[Beowulf] Moores Law is dying

Robert G. Brown rgb at phy.duke.edu
Tue Apr 14 10:00:27 PDT 2009

On Tue, 14 Apr 2009, Steve Herborn wrote:

> As far as "Desktop" machines go there hasn't been an application invented
> that needs more.  Because memory & disk storage prices fell programmers got
> sloppy & crammed in a lot more, but little to none of it was actually an
> application that truly needed more because of purpose, only poor design.

I disagree, although I think that there is an issue with lag between
capabilities and software that really uses those capabilities.

AI designs (ones where the OS/UI might exhibit something like real
"intelligence") are open ended in their computational requirement.
Really smart designs are unlikely to emerge without "unlimited" amounts
of memory and processing power, though, as there are serious problems to
be overcome.  Voice recognition, UIs that can talk and be verbally
controlled or controlled by novel interfaces (I'm waiting on my neural
implant:-) are all very processor intensive.  Graphics, and graphical
processing (e.g. OCR etc) on graphical input.  VM applications --
running native linux and a single application UNDER Windows running as a
"linux application" through the VM -- all requires a whole lot of
processor, roughly a core per VM minimum to get really smooth operation.
Encrypted networking.  Real time human connection ware.  And for the
love of god, death to the pixelated display -- give me transparently
rescalable UIs without artifacts.

And ALWAYS:  games.  High end RPgames are desktop applications (not
business desktops, but desktops nonetheless) and can use "infinite"
processing power to produce an ever more faithful simulation, ever
smarter NPCs.  Games alone can and do suck up most of the advances in
CPU and memory and network capability we've seen over the last decade.

So, it's not "only" poor design, although I agree that a whole lot of
standard daily business workflow by your average run of the mill
business employee can be done on a 400 MHz CPU with at most a few tens
or hundreds of MB of memory.  Other desktop work (and play), however,
cannot.  Even "just" decoding music and video while doing other work in
parallel is rather processor intensive, although parts of it are
nowadays offloaded to custom ASIC processors, DSPs, etc.

Part of the problem is that the primary GUI and application developers
have absolutely no real imagination, and are content to just keep on
using pixelated X rather than tackle the complexity of rescalable
everything, dithering, etc.  In Windows, in Linux.  I don't know for
sure about the Mac -- historically it has done better but not terribly
well.  Is this "poor design"?  In a sense yes, but it's more laziness
and the fact that application space is probably a decade behind the
capabilities of system hardware.

Otherwise, when I sit down at my laptop and it SEES me sit down (before
I touch my keyboard, recognizes that it is me, offers me a verbal
challenge and matches my voiceprint and appearance and maybe fingerprint
as authentication, then pops up my standard set of morning "stuff"
without being told, when I can do things like say out loud "could you
just delete all the SPAM, please" and have it do so, offering up a
handful of things it isn't quite sure about and saying "What about
these" out loud and doing the right thing when I reply "Dump them", when
I can just say "Could you please encrypt the following workspace on
disk" and have it be so, when I delete a file and say "damn, I didn't
mean to do that" and have the system reply "well, I didn't actually
throw it away yet, would you like it back?" all in ENGLISH without any
KEYBOARD OR MOUSE -- then I'll conclude that it has enough desktop

"Could" it do all of that now?  I think not, not the way I'd want it
done.  A very crude version, perhaps, with preprogrammed actions and
verbal triggers.  Nothing like Turing test stuff.


> [IMAGE]                          Steven A. Herborn
> U.S. Naval Academy
> Advanced Research Computing
> 410-293-6480 (Desk)
> 757-418-0505 (Cell)
> ____________________________________________________________________________
> From: beowulf-bounces at beowulf.org [mailto:beowulf-bounces at beowulf.org] On
> Behalf Of Bruno Coutinho
> Sent: Wednesday, April 08, 2009 11:28 PM
> To: richard.walsh at comcast.net
> Cc: beowulf at beowulf.org
> Subject: Re: [Beowulf] Moores Law is dying
> I think that even if they stop scaling down size of desktop processors due
> lack of interest in more performance,
> someone will continue doing it (even at a much slower rate) for HPC market.
> No matter how much computing power future processors will have, someone will
> invent a application that needs more.
> 2009/4/8 <richard.walsh at comcast.net>
>       ----- Original Message -----
>       From: "Ken Schuster" <ken at kschuster.org>
>       To: beowulf at beowulf.org
>       Sent: Wednesday, April 8, 2009 2:29:17 PM GMT -05:00 US/Canada
>       Eastern
>       Subject: [Beowulf] Moores Law is dying
>       >An IBM researcher says Moore's Law is running out of gas. IBM
>       Fellow Carl Anderson, who
> >oversees physical design and tools in its server division, predicted
> the end of continued exponential
> >scaling down of the size and cost of semiconductors:
> >"There was exponential growth in the railroad industry in the 1800s;
> there was exponential
> >growth in the automobile industry in the 1930s and 1940s; and there
> was exponential growth
> >in the performance of aircraft until [test pilots reached] the speed
> of sound. But eventually
> >exponential growth always comes to an end," said Anderson.
> Mmm ... he may be right, but I do not like his historical references
> which seem
> to conflate engineering and economics.  Better to refer to the
> improvement in
> magnets or something similar.  But, I like the speed of sound
> reference because
> it suggests that there is a Moore's Law barrier to be broken.  There
> is a lot of
> talk about "walls" these days ... the memory wall, the power wall, ...
> but we with
> respect compute power we have a ways to go before we reach the
> Bremermann Limit.
> rbw
> _______________________________________________
> Beowulf mailing list, Beowulf at beowulf.org sponsored by Penguin
> Computing
> To change your subscription (digest mode or unsubscribe) visit
> http://www.beowulf.org/mailman/listinfo/beowulf

Robert G. Brown	                       http://www.phy.duke.edu/~rgb/
Duke University Dept. of Physics, Box 90305
Durham, N.C. 27708-0305
Phone: 1-919-660-2567  Fax: 919-660-2525     email:rgb at phy.duke.edu

More information about the Beowulf mailing list