RayD Dr.Jit

PyPI License

RayD is a Dr.Jit-native GPU library for differentiable ray geometry and multipath simulation primitives with OptiX and pure-CUDA ray-tracing backends.

pip install rayd-drjit

RayD is not a full renderer. It exposes low-level scene, ray, edge, visibility, and reflection-path queries for building custom renderers, RF simulators, acoustic tools, and inverse-design systems without adopting a material-light-integrator framework.

Release compatibility

Release wheels cover Linux x86-64 and Windows x86-64 on CPython 3.10-3.14. They are built with CUDA 12.8 and contain native code for sm_70, sm_75, sm_80, sm_86, sm_87, sm_89, sm_90, sm_100, sm_101, and sm_120, plus compute_120 PTX for forward compatibility. sm_87 is present for Orin-compatible builds, but the published x86-64 wheels themselves are not aarch64 packages; Jetson currently requires a native/source build.

Ray-tracing backend selection

Scene() uses OptiX when the current CUDA device/context accepts an OptiX context and otherwise selects the pure-CUDA triangle backend plus the Dr.Jit/CUDA edge BVH. Use trace_backend="optix" or "cuda" and edge_bvh_backend="optix" or "drjit" to require a backend. scene.capabilities() reports the resolved choice. An explicit OptiX request fails when unavailable, and operational pipeline, allocation, or CUDA failures are not converted into a fallback. RAYD_DISABLE_OPTIX=1 forces capability discovery to report OptiX unavailable.

RayD is Dr.Jit-native and does not depend on PyTorch. Because its nanobind extension uses the CPython ABI, CI builds one wheel per Python version rather than one Python-independent wheel. The complete build matrix and release configuration are documented in ci_build_matrix.md.

Scope

RayD focuses on geometry and wave-propagation primitives:

  • differentiable ray-mesh intersection
  • scene-level GPU acceleration through OptiX or a pure-CUDA BVH
  • nearest-edge queries through a scene-global edge BVH
  • primary-edge sampling support for edge-based gradient terms
  • segment visibility and multipath reflection primitives
  • Dr.Jit arrays and Dr.Jit autodiff as the public Python API

RayD intentionally does not provide:

  • a material or BSDF system
  • emitters
  • integrators
  • scene loading
  • image I/O
  • a complete path-tracing runtime
  • alternate Python frontend wrappers

Why RayD?

RayD is for users who need fast differentiable geometry queries, but do not want a full rendering framework.

Mitsuba is excellent for physically based rendering, but it can be too high-level when the main workload is RF propagation, acoustics, sonar, visibility analysis, or custom wave simulation. In those settings, direct control over ray-scene queries, edge queries, reflection paths, and geometry gradients is often more useful than a complete renderer.

RayD keeps the API surface small: meshes, scenes, rays, intersections, edges, and multipath query results.

Core API

  • Mesh: triangle geometry, transforms, UVs, and edge topology
  • Scene: a container of meshes plus OptiX acceleration structures
  • scene.intersect(ray): differentiable ray-mesh intersection
  • scene.shadow_test(ray): occlusion testing
  • scene.nearest_edge(query): nearest-edge queries for points and rays
  • scene.set_edge_mask(mask) / scene.edge_mask(): scene-global filtering for secondary-edge queries
  • scene.trace_reflections(...): specular reflection-path tracing
  • scene.visible(...): batched segment visibility queries
  • scene.trace_refl_epc(...): equivalent-path correction primitives for reflection paths
  • scene.accum_dfr_direct(...) / scene.accum_dfr(...): native diffraction grid accumulation
  • scene.trace_dfr_paths(...): compact native first-order diffraction path export

Feature Overview

Each core query, what it computes, its input/output, and how it behaves under Dr.Jit autodiff (AD):

