2024-10-18 - IIT Kgp Qiskit Fall Fest.pdf

Challenges in Quantum Compilation - IIT Kgp Qiskit Fall Fest 2024

1. Kharagpur Quantum Informatics
& Computing Club

2. Aritra Sarkar
Senior Researcher, Fujitsu Research
(former) Postdoc Researcher, QuTech, TU Delft
Challenges in Quantum Compilation

4. Quantum Computer Engineering
B.Tech Avionics
(specialisation: Space Robotics)
Indian Institute of Space Science and Technology
2009-2013
Scientist
Satellite Onboard Control and Digitals Subsystem
Indian Space Research Organisation
2013-2016
Postdoctoral Researcher
Quantum Machine Learning group
Quantum Computing division, QuTech
2022-2024
M.Sc. Computer Engineering cum laude + Ph.D.
Quantum Computer Architecture group
Department of Quantum & Computer Engineering
2016-2018 + 2018-2022
AIT
QCA

5. Motivations
Computer
Science
Quantum
Physics
❑from understanding to control
❑from passive to active
❑sensors, communication,
simulation, computation
❑from science to technology
❑new ways of computing
❑reversible
❑resource complexity
❑natural quantum simulators
Computer
Engineering
❑need for computing ever
increasing
❑faster computers → smaller
transistors → quantum effects
→ accelerators and ASICs

6. Quantum Accelerator
Operating System
MemoryCPU
Hardware Accelerators
QPU
(gate-based)
GPU
QPU
(annealing)
DSP FPGA NPU
ALU
Control
Data
Program
Shared
Control
Shared
Memory
API
(cloud PaaS)

7. Quantum Advantage

8. Quantum Computing Stack
Q Application Code
Q High-Level Programming Language and Compiler
Q Circuit Decomposition and Simplification
Q Assembly Language
Q Instruction Set and Micro-architecture
Q Error Detection and Correction
Q Optimal Control and Error Mitigation
Q Circuit Cutting, Knitting, Mapping and Routing
Q Circuit Scheduling
Q Device Characterization and Calibration
Q Processor Fabrication and Lab Setup

9. physical manipulation of
quantum hardware
Quantum Firmware
mathematical
abstractions of
quantum
algorithms
Q Application Code
Q High-Level Programming Language and Compiler
Q Circuit Decomposition and Simplification
Q Assembly Language
Q Instruction Set and Micro-architecture
Q Error Detection and Correction
Q Optimal Control and Error Mitigation
Q Circuit Cutting, Knitting, Mapping and Routing
Q Circuit Scheduling
Q Device Characterization and Calibration
Q Processor Fabrication and Lab Setup

10. mathematical
abstractions of
quantum
algorithms
QISA
Pulses
physical manipulation of
quantum hardware
QASM
Hamiltonians
Super-operators
Decoherence time
Unitary gates
Noise models
Complexity classes
YAQQ
compiler
CISQC
microarchitecture
EOPE
analog control
Quantum hw-sw interface
Heisenberg interface
classical finite descriptions

11. Are all QC alike?
Theorist: Yes!
•Church-Turing-Deutsch principle:
•Universal quantum computation set can simulate any physical dynamics
•QTM later generalized to the circuit model and k-local gate sets ({H,T,CX}, {D(θ)}, {Toffoli,H})
•Solovay-Kitaev theorem:
•Iterated shrinking lemma: find better approx. of U by recursively inc. decomposed circuit length
•Universal QGS approximate the Hilbert space with ??????(??????�
3.97
Τ
1
??????) gates in ??????(??????�
2.71
Τ
1
??????) compiler time
•Can be generalized to n-qubits/��(??????) at exp. scaling
https://inqnet.caltech.edu/wormhole2022/
https://phys.org/news/2013
-
05
-
principle
-
nature
-
quantum.html

