In 1965, Gordon Moore, co-found of Intel, predicted in a paper published in Electronics Magazine, that computing power would roughly double every two years. The term 'Moore's law' was later coined and it has stuck ever since. But does the unlimited potential to create faster machines have a pre-assigned limit?
Indeed, for the last 44 years or so - we've watched computers gain from huge advancements in technology, as new manufacturing techniques and a doubling of the transistors available on integrated circuits pushed the boundaries of computing and lowered costs.
But not everybody shares Moore's resilient optimism. Two physicists from Boston University believe the long held law is due to hit the brakes. And according to their studies, there isn't much we can do about it all.
InsideScience reports that physicists Lev Levitin and Tommaso Toffoli are so sure of their claims, that they're willing to bet on the way we measure light itself.
The shrinkage of components - one of the main conceits of Moore's original law, will eventually need to get so small that it will eventually reach a 'quantum limit' - the general inability to get any smaller without breaking the laws of nature.
 |
| Moore's law in practice. Source: Wikicommons |
Levtin and Toffoli give us 75 to 80 years before Moore's Law needs to be rethought or swapped out for a totally new way of thinking. But even that claim is cautiously hopeful, particularly if you believe Scott Aaronson, an assistant professor of electrical engineering from MIT, who estimates the law will likely run out in less than 20 years - a tad earlier to start worrying.
The men speculate that the limit of resources and how we use these resources to continue the rate of 'doubling', will eventually reach their conclusion.
They base much of their hypothesis around the development of quantum computing, a belief that rests strongly in the notion of highly theoretical processes that could one day revolutionise the speeds of our computers and change the way we calculate data, forgoing silicon and 0s and 1s for qubits.
But until we have more concrete evidence regarding these quantum principles and how they are likely to develop in tandem to contemporary processes, Moore's law is looking pretty safe - for now.