“There’s a law about Moore’s law,” says Peter Lee, a vice-president at Microsoft Research: “The number of people predicting the death of Moore’s law doubles every two years.”
Is he right? What's your take on Moore's law and the future of IC development?
>>61930563
Moore's law is dead because it's more like 3 years now.
I don't think it's much of a concern, at least not until we physically can't go any further. Then we're fucked.
>>61930603
Do you like this image?
I think this image is very nice
>>61930563
It's already slowing down. So we're mass producing 10nms now. We'll probably get to 8 or 7 in 5-10 years. Maybe even 5 in a decade, though it will probably take longer. But you can't make these thing a infinitely small. Say we get even better than we have at mass producing transistors - it will quite literally never decrease below the 1-3 mm range.
>>61930678
>moore changing his prediction
>has made it sooner now that intel has issues with 10nm
fucking lel
also
>approx 2600
Amazing and kinda terrifying at the same time
>>61930916
See, this is one of the problems with moores law, most people think it is about transistors getting smaller
In reality it was always for x dollarydoos you can get y transistors, and y will double every 2 years
>>61930987
Now that I'm researching Moore's law and some trends, what I've noticed is how people predict the death to come and it never does.
Moore's law ultimately isn't that important and processor designers can go other ways, and they have been. Notice all the new processors coming out designed for mobile devices and the IoT? It's because that's really the only place for manufacturers to economically expand to.
Regardless, there will be breakthroughs that keep Moore's law alive. Something as 'simple' as stacking chips/memory vertically would accomplish the goal of more transistors in less space for less money.
>>61931007
>The complexity for minimum component costs has increased at a rate of roughly a factor of two per year
In his original paper in 1965 where he's looking at recent trends. I don't know if I'm stupid but this quote only made sense to me in the greater context of a few paragraphs.
He's basically stating that:
The complexity (intricate mass of components) that one receives for minimum component costs (the initial cost for the initial amount of components) doubles each year.
It was only later that Moore made the statement that transistors double every 2 years which is now so popular.