牛津仪器集团成员
扩展
Harnessing quantum mechanical weirdness to do computing

It was during my masters at EPFL Switzerland that I was first able to fully appreciate the mind boggling world of sub atomic particles. Here, classical laws proposed by Newton fail and an entirely different framework of called Mechanics is needed to understand the workings of this fascinating world. This world is full of eerie effects like particles existing in multiple states at the same time, ability to pass through walls etc. As the great Danish physicist, Neils Bohr who laid the foundations of mechanics in the 1920s aptly put it:

        “Anyone who is not shocked by has not understood it!”

Fast forward to today and scientists have not only gained a great deal of (note: I did not say complete) understanding of the world but in what is considered one of the greatest examples of human creativity and enterprise, but also beginning to harness this “weirdness” for complex computation.

Let’s take a step back to appreciate its impact. Modern day computation (yes I mean literally every device in proximity to you right now) uses classical bits which have a value of 0 or 1. Using large numbers of bits and a way to control them (logic gates) enables all classic computation. A bit or a allows its state to exist in 0,1 and states that have shades of state 0 and 1, these states are technically called superposition states. Using gates made from qubits one can create a superposition state and encode in parallel in these states. In order to understand the profound implications and sheer power of this ability to encode parallel let’s consider an example problem. There are four cards on the table. Three of them are identical say king of hearts and one is a queen. Now the cards are shuffled and placed face down on the table and the objective is to find the queen. How would you (and incidentally a classical computer) go about it? It would have to be something like: open card one, then the other etc. Considering the queen can either be found on the first or the fourth attempt, with some simple math you can deduce that it would take around 2.5 attempts on average to do find the queen over many attempts to find it. In contrast using a 2 system this can be done in a single attempt! While this experiment might not sound ground breaking let’s look at a more complex problem, say factoring a large prime number 129 digits long, this happens to form the back bone of today’s online encryption protocols. A classical computer would take an extremely long time to factor such numbers into its prime components (keys) which is why it is considered secure. In fact it took 8 months and 1,600 internet users to crack RSA129, a computer can do this in seconds!

While internet security is interesting scientists are already working on ways around it (Read about Quantum Key Distribution (QKD) if you would like to know more), its impact will be most seen in its application in drug discovery, financial risk management, AI and most importantly understanding quantum itself! As Richard Feynman pointed out in his famous Caltech lecture, we need objects to simulate .

Great!! How does one go about making such qubits?

That’s where the challenge lies. A system by its very is fickle and changes its state even with very little perturbation. Isolating and controlling a state in a and maximizing the duration of its existence before it collapses into a classical state is at the heart of research and technology development. There are several different technologies in play today:

 1. Ion traps: Ions or atoms localized by or laser beams. This works well for a few qubits but scaling is difficult and the challenge lies in initialization, control and measurement of large array of trapped ions. 
 2. Nuclear qubits: Here is coded on to nuclear states. This can be achieved by implanting dopants in atoms. While inherently difficult to scale up as it requires deterministic fabrication of coupled spins, progress is being made in this direction.
 3. Photonic qubits: Polarization of light in photons is used to carry out computation. This has traditionally required complex instrumentation in controlled environments but progress in silicon based approaches promise a great scale up potential. The main advantage of this technique is a lower level of stochastic noise compared to mater based qubits
 4. Superconducting qubits: Uses Josephson junctions in superconducting materials to create either , charge or flux (other flavours also exist today), this is the most used form of for research today.
 5. Topologically protected qubits: These are versions of qubits currently in early stage research where using certain properties of materials to protect the state in subatomic particles. Some of the popular approaches involve - junctions and insulators to create and operate on strange, yet fascinating particles called fermions.

It is clear that technology has arrived but is at an early stage of its evolution. At Instruments we are working on several technologies that are helping with the progress of this mind boggling yet powerful technology. At plasma technology we enable the fabrication of materials like superconducting Niobium Nitride for fabrication of flux qubits as well as single photon detectors. We also have a great deal of expertise in device fabrication processes for on chip , solid state lasers and deposition of -dielectrics dielectrics for fabrication. Instruments also enables by offering the  environments and measurement technologies necessary along with extremely efficient single photon detectors for photonic routes to . We would be delighted help you with your challenges so please get in touch with us at plasma-experts@oxinst.com or visit plasma.oxinst.com

Author: Dr. Ravi Sundaram

沪ICP备17031777号-1 公安机关备案号31010402003473