The first computer was as big as a room. Now they're the size of a full stop . . . and getting even smaller
New chip is the equivalent of 20 of today's PCs, writes Charles Arthur
Support truly
independent journalism
Our mission is to deliver unbiased, fact-based reporting that holds power to account and exposes the truth.
Whether $5 or $50, every contribution counts.
Support us to deliver journalism without an agenda.
![Louise Thomas](https://static.independent.co.uk/static-assets/support-us/louise-thomas.png)
Louise Thomas
Editor
Videophones for the price of a standard telephone, wristphones, and computers that recognise faces are among the likely products of a leap forward in microchip technology announced yesterday.
The advance unveiled by the US semiconductor giant Texas Instruments (TI) would cram the equivalent of 20 of today's PCs onto a single chip, which could be on sale within three years. The result would be affordable products with enormous amounts of computing power.
The announcement is the continuation of a 30-year race to pack more and more electronics into ever-tinier spaces. For example, the computing elements of the wartime computer, the Colossus, right, would today take up the same space as the full stop at the end of this sentence.
"This new technology will make it practical to use calculations which today are just theoretical, because they would take too long with existing machines," said John Carter, of the University of Southampton's Electronics and Computer Science department. "Systems that need to identify faces have to do it in seconds, not minutes or hours."
Similarly, sending video signals down phone lines requires immense computing power at each end of the link. This has kept prices of videophones around pounds 500 for some time. More powerful computer chips would make videophones both efficient and affordable.
TI will not have a monopoly, though: Japanese and American companies are hot on its trail. Hitachi, NEC, IBM and Intel - which makes the best- selling Pentium chip for PCs - are all within sight of the same target.
By boosting the power of a single chip while keeping it at the same price, the new technology will cut the cost of products, since the most expensive element of making modern computers is assembling the different components.
TI's "TImeline" design reduces the width of the silicon "wires" in the chip from today's common value of around 0.6 microns (millionths of a metre) - 200 times thinner than a piece of paper - to 0.18 microns. This dramatic reduction in size means that many more transistors, the building blocks of a computer, can be squeezed into the same area.
Using this process, TI intends to build a chip containing 125 million transistors - a huge step up from the Pentium Pro, commonly found in PCs today, which contains 5.5 million transistors. Intel presently makes chips using 0.35 micron designs. The more transistors a single chip contains, the more processes it can carry out at once.
The latest announcement by TI is the demonstration of a "law" first proposed more than 30 years ago by an American engineer, Gordon Moore.
In 1965, Moore was preparing a speech about the improvement in performance of computers' memory chips, which were then rudimentary. He noticed that each successive generation of chips contained twice as many transistors as its predecessor, and was launched 18 months or so later. He reasoned that this trend could continue almost as long as required.
Moore, now 69, went on to be co-founder of Intel, now the biggest manufacturer of semiconductor chips in the world. The company's first chip, the 4004 (used in early calculators) contained 2,300 transistors. Since then, the power of individual chips has increased almost a millionfold.
However, the inexorable progress towards smaller chips may be nearing its limit. Physicists reckon that below 0.1 micron - a size which would be reached around 2005 - significant problems could arise from the unusual phenomena known as "quantum effects", in which individual electrons can tunnel through solid barriers.
This would mean that calculations could become unreliable.
Clive Hoggar, a semiconductor product manager for TI, said: "The technology has a habit of moving immovable barriers.
''Somehow, we always get past these problems."
Subscribe to Independent Premium to bookmark this article
Want to bookmark your favourite articles and stories to read or reference later? Start your Independent Premium subscription today.
Join our commenting forum
Join thought-provoking conversations, follow other Independent readers and see their replies
Comments