Re: Hardware Progess: $165/Gflop

From: Dani Eder (
Date: Sat Jan 03 2004 - 18:29:53 MST

Moore's law originally referred to cost optimal
number of components in an integrated circuit
doubling every year. The trend for integrated
circuits subsequently has been about 18 months
per doubling. A 'Moore's Law pace' is therefore
a change of a factor of two per 18 months, and
can refer to things other than the number of
components in an IC.

> unfortunately, does not linearly fit semilog plots,
> and hasn't been for quite
> a while now.

I beg to differ. This chart:

shows that supercomputer performance has in fact
been increasing in just such a fashion.

> How do you know which sort of system is useful for
> AI work?

I don't. I'm guessing that something like attention
and short term memory can be handled by RAM and that
something like long term memory and synapse connection
tables can be handled by hard drives.

I do believe that building an AI requires both
adequate software and adequate hardware. Since
many people on this list concern themselves with
software, I've decided to track hardware trends.

I wonder why you didn't
> include GBit Ethernet in your comparison.

I don't think the motherboard I priced would
support that much LAN traffic.
> > If optimistic estimates of the required computer
> > power for human-level AI are correct at 100
> TFlop/s,
> Tee-hee.

Why do you laugh? That figure comes from Hans
Moravec. While I consider it optimistic, he has
what I think is a reasonable means of coming up
with the figure.
> This assumes that 17 M$ worth of hardware achieves
> equivalent real-world
> performance as a human. In other words, it assumes
> AI is a solved problem.

No it doesn't. Economic crossover means a computer
can perform a specific job at a lower cost than
humans. This has already occurred for tasks like
payroll processing, where all that is required is
simple arithmetic. As more powerful computers
become available, more and more jobs can be performed
by them cheaper than us. Economic crossover for
engineers would occur when a computer can do our
job cheaper than us.
> I notice that no one is currently attempting to
> build self-rep systems,
> despite seminal work being about half a century
> past. It's useless to
> speculate when something can be done, if the
> humanity collectively decides to
> not pursue that particular path.

Although I've been working on the concept of a
replicating factory in the form of a single
unit that can copy itself, our society as a whole
has been working towards automating production.

Imagine if you will that the railroads and shipyards
fully automate handling of shipping containers
(those 8x40 foot steel boxes). That factories
automate the unloading and loading of those boxes
with raw materials and finished products, etc.

Once all the pieces get automated, your entire
industrial infrastructure acts as an automated


Do you Yahoo!?
Find out what made the Top Yahoo! Searches of 2003

This archive was generated by hypermail 2.1.5 : Wed Jul 17 2013 - 04:00:43 MDT