Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!swrinde!zaphod.mps.ohio-state.edu!rpi!hiebeler From: hiebeler@cs.rpi.edu (Dave Hiebeler) Newsgroups: comp.theory.cell-automata Subject: efficiency and philosophy of CA computer simulation Message-ID: Date: 2 Mar 90 22:18:55 GMT Organization: RPI CS Dept, and LANL Center for Nonlinear Studies Lines: 81 Two things recently said here led to a few interesting thoughts. First, the quotes: 1>From kyriazis@iear.arts.rpi.edu (George Kyriazis) Fri Mar 2 16:48:38 1990 1> Will CA's be eventually fast enough to accurately simulate such a 1> circuit...? and 2>From karakots@apple.com (Ken Karakotsios) Fri Mar 2 16:48:44 1990 2> Wire world running on a serial computer is many orders of magnitude 2> slower than current logic simulators. One of the strong incentives to use CA for things like physical modeling, is that the path between the physics and the computer simulation is much more direct, i.e. instead of following the following steps: 1. make a continuum approximation of the system under study 2. write some PDE's 3. turn the PDE's into finite-difference equations on the computer 4. do lots of rounding off in the process of simulating these finite-difference eqns We just do the following with CA: 1. "assume" the universe (hence the system under study) is discrete 2. write a CA rule which captures the essentials of the laws of physics, in particular the ones that most strongly affect what we're studying, and implement the CA rule exactly on a computer (no round-off, etc) Because the CA approach is more direct, if you build a computer (e.g. a CAM) using the CA philosophy (local connections, very parallel) you will be able to do physics simulations very fast, although from a very different perspective than PDE's. Anyways, enough background, which everyone here has heard many times (I quickly repeated it for the benefit of new people). What the 2 quotes above got me wondering about was the efficiency of using CAs to simulate traditional-architecture computers. In some sense, it has to be efficient, because the traditional computer will be implemented in our physical reality, which a CA corresponds well to. But is there some advantage to be gained by using a standard (physically inefficient) serial computer to simulate another such computer? I think not, with the continuing development of CA hardware. The folks working on CAM-8 describe future CA machines as being "programmable matter". We will be able to do such incredibly detailed physical simulations, it's almost like having a "holodeck" from the new Star Trek -- we'll be able to create some virtual physical system, in some sense. I'm not getting into the virtual reality topic here; I'm saying that at some point we'll be able to do things like simulate a cubic centimeter of air at the molecular level, in something not far from real-time, using CA technology! So a *CA simulation* of a serial computer will approach the speed of a *physical implementation* of that computer! At least, it seems that way to me. This really starts to sound like a lot of philosophical hocus-pocus, but now that I've already gone that far, I might as well ask the real philosophical knockout question: do you think it is feasible that we will someday be able to simulate computers, using CA techniques, so that they will run *faster* than the physical implementation of the machines? That is, maybe CA hardware will develop so well that it will get ahead of manufacturing technology in some areas? I really don't know anything about manufacturing technology, so maybe it's a ludicrous question. It does seem like an interesting one though. I guess it's safe to say we can't simulate the universe more efficiently than it "runs" itself, since our CA machine will be embedded in the universe. But it does seem theoretically possible, I think, that we could "beat ourselves" at building things, so to speak. -- Dave Hiebeler / Computer Science Dept. / Amos Eaton Bldg. / Rensselaer Polytechnic Institute / Troy, NY 12180-3590 USA Internet (preferred): hiebeler@turing.cs.rpi.edu Bitnet: userF3JL@rpitsmts "Off we go, into the wilds you ponder..."