Computer Technology Forecast


Longevity Index

Computer Clock Speeds  
    For tonight, how about a little light froth as a break from the heavy stuff? How about a computer technology update?
    The computer clock speeds shown in the "Speed" column can probably be expected to be achieved about one year after the introduction date given in the left-most column 
    Moore's Law has slowed down from doubling every 18 months to doubling every two years, but who's complaining?
    Through 2009-2010, Intel has promised the kinds of performances shown in the chart below. The Semiconductor Industry Association roadmap calls for the numbers (in 2016) shown in the 2011-2012 row. (I'll bet on seeing them in 2011-2012.) IBM has delineated the future of disk capacities, projecting the kinds of improvements shown through 2011-2012.

    Beyond 2011-2012, it's strictly "If this goes on... " blue-sky numbers. Sometime down the road, we may shift to carbon nanotubes, or to some comparable new technology. But it's interesting to note that improvements could continue at this "doubling every two years" rate for another 30 years. Wow!  

Computer Capacities Versus Time



Clock Speed, GHz

Disk, Terabytes


2003 90     6 0.2 1
2005   65     12 0.4 2
2007   45     25 0.8 4
2009   32     50 1.6 8
2011 22     100 3.2 16
2013 16     200 6.5 32
2015   11     400 13     64
2017 8     800 25     128
2019 5.8 1,600 50     256
2021 4     3,200 100     512
2023 2.8 6,400 200     1,024
2025 2     13,000 400     2,048
2027 1.4 25,000 800     4,096
2029 1     50,000 1,500     8,192
2031 0.7 100,000 3,000     16,384

What Might We Do With These Whiz-Bang Computers?
    What would this do for us? To contemplate that, we might hearken back 30 years to 1972. Solid-state memory was in its infancy. Texas Instruments was selling a 16-bit minicomputer with 8 kilobytes of semiconductor RAM for $4,000 (equivalent to about $15,000 today). Sharp had had its ELSI (Extra-Large-Scale Integration) pocket calculator (for a large pocket) on the market for a year, at $400 ($1,500 2002 $). Hewlett Packard would introduce its HP-35 scientific calculator that could (wonder of wonders!) eliminate trig and log tables for $395 ($1,500 today) in November, 1972.  Where would it all end? Would we actually see home computers some day?
    Extrapolating to 2032, one might foresee computers in everything, and possibly even throwaway computers stamped out of silicon, or produced from organic materials. (Basically, we have "computers" in everything already, in the form of logic circuitry for clocks, microwaves, TV controls, and so forth.)
    I certainly hope that before then, we see large, cheap, 3-D wall displays with virtual-reality-class "You-are there" telepresence. I also expect to see household labor-saving devices, including visually-navigating lawn trimmers and floor cleaning equipment. And I'll be surprised if there aren't much better anthropomorphic robots, at least on display.
    Below is a forecast that appeared in News of the Ultranet in the Spring, 2000, edition of Ubiquity Magazine:

From the Spring, 2001, Issue of Ubiquity Magazine

    You've just gotten home from work. One of the walls in your family room is (or all of them are) dedicated to display. You say, "Athena, please see if you can reach my sister, Barbara." "Athena", your computer, says, in a slightly accented woman's voice, "Calling your sister, Barbara", and a few seconds later, the wall lights up with a full-size, 3-D image of Barbara. She says, "Hello", and you say, "Hi, sis. How did your day go?" You talk for a while, and hang up. Then you say, "Athena, what happened in the stock market today?" Charts of the DJIA, the S&P 500, and the NASDAQ composite (or whatever you've previously chosen) appear on the wall, together with a menu of news reports, your personal stock portfolio, and other options. After dealing with financial investment news, you might ask to see the regular news, both global and local. This would consist of formal broadcasts, with a freeze function, and with side menus to allow you to investigate specific topics in greater depth. Rather than being canned, the news is interactive. It might consist of a combination of human-delivered video news reports and a newspaper printed on e-paper or your notepad computer. You could also opt to have the printed matter read to you by an avatar-based cybercaster, like Ananova. You might already have instructed the household robot to prepare supper for you, setting the table, and serving the food. After supper, the household robot might clear the table and wash the dishes. The household robot may not be built in humanoid form, but might have grippers of some type to efficiently handle the food and the dishes. It might not be very strong--just strong enough to handle the food and the dishes. When not required, it might repair to a broom closet to recharge its batteries.
   How soon is this apt to happen?
    The household robots may not be here for 20 or 30 years, but part of this fanciful scenario could probably happen right now if you wanted it. Voice-controlled computers are here now. Large-screen displays are here now. The Ananova site already offers interactive news, albeit on a small screen. However, to make this really effective, you need a wall-sized display, and you need bandwidths of 20 megabits/second or higher. Three-D images without glasses can be achieved using various techniques that have been available for decades, but so far, such displays haven't made been widely adopted. Holographic displays require teraflops computing speeds, as well as special displays. Dimension Technologies has been working for more than a decade on systems that work like those on the covers of some children's books that use a vertically striped lenticular screen that shows a different image to each eye. (My 3-D Star Trek mousepad works this way, although the Romulan Warbirds in the background are a little fuzzy.) Dimension Technologies now offers an XGA 15" LCD display for $1,700, and an SXGA 18" display for $4,995. The other current chicken-and-egg problem is that of finding 3-D content, and for webcasts, sufficient bandwidth to allow the real-time transmission of the signals to drive XGA and SXGA displays. This we probably won't see until second-generation wideband shows up later in this decade, although such data rates are achievable now on various private networks such as Internet 2. For 3-D without glasses, we may be looking at anywhere from 10 to 30 years. But eventually, I think it's bound to come. What we're experiencing right now may be akin to the early days of radio, when listeners wore earphones and twiddled with their "cat's whiskers" to hear scratchy music broadcast by station KDKA.
     In the meantime, Dynamic Digital Depth (DDD) is offering red-and-green- glasses-based, 3-D webcasts now. (They'll send you a free pair of glasses!) They also offer better, Polaroid-based 3-D. They're obviously trying to popularize 3-D over the web. Who needs 3-D? Who needs color? Glasses-based 3-D might rest upon the availability of handy, low-cost, webcam pairs that can be set up to generate 3-D files that anyone can create and broadcast.
    Dimension Technologies and DDD have recently teamed to bring 3-D to the public.
    It's all very well to fantasize about what we might have in 20 to 30 years, but how about 2005? That's no farther away than 1997.
    I could imagine that 21", 22", and/or 23" CRT displays with resolutions of 2,000 X 1,500 or 2,300 X 1,500 will be common, and will cost a few hundred dollars. Cheap, high-resolution, large screen displays could also be here by that time, although I thought that ten years ago. Third-generation wireless will probably be here by then, with early-adopters web-surfing and using videophony on their cellphone/PDA's. Wideband will probably be widespread by 2005, and it's conceivable that second-generation wideband, at 10 to 20 megabits per second, will be creeping over the hill. Some of us may begin voice-commanding our computers if that's advantageous. (Of course, that capability has been available for the past ten years, and most of us aren't using it.) Combination video/still cameras will probably be common, perhaps in PDA's. I could possibly afford Dimension Technology's 15" LCD display by then. It may be cheaper in 2005. Their 18" display might even be within reach.
     Okay. What about 2010? By this time, I would hope that we would have the wall-sized high-resolution displays, or perhaps, laser holographic 3-D displays. I would expect that third-generation, 100-to-200 megabaud wideband connections are appearing in selected locations. It's conceivable that 1 gigabaud connections might be available in schools and libraries.