WEBVTT - How Moore's Law Works

0:00:00.320 --> 0:00:02.960
<v Speaker 1>Brought to you by the reinvented two thousand twelve Camray.

0:00:03.240 --> 0:00:08.719
<v Speaker 1>It's ready. Are you get in touch with technology? With

0:00:08.920 --> 0:00:13.159
<v Speaker 1>tech Stuff from how stuff works dot com. Hi there,

0:00:13.200 --> 0:00:15.640
<v Speaker 1>and welcome to the podcast. My name is Chris Pollette.

0:00:15.640 --> 0:00:17.680
<v Speaker 1>I'm an editor here at how Stuff Works, and with

0:00:17.720 --> 0:00:21.160
<v Speaker 1>me is writer Jonathan Strickland. Hey there, Yeah, we were

0:00:21.360 --> 0:00:25.279
<v Speaker 1>coming up with ideas for the tech Stuff podcast, and um,

0:00:25.320 --> 0:00:27.319
<v Speaker 1>instead of doing an article, we we found one of

0:00:27.320 --> 0:00:30.800
<v Speaker 1>those concepts that runs across several of our articles. And

0:00:30.840 --> 0:00:34.199
<v Speaker 1>it's called Moore's law. That's right, and uh, don't let

0:00:34.200 --> 0:00:36.320
<v Speaker 1>the name fool you. It's it's not actually a law.

0:00:36.680 --> 0:00:38.239
<v Speaker 1>You're not gonna get in trouble. If you break it,

0:00:38.520 --> 0:00:41.960
<v Speaker 1>you'll probably get a round of applause. Moore's laws is

0:00:42.000 --> 0:00:45.519
<v Speaker 1>more of a concept that's kind of been tweaked a

0:00:45.520 --> 0:00:48.800
<v Speaker 1>bit over the years. It's not exactly the same thing

0:00:48.840 --> 0:00:52.000
<v Speaker 1>as it was when Gordon Moore proposed it back in

0:00:52.120 --> 0:00:56.600
<v Speaker 1>nineteen sixty five, but in general the concept remains pretty

0:00:56.680 --> 0:00:59.280
<v Speaker 1>much the same. Yeah. As a matter of fact, Gordon Moore,

0:00:59.320 --> 0:01:03.360
<v Speaker 1>who was the founder of intel Um, actually didn't even

0:01:03.360 --> 0:01:05.959
<v Speaker 1>want to call it More's law. His friend from cal

0:01:06.040 --> 0:01:09.200
<v Speaker 1>Tech a professor cal Tech, decided to call it More's law.

0:01:09.280 --> 0:01:12.720
<v Speaker 1>It was a concept that doctor Moore Um proposed than

0:01:12.720 --> 0:01:16.080
<v Speaker 1>a paper with the exciting title Cramming More Components on

0:01:16.120 --> 0:01:19.800
<v Speaker 1>the Integrated Circuits that was published in Electronics magazine in

0:01:19.840 --> 0:01:24.040
<v Speaker 1>ninet And basically what War's law is is, you know,

0:01:24.080 --> 0:01:28.560
<v Speaker 1>he said, as you make uh semiconductors, the more transistors

0:01:28.600 --> 0:01:29.960
<v Speaker 1>that you put on a chip, you will be able

0:01:29.959 --> 0:01:32.320
<v Speaker 1>to duck to double that within the span of a year.

0:01:33.360 --> 0:01:36.040
<v Speaker 1>And uh, you know, he actually had to revise it.

0:01:36.040 --> 0:01:38.000
<v Speaker 1>It's not a law. He had to revise it because

0:01:38.040 --> 0:01:41.360
<v Speaker 1>of the way they manufacture chips. That actually went up

0:01:41.360 --> 0:01:44.520
<v Speaker 1>to uh one and a half years and then two years,

0:01:44.600 --> 0:01:47.160
<v Speaker 1>and you know, now it's starting to slip a little

0:01:47.160 --> 0:01:50.560
<v Speaker 1>bit as the laws of physics are catching up with manufacturing, right.

