Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!mnetor!uunet!seismo!rochester!PT!k.gp.cs.cmu.edu!lindsay From: lindsay@k.gp.cs.cmu.edu (Donald Lindsay) Newsgroups: comp.arch Subject: Re: how fast could disks be (Mach 1) Message-ID: <1192@k.gp.cs.cmu.edu> Date: Sun, 23-Aug-87 19:02:41 EDT Article-I.D.: k.1192 Posted: Sun Aug 23 19:02:41 1987 Date-Received: Mon, 24-Aug-87 00:06:37 EDT References: <12191@amdahl.amdahl.com> <1189@k.gp.cs.cmu.edu> <2316@mmintl.UUCP> Organization: Carnegie-Mellon University, CS/RI Lines: 28 Keywords: laser disk scan Mach >>The first commercial units will spin, but eventually, we will just scan the >>laser beam across an unmoving disk. >Doesn't this severely limit the ability to stack more than one disk surface >in a device? How close to the disk does the scanning head have to be? How >does this depend on the disk size? An unmoving read/write surface would impose geometrical limitations. However, I presume that we will eventually be able to scan beams far faster than we can ever hope to rotate any (large) object. Plus, we eliminate the disk motor, the bearings, the voice coil (or whatever), the comb (and its bearings). We would expect improvements in weight, power, reliability, component count, ruggedness, manufacturabilty, seek time, and transfer rate. For that, I would give up some density. As for HOW we do it: I believe it is too soon to say. I read "Laser Focus", but I probably misunderstand much of what I read. I get the impression that there are several ways to bend a beam under electronic control. Also, we can now modulate the color of certain lasers: who knows where that leads ? Plus, semiconductor lasers benefit from chip technology, and are most definitely still improving. Eventually, "disk" will just be "nonvolatile memory". Trust me. I do not, repeat, not, make misteaks. -- Don lindsay@k.gp.cs.cmu.edu CMU Computer Science -- Don lindsay@k.gp.cs.cmu.edu CMU Compu F F uld