Most of you guys have heard about Moore’s law by now. It is the observation that the number of transistors in circuits doubles roughly every two years, thus, increasing the power of our devices. Is this phenomenon grinding to a halt and can we do anything about that?
The light switch is a pretty simple concept really. You got a light bulb; you got electricity and a switch that allows electricity to pass through.
This is exactly how the transistor works on a computer. It’s a tiny little gate that opens and closes, allowing electricity to pass through. Because that is a binary system – it’s either off, or it’s on. ( Which translates to a 1 or a 0. ) Each 1 or 0 is a bit and 8 bits create a byte, or an ASCII code that the computer uses to process information.
The more of these tiny little switches you have in your computer, the more information it can process.
The computer you’re reading this article right now has over a billion transistors in it. This makes it right in line with one of the most accurate predictions in tech and science in the last 50 years.
A prediction, whose time may be up.
In 1965 Gordon E. Moore was the director of R&D at Fairchild Semiconductor, when Electronics magazine asked if he would write an article about the future of computing. In the article he wrote, he pointed out that the number of transistors they could fit on a chip had doubled every year, and he predicted that this would continue for the next 10 years.
And 10 years later he became one of the co-founders of Intel.
Then he predicted that it would double every two years. At this point it became known as Moore’s law.
Along the way some other factors got wrapped up in the law, including performance and cost.
David House later on revised it, saying that because of the speed of the transistors was improving; it would actually double the performance every 18 months. ( And that is what we’ve been going with ever since. )
However, Moore’s law itself was always about the number of transistors that you can fit on a single chip.
This exponential growth in computing power has revolutionized the world.
Still, there is a growing concern that Moore’s law may have reached its limit. Some people even say that we are already at the end of it.
In fact, research shows that the processor speed is getting kind of leveled out in recent years.
So, why is this slowdown happening?
What are the limits we are running into?
Well, one problem is simply heat. The more processing a chip is doing, the more electricity is flowing through it and the more heated it becomes.
The second problem is size. ( Yeah, that is what she said … oh no … wait! ) Once those gates are down to a certain size, the quantum world starts to take over. Thanks to something called quantum tunneling, electrons tend to jump the gap. ( Which is a problem in a binary system, where it has to be totally ON or OFF. )
Luckily, there are some major advancements in technology that might keep this exponential train going.
The first one is graphene processors. The reason we use silicon in chips is because of its semiconducting properties, which makes it a really great on-off switch. ( It also just happens to be one of the most abundant resources in the world. ) Graphene transistors can conduct electricity much more easily and increase clock speed by a factor of 10, or even 100. That means way lower power and far less heat.
In fact, in a paper that was published last year from a team of researchers in Moscow and Japan, graphene processors have proven to make use of that quantum tunneling effect.
Graphene, if you don’t know, is a one atom thick layer of carbon atoms that are arranged in a hexagonal pattern. Still, until we can make it as cheap as one of the most abundant materials in the planet, it’s still a ways in the future.
Another idea is tree-dimensional chips. A team at Stanford has created a three-dimensional chip that uses carbon nanotubes to stack memory on top of the processor. ( That way it’s all compacted into one place. ) They say that this makes it a thousand times faster then a normal computer chip.
The next one is molecular transistors. In 2009 the first molecular transistor was created. It used a molecule of benzene, suspended between two gold contact points to serve as an on/off switch. Another was created in 2015 that actually had the ability to transfer only one electron at a time, but it has to be super cool to work, so, not really viable in mass computing.
Next up is photon transistors. What if instead of using electrons to do the on/off switch you used photons. In 2013 scientists at MIT created the first optical transistor which used beams of light instead of flows of electrons.
It is really cool that they were even able to send one photon over at a time, but this is still very primitive and the amount of information that they can process is very little.
You simply can’t do a list of future computer technology without mentioning quantum computers.
( This disserves an article on its own. )
Basically, the reason you hear so much about quantum computing is because it would be a gigantic leap forward in how we process information. With them, we are getting away from the binary system completely.
In the quantum world you have both on and off, but then on and off at the same time. So theoretically, the computer would be able to run parallel processing, doing multiple calculations at the same time. ( Which allows for a much faster, especially data encryption. )
Ok, so , now we are going to get into some really weird territory.
One idea that’s been floating around is protein computers.
Researchers at Lund University in Sweden were able to make a working computer, the size of a book that uses proteins to do its calculations in a parallel fashion. ( Parallel processing. ) It worked and used far less energy than its traditional counterpart.
Last, but not least, we have DNA computers.
The idea that DNA can serve such purpose has been around for a little while now. ( And it kind of makes sense when you think about it, because what is DNA, but a place to store genetic information. ) The cool thing is that it consists of two different types of base pairs – guanine & thyamine, adenine & cytocine. ( And only those two with each other … so … that is a binary system )
Using gene sequencing techniques, the team at Harvard in 2012 was able to put 700 TB ( Terabytes ) on 1 gram of DNA. Just to put things into perspective this thing will weigh 150 kg with our current technology. So DNA might be the perfect information storage solution. It’s super dense, it lasts forever and it is super abundant. ( Like in every living thing. )
But, what about computing power?
Well, Leonard Aldeman first proposed DNA processors back in 1994. He even built one that was able to do some rudimentary calculations, but it was very primitive and required people to be involved, so, it wasn’t anything practical.
More recently, a team at the University of Rochester created logic gates out of DNA strands. ( Still pretty primitive, but a cool step in the right direction. ) Once perfected, they think that a teardrop sized DNA computer would be as powerful as all the computers on earth.
Computers like this that work on a molecular level, could lead to nano machines we can inject into our bodies, to repair tissue, improve brain function, and all kind of things we can’t even imagine right now.
Who knew that the most powerful computer in the world was inside us all along.
These are some exciting technologies; unfortunately, most of them are still pretty far into the future. Looks like we are going to have a slowdown to Moore’s law. ( Bummer 🙁 )
But, while Moore’s law may be coming to an end, we need to pause and reflect on just how important it is. It was never actually a law, like gravity or thermodynamics. There was no inevitability to this. It only happened, because engineers and scientists worked year after year, grinding out new innovations and new ideas to keep pace with this prediction.
In the end, there is an argument to be made. This arbitrary prediction was able to focus everybody’s attention on these goal posts, which we’ve set along the way. Let’s not forget that the guy who made the prediction, also just happened to be the lead of the biggest chip manufacturer in the world. ( That also helps. )
Still, we do think that these expectations are important. Futurists like Ray Kurzweil are often ridiculed for their fantastical visions of the future, but those predictions lay down a path for us to follow. It gives society a focus and creates a self-fulfilling prophecy that moves everybody forward. So, having a road map matters, whether as a whole society or just as an individual. Draw your own map and see where it takes you.
The post above was made from the following materials: