As someone that learned Computer Science in the 1970’s at a time of the mainframe and machine code, of 8 bit processing and partitioning CPUs, I have seen many tech waves come and go… PCs, Internet, Distributed, Handheld, Smart Phones, Blockchain, IoT and others…
How things change. Today we have entered the new Quantum Computer realm, but in reality its not that new. It is a technology built upon the laws of nature, of the universe, or so Quantum Mechanics now defines things, moving away from classical laws that have been found to be wrong.
So here is the problem. If you think you understand Quantum Computing, part of Quantum Information Sciences, you have confirmed you probably don’t. Many of the world’s leading physicists freely admit they are still coming to terms with its potential, and do not fully understand what is going on. But most universally all agree it is both a revelation and danger to humanity (for good and for bad) at the same time, and advancement so profound is itself in superposition.
It is easy to be deceived, where most people will rely on their knowledge of Classical Computing (Turing Machine), and the norms of what you think you know about Chemistry, Biology and Physics that place rules, laws and order on what appears to be natures natural order and chaotic world. Most of us seek comfort and therefore reference our past learning and experiences, and the things we were taught at school. However unfortunately none of this is helpful to understanding how the Quantum world works and it is most notably wrong...
The first thing to mention is Quantum is unlike any else. When applied to computing or computation it forms is an entirely new species. Like the creature in ALIEN, the rules of what has come before simply don’t apply. Acid for blood, no eyes (yes I didn’t realise at first), a killing machine that doesn’t breathe what we do, and propogates remotely through inplanting eggs in any living host. Something entirely new. From which drawing any parallels are void.
Humans are about to enter the world of Modern Computing…
Yes that’s right. Quantum Computing in many ways is the beginning of humankind's efforts to understand the basis of life, as we discover the world around us is not as probabilistic and deterministic (organised) as we think it is. Quantum Computing is the start of the journey. We are once again at the beginning. As Turing perfected the maths and switch logic for classical computers in the 1930’s, the first machines of the 1950’s prompted Thomas Watson at IBM to annouce the world probably won’t need more than a handful of these machines. Deja vu, you bet…
Quantum Computing in many ways defies understanding. But we can overlay a new foundation of maths, algorithms that help us control and unlock the potential of atomic particles and get them to do the work for us. But we have also seen in 1945 unlocking their enormous power can work against us all. And there is the challenge, how to work with atomic particles that do not follow it seems any rules. Not ours anyway, yet seem to follow something, organised by the things that holds the universe and life itself together.
As Quantum Mechanics in the early 1900’s tried to fathom how atomic particles behave, they soon realised photons, ions and electrons (commonly used in Quantum) do not follow the exacting rules of maths and physics. They don’t fit models, so they created new models. The test. Which path will the photon take when presented with two slits, the left, right or both at the same time? And there it is, the weirdness of what we are trying to work through.
But in another sense does it matter? Early Quantum Machines assume the particles can hold different states, which we will come back to, at the same time… You may also ask why don’t they know? The answer is of course that when you try to measure what is going on, the particles loose their special effects and you end up with a single state anyway. There is so much weirdness to consider, even Einstein struggled with certain aspects of Quantum — specifcally the concept of entanglement.
So what is all the fuss about?
I explain our journey into the Quantum realm like this… It may surprise you but the greatest invention of humankind is the Telescope. Yes that’s right. In 1600 is allowed us to understand our place in the stars and confirmed to world wasn’t flat. But then we turned it inwardly to create the ‘micron scale’ that allowed us to deeper and start to understand how things work, cells, proteins, and DNA, as we moved into the atomic realm of particles and have reach the sub atomic place of up and down Quarks.
In essence this is the world of , fields, vibrations, waves and the behaviour (random or otherwise) as very small objects that wiz around and seem to follow instructions, but seemingly not ours. Our bodies are made up of these atoms formed from cells, which Turing called Morphogensis. When you start to think about it, these small particles are influenced in some way or we would be a pile of slim and jelly on the floor. How are we held together? Is it something to do with magnetic and electromagnetic fields? How come consciousness exists? To be conscious? My head hurts already.
But here is the thing that demonstrates the true weirdness of Quantum realms. When all cells are formed they are all exactly the same. But then they start to form groups, and patterns which form different things — skins, organs, bones, eyes, other tissues, a term Turing called ‘differentiation’.
But how do they know what to do…? When they all start out the same.
So we think we understand how life is created and our place in the universe. But it is clear we do not! Which is exciting and scary at the same time. I remain in two constant states myself, most of the time. Hope at the potential and what lies ahead, and concern that in the wrong hands humanity will be taken to the brink.
Old tricks die hard
In the current world of Classical Computers the laws of mathematics, the bahavior of electricty and magnetic fields, to design our computer chips around simple input and output processes. A linear approach, stacking transistors onto silicon to increase computational power. Add more transistors, many billions, and you theoretically get exponential processing power. Well, sort of right but also wrong.
The speed of our computer processors (dual core, quad core) expressed in GHZ — Gigherz is a clock frequency that indicated cycle time processing efficiency. However in the last 10 years there has been no significant performance improvement because we are reaching a natural end point with transistors now so small, 10 atoms across, it has entered the atomic world where excessive heat (energy consumption) delivers more interference contributing to a performance drop off in performance. It is worth remembering tiny atoms are easily disturbed.
Quantum Computers are designed in a way that creates new rules and maths (algorithms) to influence the behaviour of photons, ions and electrons through interference. Yes interference. You see these little atoms don’t understand our deterministic rules and their behaviour cannot really be measured, but we do know when the conditions are right they deliver Quantum effects that can unlock ‘brute force’ computing power.