API What it computes Input -> Output AD
scene.intersect(ray) Closest-hit ray-mesh intersection rays -> Intersection (t, point, normal, uv, ids) AD
scene.shadow_test(ray) Any-hit occlusion test rays -> boolean mask Boolean
scene.nearest_edge(point) Nearest scene edge to each point points -> NearestPointEdge (distance, points, edge id) AD
scene.nearest_edge(ray) Nearest scene edge to each ray (segment on [0, tmax] when tmax is finite) rays -> NearestRayEdge AD
scene.nearest_edges(point, k) k nearest scene edges per point (k <= 16) points -> NearestEdgesTopK AD
scene.visible(start, end) Mutual visibility between two segment endpoints endpoints -> SegmentVisibility Boolean
scene.visible_pair / visible_chain / visible_edge Shared-origin pair, polyline-chain, and edge-sample visibility segments -> segment/chain/edge visibility Boolean
scene.trace_reflections(ray, max_bounces) Specular reflection paths with image sources rays -> ReflectionChain (hit points, normals, image sources, ids) AD
scene.trace_refl_epc(ray, receiver, ...) Equivalent-path-correction reflection toward a receiver rays + receiver -> ReflEpc Detached
scene.trace_refl_epc_field(tx, receiver, ...) EPC trace returning the complex reflected field tx position + receiver + ReflEpcFieldOptions(AD) -> ReflEpcField (complex E-field) Native AD
scene.accumulate_reflections(...) Accumulate reflected field/power onto a grid rays + grid + material -> AccumResult Native AD
scene.accum_dfr_direct(...) Native direct/Keller/suffix diffraction accumulation onto a grid DfrStates + DfrGrid + DfrMaterial -> DfrAccum Native AD for direct, Keller, and suffix-reflection order-1, including suffix reflector mesh-vertex gradients; detached for unsupported AD strategies
scene.accum_dfr(...) Native order-2/3 diffraction-chain accumulation onto a grid initial/recursive DfrStates + grid + material -> DfrAccum Native AD for direct, Keller, and suffix-reflection order-2/3 chains, including suffix reflector mesh-vertex gradients
scene.trace_dfr_paths(...) Compact first-order diffraction path export tx/rx + DfrStates + DfrMaterial -> DfrPaths AD

AD legend:

  • AD - differentiable geometry: geometric outputs carry Dr.Jit gradients with respect to mesh vertices and transforms; the discrete hit/edge/path selection runs detached.
  • Native AD - the native RayD multipath implementation preserves Dr.Jit gradients for continuous inputs passed as AD arrays and supplies RayD-side derivative code. Diffraction grid accumulation uses CUDA custom-op JVP/VJP over a fixed OptiX forward tape for direct, Keller, and suffix-reflection order-1 and order-2/3 chain accumulation. Discrete visibility/path-selection decisions remain detached.
  • Boolean - returns occlusion/visibility booleans and is not differentiable.
  • Detached - native fast path: runs detached only and rejects AD inputs.

Trace vs. accum:

  • trace_* APIs enumerate geometric or field paths and return per-ray/per-path records. They do not reduce contributions into receiver grids.
  • accum_* APIs launch native accumulation kernels that reduce contributions into aggregate outputs, usually grid cells, with atomic or tiled writes. They are the high-throughput path for radiomap-style workloads when individual path records are not needed.

Minimal Example

The example below traces one ray against one triangle and backpropagates the hit distance to the vertex positions.

import drjit as dr
import rayd.drjit as rd


mesh = rd.Mesh(
    dr.cuda.Array3f([0.0, 1.0, 0.0],
                    [0.0, 0.0, 1.0],
                    [0.0, 0.0, 0.0]),
    dr.cuda.Array3i([0], [1], [2]),
)

verts = dr.cuda.ad.Array3f(
    [0.0, 1.0, 0.0],
    [0.0, 0.0, 1.0],
    [0.0, 0.0, 0.0],
)
dr.enable_grad(verts)
mesh.vertex_positions = verts

scene = rd.Scene()
scene.add_mesh(mesh)
scene.build()

ray = rd.RayAD(
    dr.cuda.ad.Array3f([0.25], [0.25], [-1.0]),
    dr.cuda.ad.Array3f([0.0], [0.0], [1.0]),
)

its = scene.intersect(ray)
loss = dr.sum(its.t)
dr.backward(loss)

print("t =", its.t)
print("grad z =", dr.grad(verts)[2])

Edge Queries

RayD provides a scene-level edge acceleration structure for nearest-edge and edge-sampling workloads.

This is useful for:

  • edge sampling
  • nearest-edge queries
  • visibility-boundary terms
  • geometric diffraction models

Scene.set_edge_mask(mask) filters the secondary-edge BVH in scene-global edge index space. It does not modify scene.edge_info(), scene.edge_topology(), or scene.mesh_edge_offsets().

Multipath Queries

RayD includes low-level reflection and visibility primitives for custom wave simulators:

  • trace_reflections(...) for specular reflection chains
  • visible(...), visible_pair(...), visible_chain(...), and visible_edge(...) for segment and polyline-chain visibility
  • trace_refl_epc(...) and trace_refl_epc_field(...) for equivalent-path correction workflows
  • accumulate_reflections(...) for reflection grid accumulation workloads
  • accum_dfr_direct(...), accum_dfr(...), trace_dfr_paths(...) for native diffraction kernels

These APIs expose primitives, not a complete propagation simulator. Callers own the source model, receiver model, material policy, objective, and optimization loop.

