• Question: Does "Moore's Law" exist and if it does how does it affect your work? For Example, do you think there will be a peak in technological advantages, or will things always keep improving requiring you to keep reinventing?

    Asked by Jake M/ Corbyn to Alessandra, Dimitrios, Niamh, Becky, Stacey, Tony on 20 Jun 2017.
    • Photo: Stacey Marple

      Stacey Marple answered on 20 Jun 2017:

      Hi, Moore’s Law is named after Intel cofounder Gordon Moore. He said that the number of transistors in an intergrated circuit doubles approx every 2 years, it has existed but most computing companies agree the rate is slowing. It does not directly affect my current work but when I designed robots it was a consideration, you don’t want to deliver outdated products. Shrinking chips no longer makes them faster or more efficient, so as the rate plateaus, industry must look to make better computers in different ways. I don’t think the technologically revolution will stop because Moore’s law is finished, we will just find a different way of improving.

    • Photo: Niamh Ryall

      Niamh Ryall answered on 21 Jun 2017:

      Good question. Moore’s law is usually described as technology getting twice as fast or half as cheap every 2 years. This is because the amount of transistors (essentially on/off switches) you can fit on a chip doubles in roughly 2 years. Until now, this has been approximately true, the problem now is that we are way past what’s called the diffraction limit. Imagine shining light through a hole to create a pattern, we can only do that to about 1/2 the wavelength of the light we’re using. For green light of 500 nm, you can template like this to about 250 nm. There’s clever things we can do to break this limit, like bringing things into near contact and using X-rays (shorter wavelengths) but it’s getting more and more technically difficult. The next step might be to “grow” the 3D structure using what’s called “bottom up” techniques. Or maybe channeling gases to erode 3D structures into a chip. These kinds of sizes also means quantum mechanics becomes important! They also need to think about cooling. If you’d like to see the sizes we’re talking about, see Intels videos like “22 nm explained”.

Comments