But then clasical physics suggest one cannot observe the bahviour of what is going on. Looking inside to see if the Cat is dead or alive… Therefore taking both states at the same time. But then the observer (one of us) comprises of atoms and therefore observing merely pushes the outcome to one of two states, we observe the Cat as either Alive or Dead.
We are creating new machines, each set on trying to influence, understand and get these tiny atoms to do what we want, as they spin forming Superposition (with probability halving), as more than one state can be achieved at the same time. A series of Quantum Gates can then exploit these atomic behaviours, as Quantum effects, and it is this teams at IBM, Intel and Google are flat out trying to perfect.
As many say Quantum Computing is decades away, I disagree. The primary reason is the ‘rate of learning’ is exponentially increasing, as ML and AL merge with Quantum, and the community is also growing faster. Capital is flowing much faster, as National Security interests are heightened, as all of a sudden the potential for ‘modern encryption’ is rendered meaningless acts as an accelerator. Presumably to be first to see what everyone else really has and is hiding. Although Quantum Computing is maturing fast, the new quantum economy and industry stacks and layers are being formed, so we can all benefit from and enjoy the benefits of Quantum Computing.
But then we should not forget the Hippocratic Oath as we assume We Can, We Are pushes us ahead further into the sub-atomic world, rather than asking Should We? Something that Oppenheimer 16th July 1945 and his team from Los Alamos demonstrated. Splitting the atoms of Plutonium and Uranium unleashing death!
So what is it that gives a Quantum Machine the power…?
Many of you reading this will already know a ‘Qubit’ supports 4 states simultaneously (00, 01,10,11) where a binary Bit has only one state, a Zero or One. What happens next is entanglement, the sub atomic pairing of these particles across space and time (or in machines and on chips) where they become entwined and mirrors eachothers properties, offering more and more states to play with — in this case Spin Up and Spin Down, assigns a zero and one. But the reality is nobody really knows what is going on, but we do know when you can keep Qubits entangled you get an exponential increase computational power, as multiple states deliver the answer quickly is all that matters.
We are trying to influence the atomic world, overlaying new mathematics and algorithms so that we may understand this realm, get something from it, and yes it appears to be working.
Time to sit down, or go into a dark room. Time to contemplate how exponential scaling is just around the corner. As we know each Qubit can maintain multiple states at the same time being both a 0 and 1 and everything in between at the same time, called Superposition. A binary Bit has ‘one state’, so you repeat processes one after another, limited to clock speeds and the underlying behaviour of silicon. This is why you need so many transistors on a chip, to increase computational power, but even then performance is limited.
A Qubit starts with 4 simultaneous states, adding a Qubit doubles this to 8 states, 3 has 16, 4 has 32 and when you get to 250 Qubits, this has more states than there are atoms in the universe. £U€&!, that’s what I thought!
We have established a Qubit can assume multiple states at the same time, which gives us the exponential scaling opportunity as you add more Qubits. It is worth remembering the teams at IBM, Google and other places are trying to influence the behaviour of these atoms to keep the entangled, achieving coherence. In effect trying to control these particles by interference, to exploit the potential of these atomic particles. Whether using cold temperatures of deep space with Superconducting approaches or Trapped IONs or Photons using lasers, different magnetic fields and even creating new ‘fake atoms’ that can take on certain behaviours that can influence and keep the fidelity of entangled Qubits.
Each approach tries to influence Qubits to help them come out with the right answer and find and eliminate the wrong answers. A bit like Headphones with noise cancelling filter that identifies unwanted signals and background noise and filters this out. Imagine if you only focused on the right answers!
For me the single biggest weirdness is entanglement where photons can become essentially paired (twinned) many light years apart, across space and time, it starts to freak one out a little. This is also the very basis of a new era in Quantum Communications, where the Sender and Receiver share the properties of two photons that are paired, and where any attempt to eavesdrop interrupts the entangled state ending the channels of cimmunications. Flipping to a new channel and a new pairs of photons to continue. Think Micuis 2016 that changed espionage. It is also worth remembering an entangled state delivers the ability to generate pure random numbers, adding new levels of Quantum Encryption potential to the mix.
The universal problem for all teams developing Quantum Machines is how to keep the fidelity of the Qubits, keeping them entangled long enough to do something useful, e.g. applying (Quantum Gates) Not and Swap gate logic to deliver computation potential. Given they are extremely sensitive to external vibrations, noise, heat and other things. All of which delivers decoherance and the computational advantage collapses.
The other challenge is programming Quantum Machines is very difficult and there remains a lack of software and virtually no applications that can benefit from Quantum properties. The industry starts to mirror the early days of mainframes of the 1960’s, as we look to create a full Quantum Service Tech Stack for customer consumption.
But despite limitations which are structural at this point, Quantum Computers are really good at certain things, for example factorising big numbers, especially now we have Shor’s Algorithm. In other words Quantum Computers render all encryption completely useless, unless you use Quantum effects described above to make everything Quantum Safe. This is the attention grabber.
Many people talk about Supremacy, which is a misleading concept. This involves choosing a narrow problem that a Quantum Machine can achieve millions/billions of times faster than classical machines. This is the first stage in defining a new Quantum Age.
Most teams are working on stability and fidelity issues, to find ways of plugging (errors) and gaps to keep the number of Qubits entangled, as they plan to use hundreds of Qubits to create robust Virtual Qubits environment that will move us to the next phase. Allowing new applications and hardware to become inherently stable, thus commercially useful, that will define a new era in Distributed Quantum Computing and Networks.
It is here, Quantum Computing that is, but it is also not yet here. We are living at an All at Once time where there are many pervasive technologies arriving. When Quantum meets AI, which is starting to happen, the rate of learning will increase, and the path to SuperIntelligence will simple arrive.
Nick Ayton © 2020