pyqpanda3 v0.5.1 lets QCloud connect with a single API key, exposes correction modes and crosstalk suppression as configurable options, and makes random-circuit generation follow the chip topology; failed jobs are now reported as result state instead of exceptions.
pyqpanda3 v0.5.0 gives every CPU-family simulator a precision switch (float32/float64), dispatches SIMD kernels at runtime, rewrites the LightSABRE layout/routing engine, and adds a native RESET — on the benchpress suite it is now faster than Qiskit 2.5.2 on 1,040 of 1,044 paired cases.
pyqpanda3 v0.4.1 threads fidelity data into transpilation decisions — fidelity-driven approximate decomposition, LightSABRE routing, recoverable transpilation metrics — and fills the Heisenberg-picture expectation simulator and the matrix-product-state simulator.
pyqpanda3 v0.4.0 connects QCloud instruction fields, instruction recovery, schedule visualization, fidelity-aware transpilation, and topology-aware state preparation into one hardware workflow.
pyqpanda3 v0.3.5 introduces the visualization module for Bloch sphere plots and circuit diagrams, adds timing_info() to QCloudResult for job execution profiling, and fixes critical bugs in measure(), expectation computation, and gate serialization.
pyqpanda3 v0.3.4 introduces a unified visualization interface for quantum states, Bloch spheres, and probability distributions.
We're launching the new pyqpanda3 documentation platform — a bilingual, dark-themed site built on VitePress with full-text search, Mermaid diagrams, API reference, blog, and release notes.