12. Are all QC alike?
Experimentalist: No!
•DiVincenzo criteria:
•Scalable (bounds circuit max. width), well-characterized qubits
•Long decoherence time (bounds circuit max. depth)
•Universal native gate set
•Initialization to fiducial state
•Measure individual qubits
•Qubit connectivity:
•Bounds information interaction rate between qubits
•Needs additional gates for routing (taking some share of max. depth)
•Benchmarking performance:
•Useful quantum circuits are neither trivial nor random
•Use generative AI
https://www.classiq.io/

13. [1] Quantum Compiler Optimization
•Decoherence limits circuit depth
•QPUs have different native gate sets
(discrete for controllability; FTQC has few FT-gates i.e., often not universal - Eastin-Knill)
Universal Quantum Gate Sets are not equivalent! (at least in the NISQ era)
•Equivalence → Expressibility (still holds) / Reachability (breaks)
•Given a max. circuit depth, some states are reachable with QGS1, some others with QGS2
•Given a target state, the fidelity using QGS1 is different from QGS2
(irrespective of no noise model and full qubit connectivity)
Project collaborators: Akash Kundu, Sibasish Mishra, Matthew Steinberg, Tamal Acharya, Jaroslaw A. Miszczak, Sebastian Feld

14. [1] Quantum Compiler Optimization
Quantum Algorithmic Information Theory
•(Quantum) Algorithmic Probability
- Probability of a particular output by a random program
- Generalized to Quantum Circuit Probability
•(Quantum) Kolmogorov Complexity
- Shortest program for a universal computer that outputs s
- Most states have exp. exact circ. dcmp using discrete QGS (does not
violate SKT)
- Most states/gates are only approximately expressible using classical
programs (similar to how most real numbers are transcendental)
Project collaborators: Akash Kundu, Sibasish Mishra, Matthew Steinberg, Tamal Acharya, Jaroslaw A. Miszczak, Sebastian Feld

15. [1] Quantum Compiler Optimization
Project collaborators: Akash Kundu, Sibasish Mishra, Matthew Steinberg, Tamal Acharya, Jaroslaw A. Miszczak, Sebastian Feld

16. [1] Quantum Compiler Optimization
check it out here
Project collaborators: Akash Kundu, Sibasish Mishra, Matthew Steinberg, Tamal Acharya, Jaroslaw A. Miszczak, Sebastian Feld

17. [2] Quantum Microarchitecture Optimization
QC resources
•Time/Clocks/Gates
•Space/Memory/Qubits
•Total Complexity(Brown, Susskind, et.al.)
�
���????????????=�
�??????��.+�
����.
•Thermodynamics of Quantum Computation (Zurek, Kolchinsky, et.al.)
δ�??????→�+log
2
1
�(�|??????)
+???????????? ≳??????(??????|�)
•Logical Depth (Bennett)
�
�
′
??????= min
�
{��:�−�
∗
<??????∧(&#3627408456;&#3627408477;≈??????)}
•Levin Complexity (Levin)
??????
&#3627408481;??????=min
&#3627408477;
{??????&#3627408477;+log??????&#3627408477;:&#3627408456;&#3627408477;=??????}
Quantum
Advantage
Time
Space
Fidelity
Desc.Cplx
Energy
...
Quantum Circuit Computational Complexity
Quantum Volume for QPU qubits+noise
Project collaborators: Sibasish Mishra, Sarang Gosavi, Sebastian Feld

18. Cryogenic-CMOS for Quantum Computing
[2] Quantum Microarchitecture Optimization
CISQC: Complex (energy-efficient) Instruction Set Quantum Computer

Energy ≡ Quantum Circuit Description Complexity
≡ bits of classical information required to erase a qubit
≡ least no. of control bits to uncompute/transform the state to a known state
≡ least no. of control bits to prepare the current state from a known state
≡ coding and compressing scheme for quantum circuits
•Pragmatic application:
•Common in assembly languages and low-power embedded systems
•Optimize instruction bandwidth in Q-C Ctrl. I/F
•so far,.. optimize QEC, Routing, Pulse using ML at cryo-level
•trade off 4K-RT compressed ins. bw with ASIC/FPGA proc.
•2.46pJ/b, 40Gb/s
•Assign opcodes in QISA
Project collaborators: Sibasish Mishra, Sarang Gosavi, Sebastian Feld

