Moore's Law is on the verge of collapsing

Advertisement

Moore's Law may be about to breathe its last.

Advertisement

It's a world-famous maxim that has predicted the development of computers for decades - but, an industry roadmap suggests, it will soon no longer be viable.

Moore's Law, formulated by Intel cofounder Gordon Moore in 1965, is simple. It says that as technology improves, the number of transistors that can be fitted on a given integrated circuit will double every two or so years - and accordingly, computing power doubles too.

Complimentary Tech Event
Transform talent with learning that works
Capability development is critical for businesses who want to push the envelope of innovation.Discover how business leaders are strategizing around building talent capabilities and empowering employee transformation.Know More

For the last 50 years, it has held true, as computers have become exponentially more powerful, shrunk ever-smaller and become cheaper and cheaper.

But the 2015 International Technology Roadmap, published earlier this month by the Semiconductor Industry Association, thinks this trend will soon come to an end. (We first saw the report over on Engadget.) The size of atoms themselves mean we're pushing how far Moore's Law can go: Beyond a certain point, physical constraints - and economic costs - means conventional transistors can't be made any smaller.

Advertisement

The Semiconductor Industry Association thinks that beyond 2021, it won't be financially viable for companies to make transistors any smaller. Instead, chip builders searching for a speed boost will have to turn to alternatives like layering circuitry in 3D design. But this presents its own challenges, like adequately dissipating the heat that builds up as the circuits function.

If Moore's Law does come screeching to a halt in the next five years - and alternative new designs don't find a way to overcome this new obstacle - this doesn't mean computing devices will just stop getting better.

"Think about what happened to airplanes," computer scientist Daniel Reed told Nature earlier this year. "A Boeing 787 doesn't go any faster than a 707 did in the 1950s - but they are very different airplanes ... Innovation will absolutely continue - but it will be more nuanced and complicated."

NOW WATCH: Britain could be chemically 'growing' military drones in the future