages, such as Java, we decided to design our own
network computer to see what type of hardware and software we would
get using at the $500 price point. A $500 computer will not contain
many things which we take for granted when discussing personal
computers. For example, a hard drive becomes optional equipment, as
does a keyboard and mouse. By removing these devices presents an
interesting problem and raises many other questions. We want to find
out what these problems are and obtain a solution for them. In order
to do this, we must first design a network computer of our own to
discover exactly what $500 buys.
We priced it using off the shelf, readily available hardware. Our
implementation is only theoretical because we did not actually build
a working prototype (mainly because computer scientists and that's not
what we do). The main purpose of this exercise is to get an idea of
the performance obtainable from a network computer. We want to
determine how much of the network computer hype is real and how much
of it is marketing. For more information, refer to the NC Specification document.
...and what do we learn from this?
A lot. We learn how to estimate the performance of a system that
currently does not exist. This is a valuable skill because this type
of evaluation must be done on a new project of this type to see if the
end result will be usable. If the result product is not usable or the
performance is less than that of what you can currently get, it will
not sell. With this in mind, we want to know if the Network Computer
will revolutionize the computer industry or just be another fad that
will quickly die. Also, there are some other social aspects to this.
For example, if the machine does not contain a hard drive and you
receive electronic mail, where is the mail stored? This would be a
new paradigm for the computer industry and we are trying to determine
whether the NC is technologically superior to a personal computer or
if it is a lot of hot air.
...and how does it tie in with
High Performance Computer Architecture?
We trying to see if this computer is high performance at
all. From what we can tell, we can't see that this type of computer
provides at least the same performance as other computers in the same
price range. It is not clear that this computer will perform at a
speed which closely resembles a modern personal computer.
...and how do we measure success?
If we can get data that conclusively shows the performance of a NC, it
will be possible to estimate the overall performance of the system.
For example, if we use a NC with an ethernet connection browsing the
WWW, will it have the same performance as a personal computer browsing
the WWW. These are just some of the questions that must be answered.
...and finally, how are we progressing?
- 3/28 : Project Proposal
- 3/31 : Price the $500 computer
- 4/01 : Modify Web Browser
- 4/04 : Create Web Project Web Page
- 4/05 : Distribute Survey
- 4/12 : Gather data from web browser
- 4/18 : Checkpoint
- 4/26 : Create poster & Final Report
- 5/02 : Poster session
MULVA - Name of Jerry Seinfeld's girlfriend which rhymes with a
part of the female anatomy.
Courtesy of the "Junior Mint" episode of
Seinfeld.
Visitors since April 17th, 1996