0:01:50.600 --> 0:01:53.120
<v Speaker 1>And and the definition was tweaked also by saying, okay,

0:01:53.160 --> 0:01:56.080
<v Speaker 1>instead of just saying the number of transistors, maybe maybe

0:01:56.120 --> 0:01:59.200
<v Speaker 1>we should word that as in a computing power, like

0:01:59.240 --> 0:02:01.720
<v Speaker 1>every two years, the computing power on a on a

0:02:01.840 --> 0:02:06.800
<v Speaker 1>chip doubles um. And this has pretty much been holding

0:02:06.800 --> 0:02:10.919
<v Speaker 1>truth throughout the history of of electronics and computing, uh,

0:02:11.120 --> 0:02:14.520
<v Speaker 1>mainly because not not from a technological standpoint, although that

0:02:14.520 --> 0:02:18.680
<v Speaker 1>plays a huge role as people improve manufacturing techniques and

0:02:18.720 --> 0:02:23.080
<v Speaker 1>things like that, But really from a commercial and social standpoint,

0:02:23.080 --> 0:02:27.120
<v Speaker 1>we're talking about companies that they have to meet expectations

0:02:27.160 --> 0:02:30.359
<v Speaker 1>now because everyone knows about Moore's Law, and everyone knows

0:02:30.360 --> 0:02:32.400
<v Speaker 1>that a computer has to be twice as powerful as

0:02:32.400 --> 0:02:35.320
<v Speaker 1>it was two years ago, and now companies feel like

0:02:35.320 --> 0:02:38.399
<v Speaker 1>that that's actually a standard they have to meet. It's

0:02:38.440 --> 0:02:43.680
<v Speaker 1>no longer dictated by technology, it's dictated by expectation. Yeah,

0:02:43.680 --> 0:02:46.320
<v Speaker 1>that's that's true. Um. If you go into Intel's website,

0:02:46.320 --> 0:02:48.440
<v Speaker 1>they actually have a section of the site dedicated to

0:02:48.440 --> 0:02:50.080
<v Speaker 1>Moore's Law, and if you read through it, you know,

0:02:50.160 --> 0:02:52.360
<v Speaker 1>you don't even really have to read very deeply into it.

0:02:52.639 --> 0:02:56.919
<v Speaker 1>They talk about how they are attempting to keep that standard.

0:02:57.480 --> 0:03:00.200
<v Speaker 1>So it's it's kind of funny that, you know, it's

0:03:00.480 --> 0:03:05.120
<v Speaker 1>it's this nebulous artificial Yeah. Yeah, but they're they're trying

0:03:05.160 --> 0:03:07.959
<v Speaker 1>to make it happen. One of the funny things about

0:03:08.560 --> 0:03:12.040
<v Speaker 1>Dr Moore's prophecy, though, in in nineteen sixty five, is

0:03:12.440 --> 0:03:15.119
<v Speaker 1>um he said that in his paper he says that

0:03:15.360 --> 0:03:18.560
<v Speaker 1>if they are able to meet these standards, that we

0:03:18.560 --> 0:03:22.880
<v Speaker 1>would be able to have certain conveniences, you know, can

0:03:23.080 --> 0:03:25.480
<v Speaker 1>that are able to uh to make our lives easier,

0:03:25.520 --> 0:03:28.520
<v Speaker 1>such as you know, computers in the home or mobile

0:03:28.639 --> 0:03:36.800
<v Speaker 1>communications devices. So he predicted cell phones. Yeah. Um. The

0:03:36.800 --> 0:03:39.840
<v Speaker 1>the interesting thing here is that Moore's law can't go

0:03:39.920 --> 0:03:43.360
<v Speaker 1>on forever, and that's because of the laws of physics.

