0.10.0

In this release we have added a range of tools for simulating noisy quantum systems. It includes support for trajectory based noise simulation for our GPU-accelerated statevector and tensor network simulators, and a new sampling option for better stim performance. We have also expanded the range of noise models, and added an apply_noise "gate". This release also adds the C++ support for the CUDA-Q dynamics backend (a master equation solver), as well as operator classes for fermionic, bosonic and custom operators. This release also includes support for submitting to Pasqal backends, and adds C++ support for QuEra. Check out our docs, including new tutorials and examples, for more information.

Note: Support for CUDA 11 will be removed in future releases. Please update to CUDA 12.

What's Changed

Features and Enhancements 🎉

• Support noise trajectory simulation for the statevector targets by @1tnguyen in #2466
• Support noise trajectory simulation for tensornet targets by @1tnguyen in #2520
• Implement new noise models by @bmhowe23 in #2653
• Implementation of the cudaq::apply_noise feature by @schweitzpgi in #2635
• New cudaq::sample option: explicit measurements by @bmhowe23 in #2567
• Enable multi-GPU for dynamics target in Python by @1tnguyen in #2509
• Operator classes by @bettinaheim in #2631
• Dynamics cpp integration by @sacpis in #2683
• Support Rydberg Hamiltonian in C++ Evolve API by @nvidia-dobri in #2704
• Add support for IonQ qpu.forte-enterprise-1 and qpu.forte-enterprise-2 quantum backends by @radumarg in #2559
• Add Pasqal target by @awennersteen in #2632

Bug Fixes 🐛

• Apply LLVM customizations for Python wheel build by @bmhowe23 in #2504
• Fix a bug in tensornet backend scratch pad allocation in multi-GPU mode by @1tnguyen in #2516
• Fix issues with python range boundaries by @annagrin in #2571
• Fix issues in lift-array-alloc by @annagrin in #2570
• Fixing various Python bugs by @boschmitt in #2592
• Accept extensions other than .cpp in nvq++ by @bebora in #2514
• Bug fix for Python bridge by @amccaskey in #2615
• Fix a couple of bugs with loop normalization and unrolling by @schweitzpgi in #2618
• Fix an issue with the tensornet initialization for the remote-mqpu target by @1tnguyen in #2565
• Add Stim support for sdg gate (required for my measurements) by @bmhowe23 in #2624
• Fixes for Python dynamics handling distributed state by @1tnguyen in #2629
• Allow 'None' as a type annotation for cuda-q kernels by @cuMikeBrown in #2630
• Fix #2525: Make exp_pauli a fully functional operator. by @schweitzpgi in #2642
• Handle missing cases in ObserveAnsatz by @annagrin in #2661
• Fix decomposition patterns to handle spans of pauli words. by @schweitzpgi in #2664
• Fixes for NoiseModel when using with fp32 targets by @1tnguyen in #2672
• Fix a bug in ObserveAnsatz when handling spin.y term by @annagrin in #2676
• Fix cudaq.draw for circuits with exp_pauli by @bmhowe23 in #2678
• Fix cudaq.exp_pauli usage of pauli_word for quantum devices by @bmhowe23 in #2675
• Fixing an issue with exp_pauli in nested kernel by @schweitzpgi in #2697
• Fix bug with exp_pauli lowering. by @schweitzpgi in #2698
• Explicitly add the path for cuTensor library in the CUDAQ_DYNLIBS by @khalatepradnya in #2691
• Improve error handling for analog targets when local emulation is requested by @khalatepradnya in #2713

Breaking Changes 🛠

• C++ operator and matrix API changes by @bettinaheim in #2714

Documentation Updates ✏️

• Update Python install instructions by @bmhowe23 in #2519
• Improve photonics documentation by @Omar-ORCA in #2502
• Reorganized Backends Docs Page by @mawolf2023 in #2481
• New application notebooks for Hamiltonian simulation & Deutsch Jozsa by @zohimchandani in #2488
• Update Braket example and doc links by @yitchen-tim in #2633
• Update docs for nvidia target fuse settings by @1tnguyen in #2660
• New tutorial for adapt-qaoa algorithm by @marwafar in #2638
• CUDA-Q quantum transformer tutorial by @anthonysmaldone in #2614
• Add documentation for Pasqal integration by @awennersteen in #2695
• Docs update for Amazon Braket by @khalatepradnya in #2702
• Instructions for running on Blackwell by @mitchdz in #2740

Other Changes

• Add number of trajectories to observe_options struct by @1tnguyen in #2456
• Add ElementaryOperator.define override option by @ikkoham in #2419
• Support for float comparison in Python kernels by @sacpis in #2489
• Add a few state-related cc ops by @annagrin in #2354
• Add support for character literals to C++ kernels by @schweitzpgi in #2495
• Allow mx and my gates in OpenQasm2 translation by @annagrin in #2494
• Make controlled rank a configurable setting for tensornet backends by @1tnguyen in #2446
• Environment variable for tensornet observe calculation by @1tnguyen in #2522
• Add support for observe API for Amazon Braket by @khalatepradnya in #2539
• Normalize loops with negative upper bounds by @annagrin in #2573
• New implementation of QIR codegen. by @schweitzpgi in #2542
• Add WriteAfterWriteElimination pass by @annagrin in #2572
• Support reservation ARN for Amazon Braket by @yitchen-tim in #2553
• Make sure uccsd compiles for quantum devices by @annagrin in #2458

New Contributors

• @vtomole made their first contribution in #2423
• @caiyunh made their first contribution in #2442
• @bharat-thotakura made their first contribution in #2464
• @lmondada made their first contribution in #2140
• @yitchen-tim made their first contribution in #2553
• @cuMikeBrown made their first contribution in #2630
• @radumarg made their first contribution in #2559
• @awennersteen made their first contribution in #2632
• @anthonysmaldone made their first contribution in #2614

Full Changelog: 0.9.1...0.10.0
Release created by workflow 13913105388.