May Moore's law rest in eternal peace,,....
Intel confirms tick-tock-shattering Kaby Lake processor as Moore’s Law falters | Ars Technica
May Moore's law rest in eternal peace,,....
Intel confirms tick-tock-shattering Kaby Lake processor as Moore’s Law falters | Ars Technica
Moore's law is not dead yet. At first, moore's law said the number of transistors would double every year. Now it has become every 18 months, and now further delayed. But there is no proof yet to the exponential growth has ended.
Besides, even if it did end, what harm would come from it? It's not like we're seeing any big improvements from all those transistors anymore. Processor improvements have been only minor over the last years. What good has moore's law done us during these years?
All for the better. Never much liked Moore's Law.
For one, it's entirely a bet on guesswork which isn't something I hold close to my heart when it comes to make safe predictions.
Second, there's never been one Moore's Law, but many. It's not even many laws, or one law. There's never been a Law. Because of being pure guesswork, every few years it needed to be reexamined and changed to conform to a new guesswork since the previous one turned out wrong after a while.
Third, it has been responsible for driving industry development instead of predicting industry development, which is sad. On many cases, throughout the years, semiconductor companies actually made their mid and longterm research and development plans based on it. No wonder the "law" was law...
Originally Posted by brewbuck:
Reimplementing a large system in another language to get a 25% performance boost is nonsense. It would be cheaper to just get a computer which is 25% faster.