Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.1 6/24/83; site vax135.UUCP Path: utzoo!watmath!clyde!cbosgd!ihnp4!vax135!miles From: miles@vax135.UUCP (Miles Murdocca) Newsgroups: net.arch Subject: Re: Optical Computers Message-ID: <1333@vax135.UUCP> Date: Wed, 19-Mar-86 11:13:43 EST Article-I.D.: vax135.1333 Posted: Wed Mar 19 11:13:43 1986 Date-Received: Fri, 21-Mar-86 04:20:03 EST References: <346@sask.UUCP> Organization: AT&T Bell Labs, Holmdel, NJ Lines: 83 >>Some time ago, someone posted an article saying that someone was going to be >>unveiling an optical computer in (I think) Hanover or Hamburg. Could someone >>please repost the article or send it to me? Thanks a lot. >>If anyone else has any references to optical computers, optical bistable devices >>etc. I would appreciate them also. Thanks again. Here it is: >Path: vax135!houxm!whuxl!whuxlm!akgua!gatech!seismo!cmcl2!phri!roy >Message-ID: <2258@phri.UUCP> >Date: Thu, 27-Feb-86 19:29:50 EST >Reply-To: roy@phri.UUCP (Roy Smith) > About a year ago (I think) there was a smattering of speculation >about optical computers in this newsgroup. Well, the following just >caught my eye. Don't ask me for more details, you now know as much about >it as I do. Quoting from "Digital Design", February 1986 (Copyright 1986 >by Morgan-Grampian Publishing Company): > OPTICAL PROCESSOR MODEL -- A functional model of an optical > processor will debut at March's Hanover Fair in West Germany. > Offering potential picosecond speeds for parallel architectures, > the model results from work carried out by the European Joint > Optical Bistability project. >-- >Roy Smith, {allegra,philabs}!phri!roy >System Administrator, Public Health Research Institute >455 First Avenue, New York, NY 10016 I wouldn't rush out to Hanover to see a picosecond computer because the picosecond machine does not exist. The devices are interference filters made with ZnSe that operate at 100's of microseconds. The project is under the direction of S Desmond Smith, a professor of physics at Heriot-Watt. He never made the claim that his computer runs at picoseconds but somehow the press must have picked up the phrase "potential picosecond speeds". A stepper motor is used to move the beam to different parts of the device because the intensity of the beam destroys the material. Here at AT&T Bell Labs, we have an effort that has much more promise. We already have samples of optical devices that run at 30ns (see reference 4 below) and another device waiting in the wings that operates at 70ps (see reference 3). Both of these devices (the Multiple Quantum Well SEED and the nonlinear Fabry-Perot etalon) are grown with molecular beam epitaxy (MBE) which is an expensive process that few companies can afford. I can't give particulars about the state of the devices or the method of processing beyond what you will find in the published articles because of proprietary interest. You will find references 1 and 2 of particular interest because they go over the computing aspects of optical computing. My full-time job at AT&T Bell Laboratories is to find methodologies for programming optical computers, so I would be glad to trade ideas with anyone who has an interest. I can only talk about what material has been released for publication outside of AT&T. Note that there are only a few groups who are working on *digital* optics. Much of the work on optical computing has been in the *analog* domain. Miles Murdocca, 4G-538, AT&T Bell Laboratories, Crawfords Corner Rd, Holmdel, NJ, 07733, (201) 949-2504, ...{ihnp4}!vax135!miles 1) A. Huang, "Parallel Algorithms for Optical Digital Computers", 10th International Optical Computing Conference, IEEE catalog number 83CH1880-4, pp13-17. 2) A. Huang, "Architectural Considerations in the Design of an Optical Digital COmputer", Proc. IEEE, July 1984, pp780-786. 3) H.M. Gibbs et al, "Room temperature exitonic optical bistability in a GaAs-GaAlAs superlattice etalon", Appl. Phys. Lett., vol 41 no. 3, pp221-222, Aug. 1982. 4) D.A.B. Miller et al, "A novel hybrid optically bistable switch: the Quantum Well Self Elecrto-optic Effect Device", Applied Physics Letters, vol 45, July 1, 1984 pp13-15.