Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Learn about Cobble, a new programming language designed for quantum computational linear algebra, in this 22-minute conference presentation from PLanQC 2026. Discover how quantum algorithms for computational linear algebra can achieve exponential speedups for applications like simulation and regression, but face unique challenges in implementation since they cannot efficiently store matrices in memory like classical algorithms. Explore the complexities developers encounter when implementing matrix arithmetic expressions as quantum circuits, including navigating cost models where conventional linear algebra optimizations may be inapplicable or unprofitable. Understand how Cobble addresses these challenges by enabling developers to express and manipulate quantum matrix representations called block encodings using high-level notation that automatically compiles to correct quantum circuits. Examine the language's built-in analyses for estimating time and space usage factors, along with optimizations that reduce overhead and generate efficient circuits using advanced techniques such as the quantum singular value transformation. Review evaluation results on benchmark kernels for simulation, regression, search, and other applications, demonstrating speedups of 2.6×–25.4× that existing circuit optimizers cannot achieve.
