v1.7.0

[shi-eric]

Changelog

[1.7.0] - 2025-03-30

Added

• Support JAX foreign function interface (FFI) (docs, [REQ] Add support for JAX FFI #511).
• Support Python/SASS correlation in Nsight Compute reports by emitting #line directives in CUDA-C code. This setting is controlled by wp.config.line_directives and is True by default. (docs, Explore #line directives for Python/SASS mappings #437)
• Support vec4f grid construction in wp.Volume.allocate_by_tiles().
• Add 2D SVD wp.svd2() ([REQUEST] 2D SVD #436).
• Add wp.randu() for random uint32 generation.
• Add matrix construction functions wp.matrix_from_cols() and wp.matrix_from_rows() (Inconsistent matrix indexing inside/outside a kernel #278).
• Add wp.transform_from_matrix() to obtain a transform from a 4x4 matrix ([feature request] matrix to transform and back #211).
• Add wp.where() to select between two arguments conditionally using a more intuitive argument order (cond, value_if_true, value_if_false) (Add wp.where() with opposite convention to eventually replace wp.select() #469).
• Add wp.get_mempool_used_mem_current() and wp.get_mempool_used_mem_high() to query the respective current and high-water mark memory pool allocator usage (Expose functions to query CUDA memory pool usage #446 ).
• Add Stream.is_complete and Event.is_complete properties to query completion status ([REQUEST] Add functions to query event and stream status #435).
• Support timing events inside of CUDA graphs ([REQ] Support timing events in CUDA graphs #556).
• Add LTO cache to speed up compilation times for kernels using MathDx-based tile functions. Use wp.clear_lto_cache() to clear the LTO cache ([REQ] Improve tile compile times by caching LTO objects #507).
• Add example demonstrating gradient checkpointing for fluid optimization in warp/examples/optim/example_fluid_checkpoint.py.
• Add a hinge-angle-based bending force to wp.sim.VBDIntegrator.
• Add an example to show mesh sampling using a CDF (Add a Mesh Sampling Example #476).

Changed

• Breaking: Remove CUTLASS dependency and wp.matmul() functionality (including batched version). Users should use tile primitives for matrix multiplication operations instead.
• Deprecate constructing a matrix from vectors using wp.matrix().
• Deprecate wp.select() in favor of wp.where(). Users should update their code to use wp.where(cond, value_if_true, value_if_false) instead of wp.select(cond, value_if_false, value_if_true).
• wp.sim.Control no longer has a model attribute (Remove model attribute from the Control class #487).
• wp.sim.Control.reset() is deprecated and now only zeros-out the controls (previously restored controls to initial model state). Use wp.sim.Control.clear() instead.
• Vector/matrix/quaternion component assignment operations (e.g., v[0] = x) now compile and run faster in the backward pass. Note: For correct gradient computation, each component should only be assigned once.
• @wp.kernel has now an optional module argument that allows passing a wp.context.Module to the kernel, or, if set to "unique" let Warp create a new unique module just for this kernel. The default behavior to use the current module is unchanged.
• Default PTX architecture is now automatically determined by the devices present in the system, ensuring optimal compatibility and performance ([REQ] Use newer PTX version if the system supports it #537).
• Structs now have a trivial default constructor, allowing for wp.tile_reduce() on tiles with struct data types.
• Extend wp.tile_broadcast() to support broadcasting to 1D, 3D, and 4D shapes (in addition to existing 2D support).
• wp.fem.integrate() and wp.fem.interpolate() may now perform parallel evaluation of quadrature points within elements.
• wp.fem.interpolate() can now build Jacobian sparse matrices of interpolated functions with respect to a trial field.
• Multiple wp.sparse routines (bsr_set_from_triplets, bsr_assign, bsr_axpy, bsr_mm) now accept a masked flag to discard any non-zero not already present in the destination matrix.
• wp.sparse.bsr_assign() no longer requires source and destination block shapes to evenly divide each other.
• Extend wp.expect_near() to support all vectors and quaternions.
• Extend wp.quat_from_matrix() to support 4x4 matrices.
• Update the OgnClothSimulate node to use the VBD integrator ([REQ] Convert Kit's Cloth Simulation Sample to Use the VBD Integrator #512).
• Remove the globalScale parameter from the OgnClothSimulate node.

Fixed

• Fix an out-of-bounds access bug caused by an unbalanced BVH tree ([BUG] Bug in point queries of wp.Mesh and wp.BVH #536).
• Fix an error of incorrectly adding the offset to -1 elements in edge_indices when adding a ModelBuilder to another ([BUG] multiple cloth grids in scene #557).

---

This discussion was created from the release v1.7.0.