Naming rule: use Dfr for diffraction (DfrStates, DfrAccum, accum_dfr_direct), Refl for reflection-specific short names, keep Epc in equivalent-path-correction APIs, and reserve AD for automatic differentiation. See api_naming_standard.md.

Examples

Performance

The chart below was generated on March 25, 2026 on an NVIDIA GeForce RTX 5080 and AMD Ryzen 7 9800X3D, comparing RayD (0.1.2) against Mitsuba 3.8.0 with the cuda_ad_rgb variant.

  • RayD is consistently faster on static forward and static gradient workloads across all three scene sizes.
  • Dynamic reduced forward reaches parity or better from the medium scene onward, and dynamic full is effectively tied on the largest case.
  • On the largest 192x192 mesh / 384x384 ray benchmark, RayD vs Mitsuba average latency in milliseconds is: static full 0.162 vs 0.190, static reduced 0.124 vs 0.224, dynamic full 0.741 vs 0.740, dynamic reduced 0.689 vs 0.714, gradient static 0.411 vs 0.757, gradient dynamic 1.324 vs 1.413.
  • Correctness stayed aligned throughout the sweep: forward mismatch counts remained 0, and the largest static gradient discrepancy was 9.54e-7.

Device Selection

RayD follows Dr.Jit's current-thread CUDA device selection. Choose a GPU before constructing RayD resources:

import rayd.drjit as rd

rd.set_device(0)

Existing RayD scenes, OptiX pipelines, and BVHs should not be reused across device switches in the same process.

Building Locally

RayD is a Python package with a C++/CUDA extension.

You need Python >=3.10,<3.15, CUDA Toolkit >=11.0, CMake, a C++17 compiler, drjit==1.3.1, nanobind==2.11.0, and scikit-build-core. The nanobind pin must match the version embedded by the pinned Dr.Jit release so both extensions share one nanobind registry ABI.

On Windows, use Visual Studio 2022 with Desktop C++ tools. On Linux, use GCC or Clang with C++17 support.

Recommended environment:

conda create -n myenv python=3.10 -y
conda activate myenv
python -m pip install -U pip setuptools wheel
python -m pip install cmake scikit-build-core nanobind==2.11.0
python -m pip install drjit==1.3.1

Install from the repository root:

python -m pip install ./drjit

For editable development builds:

python -m pip install --no-build-isolation -ve ./drjit

Repository Layout

Testing

python -m unittest tests.scene.test_geometry_jit -v
python -m unittest tests.visibility.test_visibility_topk_jit -v
python -m unittest tests.reflection.test_epc_jit -v
python -m unittest tests.reflection.test_accumulation_jit -v
python -m unittest tests.packaging.test_project_metadata -v

The default development environment used by this repository is:

conda activate witwin3

Credits

RayD is developed with reference to:

Citation

@inproceedings{chen2026rfdt,
  title     = {Physically Accurate Differentiable Inverse Rendering
               for Radio Frequency Digital Twin},
  author    = {Chen, Xingyu and Zhang, Xinyu and Zheng, Kai and
               Fang, Xinmin and Li, Tzu-Mao and Lu, Chris Xiaoxuan
               and Li, Zhengxiong},
  booktitle = {Proceedings of the 32nd Annual International Conference
               on Mobile Computing and Networking (MobiCom)},
  year      = {2026},
  doi       = {10.1145/3795866.3796686},
  publisher = {ACM},
  address   = {Austin, TX, USA},
}

License

BSD 3-Clause. See LICENSE.

Key dates

PyPI data
Data sourced directly from PyPI's database.
  • Released:
Latest release

1 maintainer

PyPI data
Data sourced directly from PyPI's database.
Avatar for Asixa from gravatar.com Asixa

Credits

Author: Xingyu Chen

License

BSD-3-Clause

Requires

Python <3.15, >=3.10

Report project as malware

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distributions

If you're not sure about the file name format, learn more about wheel file names.

rayd_drjit-0.8.0-cp314-cp314-win_amd64.whl (26.0 MB view details)

Uploaded CPython 3.14Windows x86-64

rayd_drjit-0.8.0-cp314-cp314-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (24.4 MB view details)

