ram@elmgate.UUCP (Randy Martens) (09/08/87)
In article <723@elmgate.UUCP> jdg@elmgate.UUCP (Jeff Gortatowsky) writes: >.. I choose to encourage people like RMS(of the free software foundation). > Hopefully, someday we will ALL own the type of machines RMS is writing for. > >Which brings me to my next comp.arch question. >*WHAT ARE THOSE MACHINES GOING TO BE??*. >Barring RISC vs. CISC debates. What is *THE* next great advance in >computer technology. Better hardware? Not 64 bits.(?) I meant more >in design. What do you guys at MIPS, LISP Machines, Motorola, Intel, >NEC, Sun (yes they are now in the CPU game), etc.... see as the most >promising area of exploration? > >Hopefully this is a more appropriate line of discussion..... 8^) > >Jeff >-- >Jeff Gortatowsky {seismo,allegra}!rochester!kodak!elmgate!jdg >Eastman Kodak Company >These comments are mine alone and not Eastman Kodak's. How's that for a >simple and complete disclaimer? ----- Hi Jeff ! Just to throw my own $.02 in, I will offer the following flight of fancy / dream system on which to run the next generation of OpSys ... Optical Computing. There are several problems with the direction that computer hardware is taking currently. Mind you, I would not be surprised to see somebody solve any or all of these difficulties by some innovation ! First, as the size of components decreases, heat becomes more and more of a problem. Also the smaller components are generally more vunerable to external events, i.e, shocks or stray radiaton. Third, the smaller the part, the tighter the controls during manufacture, the lower the yields, and therefore the more expensive the component. Lastly, we are still being strangled by the VonNeumann bottleneck. Yes, spiffy new architectures are stating to show us the way around this problem, but wouldn't a CPU with true multitasking be nice ? Well ... The answer to all of these problems, and many more that I probably havent proposed yet, lies in optical computing. Heat : less of a problem because with light moving through conductors or channels, there is no resistive heating, as there is in semi-conductors. Hardness : optical components themselves should be very radiation hard naturally, and should also be less vunerable to power problems. Physical impacts - well that could be a problem (shatter !). Manufacturing : Initially, of course, it will be expensive. Everything is expensive when it first comes out. But because of the inherant speed adavntage of optical computing, it will probably be unecessary to make the components as tiny as they are now. Also optical computing may have a truly astounding advantage in terms of manufacture: minimal hardware . (see discussion below) Multi processing : Optical computing would permit true multi-tasking by the simple trick of running different processes at different frequencies. AT&T is already using this trick in it's fiber optic communications reasearch, and has achieved (I beleive) 20 GIGABAUD over a single fiber. (Anyone out there at bell labs care to comment?) And now the ultimate advantage of optical computing : Holographic Hardware. You see, it may be possible to perform computing in a CPU that is a holgram, that would be capable of instantly reconfguring itself to the task(s) at hand. This would take the line between software and hardware and sort of wipe it out of existance. Some research has been done on holographic optical elements. I have seen the write ups in Scientic American. The possibilities are mind boggling ... Anyway, that's my random rambling on the topic. I am not an expert in any of these things, but I am interested enough to try and learn. If anybody out there knows more, please post ! Thanks. "Boy the system sure is running slow ! That kernel make took 2 milliseconds !" ----- Randy Martens - the opinions, facts, innuendos, lunatic ravings, and factoids expressed herein are strictly the products of my own demented mind, and certainly not those of anyone else demented mind. Fnord. "Reality - What a Concept !" - R.Williams
peter@sugar.UUCP (09/11/87)
Wow. First polled multitasking wasn't "True Multitasking". Now single CPU multitasking isn't "True Multitasking". What's next to fall... loosely coupled systems? -- -- Peter da Silva `-_-' ...!seismo!soma!uhnix1!sugar!peter -- 'U` <-- Public domain wolf.