Recognition: unknown
qHiPSTER: The Quantum High Performance Software Testing Environment
read the original abstract
We present qHiPSTER, the Quantum High Performance Software Testing Environment. qHiPSTER is a distributed high-performance implementation of a quantum simulator on a classical computer, that can simulate general single-qubit gates and two-qubit controlled gates. We perform a number of single- and multi-node optimizations, including vectorization, multi-threading, cache blocking, as well as overlapping computation with communication. Using the TACC Stampede supercomputer, we simulate quantum circuits ("quantum software") of up to 40 qubits. We carry out a detailed performance analysis to show that our simulator achieves both high performance and high hardware efficiency, limited only by the sustainable memory and network bandwidth of the machine.
This paper has not been read by Pith yet.
Forward citations
Cited by 6 Pith papers
-
A Controlled Study of Memory Hierarchy Transitions in Quantum Circuit Simulation on Apple M4 Pro Unified Memory Architecture
Quantum circuit simulations on Apple M4 Pro unified memory exhibit a reproducible 4.46x slowdown at 29 qubits and GPU speedups of 3-10x that exceed STREAM bandwidth predictions, with larger gaps for irregular access patterns.
-
A Controlled Study of Memory Hierarchy Transitions in Quantum Circuit Simulation on Apple M4 Pro Unified Memory Architecture
Quantum circuit simulations on Apple M4 Pro show a reproducible 4.46x timing discontinuity at 29 qubits and access-pattern-dependent speedups (3.1-10x) that exceed peak bandwidth predictions.
-
PennyLane: Automatic differentiation of hybrid quantum-classical computations
PennyLane is a software library extending automatic differentiation to hybrid quantum-classical systems for variational quantum algorithms.
-
Extending UNIQuE: Quantum Simulation Speedup for the HHL Algorithm
Classical emulation of the HHL algorithm via extended UNIQuE scales exponentially only with qubit count and shows runtime advantage over state-vector simulation for small linear systems.
-
Large-Scale Quantum Circuit Simulation on HPC Cluster via Cache Blocking, Boosting, and Gate Fusion Optimization
New merge booster and diagonal detector components, combined with cache blocking and gate fusion, deliver up to 160x speedup on circuit benchmarks and 34x on diagonal-heavy gates versus prior simulators.
-
Accelerating Quantum State Encoding with SIMD: Design, Implementation, and Benchmarking
Hybriqu Encoder delivers 5.4% faster pure angle encoding at 64 qubits on Apple Silicon by using AVX SIMD and cache-friendly precalculations, with gains increasing beyond L1 cache size while full-state updates remain m...
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.