Re: 1teraflops cell processor possible?

"Robert Myers" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
snip
>
> Let's see.  How many things will I be accused of being?  Hysterical.
> Alarmist.  Hyperbolic.  What have I left out?
>
> There are some things about infectious diseases we already know:
>
> 1.  The infrastructure of vast portions of an entire continent,
> Africa, is on the verge of collapse on a scale that will make any
> other modern human catastophe pale by comparison because of HIV.  The
> glowing economic future of China may not ever materialize because of
> HIV.

One could point out that HIV can be controlled by behaviour
modifications or draconian societal measures more easily than by
scientific miracles.

>
> 2. Infectious diseases arise as if out of nowhere and because of
> modern transportation girdle the globe in a matter of days.  It is
> astonishing that HIV hasn't transformed itself into something more
> durable, virulent and deadly because of the frequency with which it
> mutates and reproduces itself.  It seems as if the SARS epidemic was
> an exceedingly close call.

Maybe we need more isolationism rather than computers built out of
unobtainium.

>
> 3. It is clear that infectious diseases can be used as weapons of mass
> destruction and that people have made preparations for using them in
> that way.
>
> 4. There are now bacterial infections that are incurable because there
> are strains of bacteria that are resistant to all known drug
> therapies.  It is not impossible to imagine reaching the point that we
> will simply not be able to operate hospitals because they are such
> effective repositories for such bacterial strains.

Hospitals can be operated.  Today's hospitals are really surprisingly
primitive, compared to what folks know how to do.  Compare the
cleanliness of an operating room to the cleanliness of a semiconductor
plant.  Do operating rooms these days have have laminar airflow with
HEPA filters?  If they don't they ought to.

>
> While I don't want to turn this into a tussle with Mr. Hinds or anyone
> else about what is and is not possible, the fact is that we know the
> ab initio equations that apply, and we know how to solve them.  The
> kind of grimy guesswork that Mr. Hinds is familiar with is
> necessitated not by any lack of fundamental understanding, but by a
> lack of raw computing power.

We don't usually design circuits by solving Maxwell's equations.

>
> Yes, you have to make choices.  If you spend money on very fast
> computers and on very basic research, that is money that will not be
> available to be spent on front-line clinical research that probably
> will save lives against known threats faster.

Interesting question, whether to spend the money on projects with
immediate payoff or go for the home run.  Cancer research is analogous.
>
> In fact, the most effective ways to slow the spread of the AIDS virus
> have very little, if anything, to do with _any_ kind of cutting-edge
> research or indeed even with ordinary clinical medicine.
>
> The irony is that, as Mr. Herbert has pointed out in a different post,
> there are people who understand the immense, almost unimaginable,
> opportunities for profit that exist and who are willing to venture
> large sums of money in the hope of capturing some of that potential
> profit.  Those with great faith in free markets will see it as a sign
> that the problems I have mentioned should be left to the inventiveness
> of free markets.

Until the people demand price controls.  As it is, it seems as if the
americans and to some extent the europeans are funding these advances.

>
> I have great faith in free markets, but there are some problems that
> cannot be left to them, and public health is one of them.
>
> It is possible to pour *so* much money into a given area of research
> that it becomes like pushing on the end of a rope.  Research into HIV
> has sometimes seemed like that because research on that virus diverted
> a finite pool of competent talent away from other important problems
> like cancer.  Because of the vast explosion in infrastructure that has
> arisen, partly in response to the AIDS epidemic, I do not believe that
> such a thing is currently happening.
>
> The opposite circumstance is manifestly the case in semiconductor
> physics, computer architecture, computer science, and related fields.
> There are more competent bodies than there are jobs.  Let's put that
> vast pool of talent to work.

OK, let's see some evidence that these radical architectural concepts
will actually work better on the problems people are interested in.  I'm
an agnostic about wizbang new architectures.
>
> I will now go to a safe place and wait for the incoming artillery.

Not from me dood.  If you can show it works better you ought to be able
to get it funded.
>
> RM