1. Quantum computing has arrived
Sergii Nechuiviter
AI Club

2. This lecture is NOT about:
• Quantum mechanics
• Mathematics
• Algorithm complexity
It mostly about state of industry of quantum
computing

3. Quantum computing?
And what?

4. Quantum computing is cool!
45 Algorithms with substantial acceleration
• Polynomial (for Searching, etc.)
• Superpolynomial (for Discrete-Log and
Quantum Simulation, etc.)
Other solvable NP-complete problems:
• Boolean satisfiability problem
• Knapsack problem
• Travelling salesman problem
• Subgraph isomorphism problem
• Subset sum problem

5. Practical applications
• Hacking open-key cryptography
• PageRank (Google in a Quantum Network)
• Searching a database of molecules
• Lattice protein folding
• Sentiment analysis
• Car detection
• Labeling of news stories and images
• Video compression
• Machine creativity
• Physics simulation
(30 quantum particles simulation is limit to our days computers)

6. If it is so cool
Why haven`t I heard about it?

7. Brief recognized history of
quantum computation research
• In 1981 Richard Feynman described a theoretical device that he called a “quantum computer”
• In 1994 Peter Shor described the first quantum algorithm
• In 1993 Seth Lloyd proposed the first feasible design for a quantum computer
• In 1998 IBM-Berkeley, Gershenfeld, Isaac Chuang, Mark Kubinec – the first real experiment: 2 qubits
• In 2001 IBM-MIT Chuang a 7-qubit NMR computer (prime factors of 15)
• In 2000 MIT Edward Farhi, Jeffrey Goldstone, Michael Sipser, Northeastern’s Sam Gutmann
proposed an adiabatic quantum computer
• In 2005 IBM, Hewlett-Packard and NEC not seems likely to deliver a working machine in less than 10 years.
• In 2007 Japanese electronics giant NEC Corp, controlled 2-qubit coupling
• In 2011 University of California, 2 qubits, quantum bas, 2 quantum RAM
arxiv.org/abs/1109.3743: Implementing the Quantum von Neumann Architecture with Superconducting Circuits
• In 2011, arxiv.org/abs/1103.1528: A Single-Atom Quantum Memory, can store qubits for only 180 microseconds
and has an overall efficiency of 9 per cent
• In 2011 Experimental realisation of Shor's quantum factoring algorithm using qubit recycling, 1 qubit
• In 2011 IBM-MIT Chuang factorized 143 using 4 qubits
• In 2012 IBM-Yale stated that they work on 3D qubits
• In 2012 arxiv.org/abs/1205.4899: Controlled-NOT Gate Operating With Single Photons
• In 2013 University of New South Wales in Sydney, single phosphorus atom in silicon (Kane design, 1998)

8. State of the Art: Grail of quantum computing research

9. NIST quantum
simulator (350 qubit)
Rose’s Law: D-Wave success story
Dairy Cow
Research here?

10. D-Wave impact
• 2006 4 qubit Calipso
• 2007 16-qubit Europa II/Orion demonstration of “the world’s first commercial adiabatic quantum computer.”
• 2008 28-qubit Leda
• 2011 128-qubit Rainer based “D-Wave One” for Lockheed Martin
• In 2011 University of California, 2 qubits, quantum bas, 2 quantum RAM
arxiv.org/abs/1109.3743: Implementing the Quantum von Neumann Architecture with Superconducting Circuits
• In 2011, arxiv.org/abs/1103.1528: A Single-Atom Quantum Memory, can store qubits for only 180 microseconds
and has an overall efficiency of 9 per cent
• In 2011 Experimental realisation of Shor's quantum factoring algorithm using qubit recycling, 1 qubit
• In 2011 IBM-MIT Chuang factorized 143 using 4 qubits
• In 2012 IBM-Yale stated that they work on 3D qubits
• In 2012 arxiv.org/abs/1205.4899: Controlled-NOT Gate Operating With Single Photons
• In 2012 NIST 350-qubit quantum simulator (news link)
• In 2013 University of New South Wales in Sydney, single phosphorus atom in silicon (Kane design, 1998)
• 2013, upgrade to the 512-qubit D-Wave Two for LM
• 2013 USC Dr. Daniel Lidar about error correction
• 2013, 512-qubit D-Wave Two for The Google / NASA Quantum Artificial Intelligence Lab
Applying for time on the D-Wave Two at the Quantum Artificial Intelligence Lab here
• 2013 As of September 20, the quantum photonic processor housed at the Centre for Quantum Photonics at the
University of Bristol will be available to researchers from anywhere in the world via remote access over the
Internet

11. D-Wave’s Quantum Computer Goes to
the Races, Wins
• 1999 D-Wave’s technology has been dogged by controversy during the 14 years
(from 1999) it has been in development, with quantum computing
researchers questioning whether the company’s technology truly is exploiting
quantum effects.
• In 2009 Google post “NIPS 2009 Demonstration: Binary Classification using
Hardware Implementation of Quantum Annealing” for car detection, 100ms per
iteration, QBoost
• 2011 Nature “Quantum annealing with manufactured spins”
• 2012 84 qubits Experimental Determination Of Ramsey Numbers With Quantum
Annealing in 250 ms
• 2013 A team of scientists at USC has verified that quantum effects are indeed at
play in the first commercial quantum optimization processor.
• 2013 Catherine McGeoch, Amherst College
D-Wave vs Lenovo workstation with a 2.4GHz quad core Intel processor, 16GB RAM
Quadratic unconstrained binary optimization (NP hard)
about 3,600 times more quickly: <½ seconds vs 30 minutes

12. Okay, it is fast
What is this D-Wave exactly?

13. Watch video
D-Wave Systems 128 qubit processor - Inside the chip
http://youtu.be/PqSgmCg1kew

14. Superconducting loop qubit

15. How does quantum mechanics help?

16. Principle of adiabatic computation

18. 8 qubit cell

19. Chimera 4x4x8=128 qubit topology

20. Vesuvius 8x8x8=512 qubit topology

23. Calibrating two D-Wave Two

24. Lovely
How can it be programmed?

25. Something familiar?

26. The crux of the matter

27. The energy program

28. Software stack of D-Wave

29. The Blackbox compiler

30. Tasks which can be solved
• Travelling salesman
• Hadamar matrix search
• Eternity II puzzle
• Weighted Maximum Independent Set
• Molecule Maximum Common Graph
• Quantum feature learning
• Binary classifier
• Neural networks
• NAND gate
• Physics simulation
• Etc.

31. Quantum computing has already arrived

32. ANNEX

33. QUANTUM COMPUTING AND
LIGHT SWITCHES
How it works

34. The light switch game

35. Bias summand

36. Interaction summand

38. 8
14
79
13
10
3
6
1
12
11
2
4
5

39. Evaluating Ramsey number R(3,3)
Note that the Ramsey experiments were carried out with an annealing time of
1000µs per run.

40. Example: Embedding of the primal graph of
ℎ3,3
6
(0, 𝑎2, 𝑎3, … , 𝑎15) into hardware.
(128 qubit chip used)