0:03:43.440 --> 0:03:47.200
<v Speaker 1>The laws of physics state, specifically quantum physics that when

0:03:47.200 --> 0:03:50.680
<v Speaker 1>you get down to a certain level, uh, things, the

0:03:50.760 --> 0:03:52.560
<v Speaker 1>things that a circuit needs to be able to do,

0:03:52.960 --> 0:03:55.440
<v Speaker 1>you just can't do at a certain size. And we're

0:03:55.480 --> 0:03:58.160
<v Speaker 1>talking about just a few nanometers in size, and a

0:03:58.240 --> 0:04:01.000
<v Speaker 1>nanometer is a billion of the eater, so it's very

0:04:01.120 --> 0:04:06.920
<v Speaker 1>very tiny. Well, at five nanometers, a logic gate can

0:04:06.960 --> 0:04:10.120
<v Speaker 1>no longer perform, electron gate can no longer perform what's

0:04:10.200 --> 0:04:11.920
<v Speaker 1>meant to do? Now? These gates, what they do is

0:04:11.920 --> 0:04:15.240
<v Speaker 1>they allow electrons to pass through or not pass through.

0:04:15.720 --> 0:04:18.279
<v Speaker 1>That ends up translating into a one or a zero

0:04:19.120 --> 0:04:23.080
<v Speaker 1>basic bits here. But an electron gate that's five nanometers

0:04:23.080 --> 0:04:25.520
<v Speaker 1>are smaller in size, can't do that because an electron

0:04:25.600 --> 0:04:28.839
<v Speaker 1>can tunnel through it. It doesn't actually make a hole,

0:04:28.960 --> 0:04:31.719
<v Speaker 1>it just passes through. It's as if it's on one

0:04:31.720 --> 0:04:33.640
<v Speaker 1>side at one moment and on the other side of

0:04:33.720 --> 0:04:36.800
<v Speaker 1>the other, and and somehow it just passes through. It's

0:04:36.800 --> 0:04:39.320
<v Speaker 1>one of those weird things about quantum mechanics. So you

0:04:39.400 --> 0:04:43.480
<v Speaker 1>can't make anything that size because you can't control the electrons.

0:04:43.520 --> 0:04:46.920
<v Speaker 1>So that that's that's really the limit. Even if we

0:04:47.040 --> 0:04:49.640
<v Speaker 1>figure out a way of making things that's small, it

0:04:49.680 --> 0:04:53.440
<v Speaker 1>won't work. I also read that of course, as you

0:04:53.520 --> 0:04:56.400
<v Speaker 1>increase the number of transitions, you're also increasing the amount

0:04:56.400 --> 0:04:59.400
<v Speaker 1>of electricity that it needs, and that also affects the

0:04:59.400 --> 0:05:01.440
<v Speaker 1>amount of cool ing that you're going to have to

0:05:01.480 --> 0:05:04.839
<v Speaker 1>have because that everything up so um. It's just a

0:05:04.880 --> 0:05:07.400
<v Speaker 1>matter of time, I guess, before we reach the end

0:05:07.440 --> 0:05:09.080
<v Speaker 1>of Moore's law, and then we'll have to see what

0:05:09.480 --> 0:05:13.560
<v Speaker 1>manufacturers do to to, you know, just pump up computers

0:05:13.560 --> 0:05:17.440
<v Speaker 1>and electronics just that much more, right, whatever tricks they

0:05:17.440 --> 0:05:19.839
<v Speaker 1>pull out of their bags. Well, if you'd like to

0:05:19.880 --> 0:05:23.040
<v Speaker 1>read more about articles that in which we mentioned More's law,

0:05:23.040 --> 0:05:26.920
<v Speaker 1>and you can read Jonathan's how quantum computing will work

0:05:27.600 --> 0:05:30.800
<v Speaker 1>and how PCs work, And there are several others on

0:05:30.880 --> 0:05:32.880
<v Speaker 1>how stuff works dot com that are up today that

0:05:32.960 --> 0:05:35.360
<v Speaker 1>you can check out. So for now, thank you. Very

0:05:35.440 --> 0:05:39.279
<v Speaker 1>much for listening. For more on this and thousands of

0:05:39.279 --> 0:05:42.719
<v Speaker 1>other topics, does it how stuff works dot Com Let

0:05:42.839 --> 0:05:45.440
<v Speaker 1>us know what you think. Send an email to podcast

0:05:45.720 --> 0:05:51.480
<v Speaker 1>and how Stuff Works dot Com brought to you by

0:05:51.520 --> 0:05:55.000
<v Speaker 1>the reinvented two thousand twelve camera. It's ready, are you