Uploaded CPython 3.14manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

rayd_drjit-0.8.0-cp313-cp313-win_amd64.whl (26.1 MB view details)

Uploaded CPython 3.13Windows x86-64

rayd_drjit-0.8.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (24.4 MB view details)

Uploaded CPython 3.13manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

rayd_drjit-0.8.0-cp312-cp312-win_amd64.whl (26.1 MB view details)

Uploaded CPython 3.12Windows x86-64

rayd_drjit-0.8.0-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (24.4 MB view details)

Uploaded CPython 3.12manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

rayd_drjit-0.8.0-cp311-cp311-win_amd64.whl (26.1 MB view details)

Uploaded CPython 3.11Windows x86-64

rayd_drjit-0.8.0-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (24.4 MB view details)

Uploaded CPython 3.11manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

rayd_drjit-0.8.0-cp310-cp310-win_amd64.whl (26.1 MB view details)

Uploaded CPython 3.10Windows x86-64

rayd_drjit-0.8.0-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl (24.4 MB view details)

Uploaded CPython 3.10manylinux: glibc 2.27+ x86-64manylinux: glibc 2.28+ x86-64

File details

Details for the file rayd_drjit-0.8.0-cp314-cp314-win_amd64.whl.

File metadata

  • Download URL: rayd_drjit-0.8.0-cp314-cp314-win_amd64.whl
  • Upload date:
  • Size: 26.0 MB
  • Tags: CPython 3.14, Windows x86-64
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for rayd_drjit-0.8.0-cp314-cp314-win_amd64.whl
Algorithm Hash digest
SHA256 a670a9c59dc9b8856017ba03309ac9f2009a60791c1e534bba733fffe92cbe8d
MD5 ecd769b170d7e014994bc89213c6c3f0
BLAKE2b-256 a8a45abb6c392a53e4aa74a12ab066c0978c928df73a30abe6cbee721ab500ba

See more details on using hashes here.

Provenance

The following attestation bundles were made for rayd_drjit-0.8.0-cp314-cp314-win_amd64.whl:

Publisher: pypi.yml on Asixa/RayD

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file rayd_drjit-0.8.0-cp314-cp314-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

File hashes

Hashes for rayd_drjit-0.8.0-cp314-cp314-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 b17043cf160cd17209016ed7ac69dbaa0903ec286d0bcdc0f14cacd47e6976ec
MD5 d7ebe8cb481b9532975ee8672fdd0068
BLAKE2b-256 186fe4215a30a1c41b3b8504da13115199493514a655ff2a5f4cec2e5299aaee

See more details on using hashes here.

Provenance

The following attestation bundles were made for rayd_drjit-0.8.0-cp314-cp314-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: pypi.yml on Asixa/RayD

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file rayd_drjit-0.8.0-cp313-cp313-win_amd64.whl.

File metadata

  • Download URL: rayd_drjit-0.8.0-cp313-cp313-win_amd64.whl
  • Upload date:
  • Size: 26.1 MB
  • Tags: CPython 3.13, Windows x86-64
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for rayd_drjit-0.8.0-cp313-cp313-win_amd64.whl
Algorithm Hash digest
SHA256 ed87f762fa58a39919857e5847aef938d8834bddc672046cc788044a3a1deb8a
MD5 923110c9197ca7626df2e2f1700b6adc
BLAKE2b-256 727bc00ede9e7a8b0fed9efb637b8f7fed83e4763d445918f6a1b525c37407fd

See more details on using hashes here.

Provenance

The following attestation bundles were made for rayd_drjit-0.8.0-cp313-cp313-win_amd64.whl:

Publisher: pypi.yml on Asixa/RayD

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file rayd_drjit-0.8.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

File hashes

Hashes for rayd_drjit-0.8.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 c2b75da5fc20d53a129d9142c32e8f01875848a8c4ddf3f3de10a5b4d859b252
MD5 1486fd44e5c271aaab90412ee33b6f44
BLAKE2b-256 5f843fc216db584f21679b22fa1d82b9684d7c82823f7e1432700840e432b60b

See more details on using hashes here.

Provenance

The following attestation bundles were made for rayd_drjit-0.8.0-cp313-cp313-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: pypi.yml on Asixa/RayD

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file rayd_drjit-0.8.0-cp312-cp312-win_amd64.whl.