19. [2] Quantum Microarchitecture Optimization
Coding theory and lossless compression
•High probability of symbol -> Low code size
e.g., Morse code
•Use statistical compression algorithms of QASM files
•Compress a code that the micro-architecture can decode
•i.e., at the QISA level instead of QASM
•QISA instead of Pulse for FTQC (RT QEC)
Project collaborators: Sibasish Mishra, Sarang Gosavi, Sebastian Feld

20. [2] Quantum Microarchitecture Optimization
Coding and Concept Discovery
•Ver-0: Binary coded Native Gate Set
•Ver-1: Huffman coded Native Gate Set
•Ver-2: Huffman coded SK Basis Approx.
•Ver-3: Huffman coded SK Basis Approx. + cutoff
Project collaborators: Sibasish Mishra, Sarang Gosavi, Sebastian Feld

21. Quantum Speed Limits for U
•Discrete: Σ (# gates * gate-time)
•Continuous: Energy = Geodesics on Riemannian manifold = Lagrangian mechanics
•Optimal for constant axis, constant ang. momentum
Accessible controls
Optimized universal controls
[3] Quantum Control Optimization
Project collaborators: Sebastiaan Fauquenot, Sebastian Feld
DOI: 10.1103/PhysRevX.9.011034
DOI: 10.1088/1367-2630/ac6821

22. [3] Quantum Control Optimization
Optimal quantum control – pulse-level gate synthesis
•Trade off fidelity and energy
•Energy-optimized Gradient Ascent Pulse Engineering (EO-GRAPE)
Project collaborators: Sebastiaan Fauquenot, Sebastian Feld
8% ↓ energy
1% ↓ fidelity (< P
thresh)

23. [3] Quantum Control Optimization
deep reinforcement learning agent
... learns pulse-level unitary synthesis
...... from measurement feedback
......... on noisy quantum processor
Project collaborators: Sebastiaan Fauquenot, Sebastian Feld
U = H
w
fe = [0.8, 0.2]

24. Quantum-accelerated Genome Sequence Reconstruction
Project collaborators: Carmen G. Almudever, Zaid Al-Ars, Koen Bertels

25. Algorithmic Information and Quantum Computation
Project collaborators: Zaid Al-Ars, Koen Bertels

26. Quantum Information Theory guided Quantum Compilation
Project collaborators: Abhishek Sadhu, Akash Kundu, Matthew Steinberg, Medina Bandic, Sacha Szkudlarek, Carmen G. Almudever, Sebastian Feld

27. Generative AI for Quantum Computer Engineering
Project collaborators: Boran Apak, Medina Bandic, King Yiu Yu, Ryoichi Ishihara, Sebastian Feld

28. eXplainable Quantum Machine Learning
Project collaborators: Shubing Xie, Akash Kundu, Abhishek Sadhu, Vedran Djunko, Sebastian Feld

29. other Q Algo and Q Compiler projects…
Project collaborators: Tamal Acharya, Akash Kundu, Hongfeng Zhang, Koen Bertels, Emile Larroque, Zaid Al-Ars, James Boyd, Bao Bach, Anneriet Krol, Alan Yu,
Maximillian Russ, Abhishek Sadhu, Deepika Rajan, Nikiforos Paraskevopoulos, David Hamel, Matthew Steinberg, Sebastian Feld
•QaCHT: Causal modeling of gene regulatory networks
•QOptRNA: mRNA codon optimization
•QraftRNA: Inverse RNA folding
•RuliadTrotter: Modeling metamathematical observers
•VQCP: Visualizing quantum circuit probabilities
•OpenQL_ud: Efficient unitary decomposition
•OpenQL_pc: Parametric circuits compilation
•MLC4QC: Machine learning control for noisy gate-based quantum computers
•AutoQECC: Automated discovery of Quantum Error Correction codes
•ArtA: Artificial Architect for spin qubit architectures
•SpinAME: Design space exploration of spin qubit processor architectures

30. Way ahead…
AutoQC