Skip to content

Release 0.40.0

Latest
Latest
Compare
Choose a tag to compare
@multiphaseCFD multiphaseCFD released this 14 Jan 15:19
v0.40.0
f634c1e

New features since last release

  • Add Exact Tensor Network (TN) C++ backend to Lightning-Tensor. (#977)

  • Add native N-controlled generators and adjoint support to lightning.gpu's single-GPU backend. (#970)

  • Add a Catalyst-specific wrapping class for Lightning-Qubit. (#960) (#999)

  • Add native N-controlled gate/matrix operations and adjoint support to lightning.kokkos. (#950)

  • Add native N-controlled gates support to lightning.gpu's single-GPU backend. (#938)

Breaking changes

  • Remove all instances of the legacy operator arithmetic (the Hamiltonian and Tensor deprecated classes in PennyLane). (#994) (#997)

  • Remove all instances of qml.QubitStateVector as deprecated in PennyLane. (#985)

Improvements

  • Add CI wheels checks for aarch64 wheels of Lightning-GPU and Lightning-Tensor. (#1031)

  • Replace the dummy_tensor_update method with the cutensornetStateCaptureMPS API to ensure that applying gates is allowed after the cutensornetStateCompute call. (#1028)

  • Add unit tests for measurement with shots for Lightning-Tensor (method="tn"). (#1027)

  • Add CUDA dependencies to Lightning-GPU and Lightning-Tensor Python wheels. (#1025)

  • Update the python layer UI of Lightning-Tensor. (#1022)

  • Catalyst device interfaces support dynamic shots, and no longer parses the device init op's attribute dictionary for a static shots literal. (#1017)

  • Update the logic for enabling grad_on_execution during device execution. (#1016)

  • Reduce flaky test and increase test shots count. (#1015)

  • Add Python bindings for Lightning-Tensor (Exact Tensor Network). (#1014)

  • Reverse Lightning-Qubit generators vector insertion order. (#1009)

  • Update Kokkos version support to 4.5. (#1007)

  • Enable N-controlled gate and matrix support to lightning.gpu simulator for Catalyst. (#1005)

  • Generalize seeding mechanism for all measurements. (#1003)

  • lightning.qubit, lightning.gpu, and lightning.kokkos now define
    the eval_jaxpr method for integration with the experimental
    capture project. (#1002)

  • Update Kokkos version support to 4.4.1 and enable Lightning-Kokkos[CUDA] C++ tests on CI. (#1000)

  • Add C++ unit tests for Lightning-Tensor (Exact Tensor Network). (#998)

  • Add native BLAS support to the C++ layer via dynamic scipy-openblas32 loading. (#995)

  • Update installation instruction for Lightning-GPU-MPI to explicitly ask users to add path/to/libmpi.so to the LD_LIBRARY_PATH. Update the runtime error message to ensure users know how to fix. (#993)

  • Update the Lightning TOML files for the devices to use the new schema for declaring device capabilities. (#988)

  • Optimize lightning.tensor by adding direct MPS sites data set with qml.MPSPrep. (#983)

  • Unify excitation gates memory layout to row-major for both Lightning-GPU and Lightning-Tensor. (#959)

  • Update the Lightning-Kokkos CUDA backend for compatibility with Catalyst. (#942)

Documentation

  • Update and improve README.rst and documentations. (#1035)

  • Add the exact tensor network to the README.rst and update link to HIP. (#1034)

  • Add the exact tensor network to the Lightning-Tensor docs. (#1021)

  • Update Lightning-Tensor installation docs and usage suggestions. (#979)

Bug fixes

  • Fix Python CUDA dependencies by adding path to nvidia/nvjitlink/lib to RPATH. (#1031)

  • Add RTLD_NODELETE flag to dlopen in order to mitigate the segfault issues for arm64-macos Catalyst support. (#1030)

  • Set RPATH with @loader_path instead of $ORIGIN for macOS. (#1029)

  • Update CUDA version to 12.4 for GPU tests on CI. (#1023)

  • Pin jax[cpu]==0.4.28 for compatibility with PennyLane and Catalyst. (#1019)

  • Fix Lightning-Kokkos [[maybe_unused]] and exp2 errors with hipcc. (#1018)

  • Fix Lightning-Kokkos editable mode path. (#1010)

  • Fix the missing ninja in Linux AARCH64 wheels recipe. (#1007)

  • Fix version switch condition the GPU workflow tests for LGPU and LKokkos. (#1006)

  • Fix issue with lightning.gpu Rot operation with adjoint. (#1004)

  • Fix the Lightning-Kokkos[CUDA] C++ API cudaFree segfaults in applyMatrix. (#1000)

  • Fix issue with adjoint-jacobian of adjoint ops. (#996)

  • Fix the upload-pypi token issues for Linux and MacOS (x86_64 & AARCH64) wheels. (#989)

  • Fix Pennylane dependency branch (v0.39_rc0 to master). (#984)

  • Fix PTM stable latest. Removing FIXME patch for v0.39. (#982)

Contributors

This release contains contributions from (in alphabetical order):

Ali Asadi, Astral Cai, Amintor Dusko, Christina Lee, Joseph Lee, Anton Naim Ibrahim, Luis Alfredo Nuñez Meneses, Andrija Paurevic, Shuli Shu, Raul Torres, Haochen Paul Wang

Assets 46
Loading