KFUPM SQS² Quantum Software Stack

SQS² Quantum Software Stack

A national, hardware-agnostic stack that takes quantum programs from Python, C++ or Rust down to the pulses that drive the qubits, and runs them alongside high-performance computing.

Join the project How the stack works
01PROGRAMbell.py1from sqs2 import Circuit, run23qc = Circuit(2)4qc.h(0)5qc.cx(0, 1)6qc.measure_all()7run(qc, backend="auto")02COMPILE · AI-POWEREDq0q1√X√XRZ(π)SCHEDULE · ALLOCATE03EXECUTECPUclassicalGPUacceleratedQPUquantum

Overview

Quantum computers will work beside supercomputers as accelerators. Using them takes software that compiles one program for different quantum hardware, runs it reliably and coordinates it with HPC.

SQS² builds that layer as national infrastructure: vendor-neutral, based on open standards, and a stack national institutions can operate, extend and trust.

"We are building a sovereign AI-powered quantum software stack—deeply integrated with HPC—to accelerate the next era of national innovations in quantum computing"

Why it matters

A sovereign stack lets national institutions adopt quantum computing on their own terms and builds national expertise in quantum software, in support of Vision 2030 goals in energy, materials, finance and logistics.

How it works

Five layers lower a program from the user’s code to the pulses that drive the qubits. Figure 1 follows one program through them, and Figure 2 shows the middleware that links them to HPC and to classical and quantum hardware.

  1. L1 Applications & SDKsPython, C++ and Rust interfaces, with support for common quantum frameworks.
  2. L2 Quantum programming interfaceStandard APIs for submitting jobs and discovering available devices.
  3. L3 AI-powered toolchainCompilation and circuit optimization that adapt to the current state of the hardware.
  4. L4 Platform managerOne abstraction over all quantum processors, with scheduling across them.
  5. L5 Quantum control planeReal-time control and measurement of one or more quantum processors.
  6. QPU Quantum processorsSuperconducting and neutral-atom devices, driven by the pulses that L5 generates.
(a) Program L1–L2 
OPENQASM 3.0;include "stdgates.inc";qubit[2] q;bit[2] c;h q[0];         // superpositioncx q[0], q[1];  // entanglec = measure q;

A Bell-state program in OpenQASM 3.

(b) Compiled to native gates L3
q0q1CZ · 48 ns√X√X√XRZ(π)virtual
q0q1CZ · 48 ns√X√X√XRZ(π)virtual
(c) Pulse schedule, same time axis L5
D0q0 driveD1q1 driveCcoupler fluxMreadoutAacquireRZ(π): phase shift, 0 nsCZq0 + q1 · 1.2 µs04080120 ns1.32 µs
D0q0 driveD1q1 driveRZ(π): phase shift, 0 nsCcoupler fluxCZMreadoutq0 + q1 · 1.2 µs →Aacquire04080120 ns
In-phase (I) Quadrature (Q), DRAG correction Flux pulse Acquisition window
Figure 1. One program through the stack. (a) Source. (b) Compiled to the native gates √X, RZ and CZ; the RZ(π) is a zero-time phase shift. (c) The pulses the control plane plays, aligned with the gates above.
Middleware layer L3–L4
HPC environmentJob schedulers · storage · MPI
Quantum ApplicationsSimulation · optimization · machine learning
SQS² middleware
Scheduling
Resource allocation
Run-time management
Compilation and transpilation
AI-powered
CPUGPUQPU
Classical and quantum resources
Figure 2. Generic SQS² middleware. Workloads from the HPC environment are scheduled, given resources, managed at run time and compiled for the CPU, GPU or quantum processor best suited to each task.

Research directions

Quantum algorithm development

Designing, testing and benchmarking quantum and hybrid algorithms for practical problems, including but not limited to:

  • Combinatorial optimization
  • PDE solvers
  • Finance
  • Cybersecurity

Hardware-agnostic programming

Open intermediate representations, OpenQASM 3 and QIR, let one program target devices from different vendors.

HPC integration

Hybrid quantum–classical jobs run through the schedulers HPC centres already use, such as Slurm.

AI across the stack

Machine learning for compilation, calibration, readout and error mitigation.

Distributed quantum control

Coordinating several connected quantum processors as one resource.

Security of quantum computing

Protecting quantum programs, devices and remote access: secure job submission, isolation between users on shared machines, and trustworthy results.

Hardware platforms

SQS² is developed and tested on laboratory hardware, from cryogenic systems to control electronics, and targets two qubit technologies.

Superconducting qubits

Microwave circuits on a chip, cooled to about 10 mK. Each gate is a shaped pulse tens of nanoseconds long, so the lowest layer of the stack works in pulses, not gates.

Neutral atoms

Single atoms held in optical tweezers. Loading is random, so atoms are rearranged into a defect-free register before each run.

300 KRoom temperature50 K4 K800 mKStill100 mKCold plate10 mKMixing chamberDriveFluxReadout inReadout outNbTi coax20 dB20 dB20 dB10 dBLPF20 dB20 dB20 dBHEMTIsolatorsControl electronicsFPGA, DAC and ADC, driven by the SQS² control plane (L5)Quantum processor
300 K50 K4 K800 mKStill100 mKCold plate10 mKMixing chamberDrive andreadout inFluxReadoutoutNbTi coax20 dB20 dB20 dB10 dBLPFHEMTIsolatorsControl electronicsFPGA, DAC and ADCSQS² control plane (L5)Quantum processor
Typical wiring of a superconducting-qubit cryostat, after Krinner et al., EPJ Quantum Technology (2019).

Cryogenics and control

A dilution refrigerator cools the processor in stages from 50 K to 10 mK. Controllers and FPGA boards generate and read the pulses in real time.

People

Project lead Dr. Muhamad Felemban Assistant Professor
Computer Engineering · Information & Computer Science
Director, IRC for Intelligent Secure Systems

Faculty

  • Dr. Saleem Rao Associate Professor, Physics Superconducting circuits and device physics
  • Dr. Sajjad Mahmood Associate Professor, Information & Computer Science Quantum software engineering
  • Dr. Alawi Alsaggaf Senior Lecturer, Information & Computer Science Post-quantum cryptography and information security

Graduate students

  • Urooj Ghani PhD student Quantum software engineering
  • Bushra Alowaiid PhD student Distributed quantum computing
  • Muhammad Usama Riaz PhD student Quantum Machine Learning
  • Ibraheem AlYousef MS student Quantum simulations
  • Abdulrahman Aldubaikel MS student Quantum Combintorial Optimization
  • Badr Al-Ulaiw MS student QC Algorithm for Geophysical Problems
  • Mutab Alqadda MS student Graph Optimization using Neutral Atom
  • Feras Alsayegh MS student Distributed Quantum Computing
  • Arwa Alsuwailem MS student Hybrid Quantum-classical workflow Orchestration
  • Hashim Habtur MS student Quatnum Compilation using LLMs
  • Bandar Hanbash MS student Quantum-HPC Inegration

Collaborators

  • Aramco
  • HUMAIN
  • Bull
  • Pasqal
  • Brightskies

Join us

We welcome PhD and MS applicants, postdoctoral researchers, and collaborators from universities and industry. Send a short note and we will reply by email.

Dr. Muhamad Felemban Project lead, SQS² mfelemban -at- kfupm.edu.sa King Fahd University of Petroleum & Minerals
Dhahran