Documented capabilities

  1. 01Bloch-cell, sparse, dense, symmetry, and amplitude-class state representations
  2. 02Standard gates, arbitrary 2x2 and 4x4 operations, measurement, and noise trajectories
  3. 03Gate fusion, parallel register batches, bulk readout, and save/restore through QSC

QSA 0.1.6

Measured release workloads

Results apply to the named benchmark and representation. They are not a universal multiplier for arbitrary circuits.

01

1,000 independent Bell pairs

QSA 0.1.6 accelerated engine vs QSA 0.1.0

32.12x faster
02

20 dense CNOT gates / 65,536 amplitudes

Specialized dense kernel vs QSA 0.1.0

107.18x faster
03

50,000 Python gate calls in one native plan

Optimized plan vs individual Python calls

629.61x faster
04

Compressed 16-qubit Grover search

Exact compressed path vs dense exact path

160,665.80x faster

Release validation

Numerical, compatibility, packaging, and hostile-input checks

Open validation record ↗
  1. 0112 native release tests
  2. 0214,400 randomized NumPy comparison operations
  3. 03200 Grover searches and 41,500 amplitude checks
  4. 0420,000 symmetry operations and 39,272 amplitude checks
  5. 05AddressSanitizer and UndefinedBehaviorSanitizer builds
  6. 06Tagged-archive rebuild and installed-package checks