File metadata

  • Download URL: rayd_drjit-0.8.0-cp312-cp312-win_amd64.whl
  • Upload date:
  • Size: 26.1 MB
  • Tags: CPython 3.12, Windows x86-64
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for rayd_drjit-0.8.0-cp312-cp312-win_amd64.whl
Algorithm Hash digest
SHA256 c9f9463847bf60c961bff58d1e6c281e2cbdb273c9e96c275cd8957e62105570
MD5 b15b1b4bd61ed5cdfe681f30ae8e9eab
BLAKE2b-256 6f03aa2a2baa4324a4b545c0c859b551ae72109798edef397a215b6935aad998

See more details on using hashes here.

Provenance

The following attestation bundles were made for rayd_drjit-0.8.0-cp312-cp312-win_amd64.whl:

Publisher: pypi.yml on Asixa/RayD

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file rayd_drjit-0.8.0-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

File hashes

Hashes for rayd_drjit-0.8.0-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 68d542cf2142405579d5b784979b59256a2fc490b0b066c393fd0504ea8cc152
MD5 412b05938d1041a7f15681e771b06f62
BLAKE2b-256 ebbb92fb4d3c5279de7de171275b1301fb560a112607c6a5ee57d77bf537002c

See more details on using hashes here.

Provenance

The following attestation bundles were made for rayd_drjit-0.8.0-cp312-cp312-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: pypi.yml on Asixa/RayD

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file rayd_drjit-0.8.0-cp311-cp311-win_amd64.whl.

File metadata

  • Download URL: rayd_drjit-0.8.0-cp311-cp311-win_amd64.whl
  • Upload date:
  • Size: 26.1 MB
  • Tags: CPython 3.11, Windows x86-64
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for rayd_drjit-0.8.0-cp311-cp311-win_amd64.whl
Algorithm Hash digest
SHA256 2bbfec889356a6cafce4a320de00a21ae0f723ee7fdfddda7e341c917ac19a74
MD5 721f362cb6d0cc614f721cdf06a4eecf
BLAKE2b-256 6efdc0c4b921256fe6e9a29d6c972da4413910d9d05b2d8f1c5299591dddcac4

See more details on using hashes here.

Provenance

The following attestation bundles were made for rayd_drjit-0.8.0-cp311-cp311-win_amd64.whl:

Publisher: pypi.yml on Asixa/RayD

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file rayd_drjit-0.8.0-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

File hashes

Hashes for rayd_drjit-0.8.0-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 0b8ca3d2605459e90b7364fd0b8cb17d897a7f740f660d46c4e38b656c689de4
MD5 fe02e9894e9c8bc8f99030667ded3370
BLAKE2b-256 2d6e8de0819e6e3b4259c423c07c539de220c46853a07e05095bae79a7389d96

See more details on using hashes here.

Provenance

The following attestation bundles were made for rayd_drjit-0.8.0-cp311-cp311-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: pypi.yml on Asixa/RayD

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file rayd_drjit-0.8.0-cp310-cp310-win_amd64.whl.

File metadata

  • Download URL: rayd_drjit-0.8.0-cp310-cp310-win_amd64.whl
  • Upload date:
  • Size: 26.1 MB
  • Tags: CPython 3.10, Windows x86-64
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/7.0.0 CPython/3.13.14

File hashes

Hashes for rayd_drjit-0.8.0-cp310-cp310-win_amd64.whl
Algorithm Hash digest
SHA256 fb89bf86adb7dd84a86b11e2ec30e7ebf2065928fd91bd855503f88f15d1252d
MD5 354228017745833283b8f4530aa1d341
BLAKE2b-256 6018f3edc2ab741fde0d58bf01e0858dbeb25c521eace5b767fed2f884d50f01

See more details on using hashes here.

Provenance

The following attestation bundles were made for rayd_drjit-0.8.0-cp310-cp310-win_amd64.whl:

Publisher: pypi.yml on Asixa/RayD

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file rayd_drjit-0.8.0-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl.

File metadata

File hashes

Hashes for rayd_drjit-0.8.0-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 dac45daa04af996a9a9a1e9958b4fc9c0176abc66dfecc2cbb1b487317cee937
MD5 6569eade4dfe8a108f0331b188fec069
BLAKE2b-256 52f4149cbe7b934f17e5ef713ec62bf430c0ebeba1df62cc2f3d8722d1d0f40a

See more details on using hashes here.

Provenance

The following attestation bundles were made for rayd_drjit-0.8.0-cp310-cp310-manylinux_2_27_x86_64.manylinux_2_28_x86_64.whl:

Publisher: pypi.yml on Asixa/RayD

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

Release history Release notifications | RSS feed

This release

0.8.0 This release

10 files

0.7.0

10 files

0.6.0

10 files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page