As Moore’s law turns 50, what does the future hold for the transistor?

"The future of integrated electronics is the future of electronics itself."

When Intel co-founder Gordon Moore began his now-famous 1965 paper (PDF) "Cramming More Components Onto Integrated Circuits" with his bold proclamation about the future of electronics, few would have believed it—especially given the cost of integrated circuits at the time. And yet, 50 years on, Moore's three-page paper has come to define the computing industry. Its most famous prediction, that the number of components on an integrated circuit would double every year (later revised down to two years a decade later) has become a self-fulfilling prophecy for the computing industry, a solid goal for the world's semiconductor manufacturers to reach for.

For the most part, it's a goal that's been ably reached. The rapid pace of technological advancement caused by Moore's law has enabled smartphones and tablets to usurp the desktop PC as the consumer's platform of choice, the likes of the PlayStation Vita to put the graphical horsepower of a PlayStation 2 (and sometimes 3) in the palm of your hand, and for AI like IBM's Watson to wipe the floor with some Jeopardy veterans. When Intel released its 8088 CPU back in 1979, the same CPU used in the original IBM PC, it came packed with 29,000 transistors built on a three-micrometer process to reach its 4.77 MHz clock speed. Today, a modern four-core Haswell processor packs in around 1.4 billion transistors built on a 22-nanometer process to reach 3GHz clock speeds.

Read 4 remaining paragraphs | Comments