Quantum Computing: Don’t Panic
This article is part of the Developer Developer Developer 13 series
Meanwhile there were three other talks going on. My brother Peter was at Quantum Computing: Don’t Panic by Anita Ramanan and Frances Tibble and made notes on it.
Anita and Frances will also be presenting the talk at .NET Oxford on 3rd July 2018.
I didn’t understand even the basic concept of Quantum Computing at first, but then I watched another short talk by Dario Gil of IBM Research which explains the basics nicely. Microsoft’s Top 3 breakthroughs with Krysta Svore explains also explains why this technology is truly revolutionary.
Quantum Computing is a new and exciting technology, based on Quantum bits, or Qubits for short, which are based on the concept of superposition, a strange quantum mechanics phenomena that allows for matter to be in two places at once. It is fundamentally very different from classical computing.
With just a small number of Qubits, an enormous number of paths can be explored at the same time: 2 to the power of N. So for example 20 Qubits allows you to explore 1073741824 paths.
A Quantum computer requires cryogenic cooling, so it is unlikely that we will ever see Quantum mobile phones, but it we can expect to see Quantum computers that will be available for hire over the cloud.
Microsoft’s involvement with Quantum Computing was first announced by Satya Nadella at MSIgnite 2017. There is now a set of videos available on Microsoft Quantum Computing.
Possibly uses of Quantum Computing include nitrogen fixation, carbon capture, materials science and machine learning. Solving these problems will have major world changing effects.
Factorise huge numbers in polynomial time — we will soon be entering a new age where the today’s cryptography is useless. With classical computers, encryption information on the Internet is that it takes millions of years to brute force, but with a powerful Quantum computer this will only take minutes. We need to invent completely new cryptographic solutions that cannot be brute forced even with a Quantum computer.
Harber and Bosch — nitrogen in the air to ammonia for fertiliser — but very energy intensive
Enzyme in beans works at room temp, but can’t simulate in enough detail with classical computer — would need 1.5 x 1⁰⁵¹ bits — bigger than our planet?!
Or just 170 qubits (if perfect, in reality we need more because of instability/decoherence)
Theseus and the Minotaur — labyrinth, try one by one different paths — like search algorithm — quantum superposition of all possible paths — use interference to dampen incorrect paths and amplify correct path
Maths: pointy brackets called Ket
Alpha and beta: probability amplitudes — can use to say what is chance of 0 or 1
Entanglement: spooky at a distance
Biggest are 50–75 qubits right now — nowhere near what we need to really do useful things
Quantum error correction — spare qubits, but adding more qubits accelerates the problem
Scalable, Universal, Easy reset, Stable, Measurable…
Ion traps have some of the longest gate times, i.e. most long-lived, but still not stable enough yet
We just need them to last long enough for our operation
Atomic qubits, light qubits…
Microsoft taking a slightly different approach: topological quantum bits — much harder to perturb
Quantum computers sit in a “Quantum fridge” which is a lot colder than standard cryogenic computers — each layer gets colder — 15 milli kelvin — just above absolute zero, or around -452 degrees faherenheit!
Incredibly new physics — only really discovered in 2012
Moving towards having working qubits
QDK stands for Quantum Development Kit, and there is a new Q Sharp programming language.
The technology is Open source, is now available for Linux and Mac as well as Windows, and is interoperable with Python. There is a short 5 minute video available where program manager Jeff Henshaw summarizes the new updates.
You can simulate up to 30 qubits, or use Azure simulator for more than 40 qubits, but you will pay a lot for that due to the exponential computing that is involved, because these computations are simulated on a regular classical computer, not an actual quantum computer.
See Anita’s blog post The Hitchhiker’s Guide to the Quantum Computing and Q# Blog and her Quantum Workshop at github.com/frtibble/QuantumWorkshop
Also follow Microsoft Quantum on Twitter
The Speakers twitter links are @whywontitbuild and @frances_tibble
Microsoft expect to have a working qubit within a year, and a useful quantum computer within 5 years.
