Moore’s law

Moore’s law is the observation that the number of transistors on an integrated circuit doubles roughly every two years, with a corresponding increase in computing performance at roughly constant cost. It is named for Gordon Moore, co-founder of Intel, who first described the trend in a 1965 paper and revised the doubling period in a 1975 update.

Moore’s law is not a law of physics but an empirical observation of an industry trend, sustained for decades by continuous advances in semiconductor fabrication that shrank transistor size generation after generation. It became a self-fulfilling target as much as a prediction: chip manufacturers planned their research and manufacturing roadmaps around sustaining the doubling rate, which helped keep the trend alive long after it might otherwise have stalled.

The law has driven much of the economics of computing for half a century. It underwrote decades of software’s ability to grow more resource-hungry without users perceiving it as slower, since hardware gains largely kept pace. Wirth’s law is a direct counterpoint, observing that software has tended to expand to consume those hardware gains and then some, so that felt performance has not improved as fast as raw transistor counts would suggest.

Since the 2010s, the physical limits of transistor miniaturization have slowed the classical doubling rate, and much of the industry’s continued performance growth now comes from other sources: specialized hardware such as GPUs, parallelism, and architectural changes, rather than from transistor density alone.

See also

• Laws and principles

References

• Moore, Gordon E. (1965). Cramming More Components onto Integrated Circuits. Electronics, 38(8).