Trushal Chokshi, Senior MEMS Systems Development Engineer at SiTime, discusses how MEMS technology affects Moore's law in this Embedded Computing Design article.
Moore’s law, the prediction that the number of transistors packed into an integrated circuit (IC) would double every two years while computing costs are halved, has had a remarkably long run. When Gordon Moore, co-founder and chairman emeritus of Intel, made this bold observation in 1965, few could have imagined that state-of-the-art microchips would one day contain a billion times more circuits. Thanks to ever-shrinking circuits, computing devices have steadily grown smaller, faster and cheaper, year after year, for more than five decades.
But Moore’s law has run up against the laws of physics. Leading foundries and chipmakers are reaching the physical limits of process technology, at least as we know it today. Migration from the 7-nanometer (nm) node to 5 nm and smaller has become exponentially difficult and costly. Gains in IC performance, density and cost reduction are now incremental. As Moore’s law continues to evolve, meaningful leaps in compute power and efficiency will be achieved through advances in multicore architectures, software, AI and machine learning, interconnects, packaging and materials science.