下一个科技热点在哪里？

    据说，在1964年，当时还在飞兆半导体公司（Fairchild Semiconductor）工作的戈登•摩尔博士正在为《电子学》（Electronics）杂志准备一篇有关半导体存储芯片演化史的论文。他决定把半导体存储芯片的容量和它们的生产年份用坐标图标释出来。当时这张图上只有五六个数据点，因为半导体芯片刚刚诞生五年，每张芯片上只有几百个晶体管。

    将所有的数据点连起来之后，摩尔发现，这条曲线一开始的走向比较平缓，然后迅速大幅上扬。可惜那条曲线最后跑出了草纸的顶部。于是摩尔又画了一张对数坐标图，这次他惊奇地发现，将一条坐标用10的N次方来计数之后，存储芯片的发展趋势呈一条几乎水平的直线。作为硅谷历史上最聪明的人之一【也是未来英特尔公司（Intel）的共同创始人】，摩尔不仅明白了这张图说明了什么，更重要的是明白了它意味着什么。

    首先，这张座标图说明，半导体存储芯片的发展速度，是人类在任何产品上都不曾见过的——如果这种速度能够继续保持下去，很快就会发生重大的代际跨越。后来这根趋势图被总结为“摩尔定律”，即半导体存储芯片的性能，每一两年就会翻一番。

    但是摩尔定律背后的重大意义在于，我们首次有了一张通向未来的路线图。你可以根据摩尔定律预测几十年后的情形，而且直到现在，摩尔定律与现实辉映得都一直十分准确。这意味着你可以对未来的某个日子进行规划和设计。可以说，它是让你在未来获得成功优势的一把神奇钥匙。

    摩尔定律很快从存储芯片扩展到了逻辑芯片，然后又渗透到了半导体行业的其它领域，而且很快使芯片行业成为增长最快的一个行业，同时也是最具价值的一个行业。

    但是当时，包括摩尔本人在内，所有人都没有预测到，到了八十年代和九十年代之交的时候，全球的芯片数量数以百亿计，摩尔定律已经具有了冲出电子行业，进入国民经济其它领域的能力。从汽车到基础建设，从基因研究到电话业务——各大企业、实验室和政府机构发现，只要他们有能力与摩尔定律挂钩，就能实现指数级增长。其中的一个结果就是我们这个时代的重大革新技术——互联网。

    在撰写关于英特尔公司历史的新书《英特尔三杰》（The Intel Trinity）的过程中，我认识到我们犯了一个严重的错误，那就是过于相信摩尔划出的这条平滑曲线。这个错误始于戈登•摩尔更换了坐标系的那一瞬间。那是因为在对数坐标系上看起来甚为平缓的直线背后，还隐藏着一条大起大落的抛物线。而过去几十年的大部分时间里，我们都忘掉了这个事实。

    那么，那条陡峭的曲线究竟隐藏着什么？像所有抛物线一样，它一开始也显示出带有迷惑性的平缓趋势。在头40年里，摩尔定律还处于比较温和的水平。这条相对平缓的曲线在微型计算机、微处理器、数码计算器、电脑游戏、个人电脑、互联网、机器人、无线电话、智能手机和电子商务等领域的发展过程中都能看得到。换句话说，光是这条曲线最平缓的部分，就已经将我们居住的世界完全改造了。

    As the legend goes, in 1964 Dr. Gordon Moore, then at Fairchild Semiconductor, was preparing a paper for Electronics magazine on the evolution of semiconductor memory chips. He decided to plot the capacity of those chips, versus their year of introduction, on some graph paper. There were only a half-dozen or so data points, as memory chips at that point were less than five years old and only contained a few hundred transistors each.

    Connecting the dots, Moore noticed a familiar parabolic curve – shallow at the beginning and then quickly turning upwards. Unfortunately, that curve also quickly went straight off the top of the page. So Moore switched to logarithmic paper – that is, with one side in powers of ten — and, stunningly, the memory chips tracked along a straight, nearly horizontal line. Moore, one of the most brilliant individuals in Silicon Valley history (and future Intel INTC -2.74% co-founder), not only knew what this said, but more important, what it meant.

    What it said was that semiconductor memory was progressing at a pace never before seen in any product in human history – and if that pace could be maintained the generational leaps would soon be gigantic. This trajectory – at first defined as the doubling of the performance of semiconductor chips every couple years – became known as Moore’s Law.

    But what Moore’s Law meant was that for the first time, perhaps in any industry anywhere, there was now a map into the future. You could track that line decades out into the future – and know exactly what memory chips would be like on any date. And that meant you could plan for that date, and you could build for it. It was a magic key to competitive success.

    Moore’s Law quickly spread from memory to logic chips and then to the rest of the semiconductor industry – and quickly made the chip business the fastest growing industry. And soon, the most valuable.

    What no one, not even Moore himself, saw coming was that, by the 1980s and 1990s, with tens of billions of chips out in the world, Moore’s Law would break out of electronics and into the rest of the economy. From automotive to infrastructure to genetic research to telephony – companies, laboratories and government agencies discovered that if they could find any way to hook up to Moore’s Law they too could experience exponential growth. One result was the great transformational technology of our time, the Internet.

    In writing my new book on the history of Intel Corporation, The Intel Trinity, I became convinced that we have made a serious mistake being so comfortable with that shallow line. And that mistake begins with Gordon Moore’s change of graph paper. That’s because behind the gently sloping straight line there still lies that dizzying parabolic curve. It is this reality that has been largely forgotten over the last few decades.

    What lies in that steep arc? Like all parabolic curve, it begins deceptively flat: for the first 40 years, Moore’s Law is a gentle grade. Yet under that comparatively flat curve can be found the minicomputer, the microprocessor, the digital calculator, computer gaming, the personal computer, the Internet, robotics, wireless telephony, the smart phone and electronic commerce – in other words, our world has been utterly transformed by just the shallowest section of this curve.