v0.7.0

ImplicitBVH v0.7.0

Diff since v0.6.0

• Added new Leaf-vs-Tree contact and raytracing traversal algorithm which is more scalable
• Added alg as a new argument to traverse
• Made options, start_level, and cache keyword arguments
• Made Morton encoding algorithm-specific (for future e.g. extended Morton codes) and moved to BVHOptions
• Added optional narrow-phase checking function before considering a contact
• Bundled bounding volumes, their user-defined index and Morton code into a single BoundingVolume struct which allows in-place sorting; this makes BVH construction between frames faster when most bounding volumes are still roughly sorted

Breaking changes

• Old interfaces are kept for compatibility, but defaults are different: BBox{Float32} nodes, UInt32 Morton codes, Int32 indices.
• BVHTraversal can have different cache types depending on the algorithm; previous code which was not making use of internals should still work.
• Removed order from BVH as now vectors of BoundingVolume are sorted in-place

Merged pull requests:

• More scalable Leaf-vs-Tree traversal (#19) (@anicusan)

v0.6.0

ImplicitBVH v0.6.0

Diff since v0.5.4

Using AcceleratedKernels-0.4, all steps in the multithreaded CPU backend are parallelised, including the previous Morton sorting bottleneck

Breaking Changes

Removed the scheduler keyword following its removal from AcceleratedKernels-0.4. To update, simply remove that argument - everything else stays the same.

Merged pull requests:

• Parallelised all steps for CPUs, including contacts repacking and extrema finding. Removed scheduler; use AK-0.4 with parallel sorter. Updated benchmarks. (#17) (@anicusan)

Closed issues:

• Storage of BVH leaves (#15)

v0.5.4

ImplicitBVH v0.5.4

Diff since v0.5.3

Merged pull requests:

• Added unsafe_indices and uniform following KA=0.9.34 changes (#16) (@anicusan)

v0.5.3

ImplicitBVH v0.5.3

Diff since v0.5.2

Merged pull requests:

• fix to error in raytracing intersection calculation for points inside… (#14) (@Jack-Grogan)

v0.5.2

ImplicitBVH v0.5.2

Diff since v0.5.1

v0.5.1

ImplicitBVH v0.5.1

Diff since v0.5.0

Merged pull requests:

• Add Atomix.jl v1 compat (#11) (@christiangnrd)

Closed issues:

• BVH building is not compatible with CUDA.jl (#10)

v0.5.0

ImplicitBVH v0.5.0

Diff since v0.4.1

• GPU acceleration using AcceleratedKernels.jl
• Ray-tracing!

Internal

• Coherent index types used in kernels
• Exposing cache from BVH builder for memory reuse

Merged pull requests:

• GPU Parallelisation with AcceleratedKernels.jl (#2) (@anicusan)
• Implementation of raytracing in ImplicitBHV.jl (#4) (@Jack-Grogan)
• switch to using AK TaskPartitioner. Added logo (#5) (@anicusan)
• Update raytrace_cpu.jl (#7) (@Jack-Grogan)
• Update raytrace_cpu.jl (#8) (@Jack-Grogan)
• Fix to error in raytrace_cpu.jl cpu parallelisation script (#9) (@Jack-Grogan)

Closed issues:

• Ray tracing issue (#6)

v0.4.1

ImplicitBVH v0.4.1

Diff since v0.4.0

v0.4.0

ImplicitBVH v0.4.0

Diff since v0.3.0

v0.3.0

ImplicitBVH v0.3.0

Diff since v0.2.0