CUDA: Add chunked processing for winding number algorithms #4195
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Grantim
reviewed
Feb 25, 2025
source/MRViewer/MRCudaAccessor.cpp
Outdated
Comment on lines
121
to
134
| size_t CudaAccessor::fromGridMemory( const Mesh& mesh, const Vector3i& ) | ||
| { | ||
| return fastWindingNumberMeshMemory( mesh ) + size_t( dims.x ) * dims.y * dims.z * sizeof( float ); | ||
| return fastWindingNumberMeshMemory( mesh ); | ||
| } | ||
|
|
||
| size_t CudaAccessor::fromVectorMemory( const Mesh& mesh, size_t inputSize ) | ||
| size_t CudaAccessor::fromVectorMemory( const Mesh& mesh, size_t ) | ||
| { | ||
| return fastWindingNumberMeshMemory( mesh ) + inputSize * ( sizeof( float ) + sizeof( Vector3f ) ); | ||
| return fastWindingNumberMeshMemory( mesh ); | ||
| } | ||
|
|
||
| size_t CudaAccessor::selfIntersectionsMemory( const Mesh& mesh ) | ||
| { | ||
| return fastWindingNumberMeshMemory( mesh ) + mesh.topology.faceSize() * sizeof( float ); | ||
| return fastWindingNumberMeshMemory( mesh ); | ||
| } |
Contributor
There was a problem hiding this comment.
should we add here some minimum fixed amount?
source/MRMesh/MRChunkIterator.h
Outdated
Comment on lines
14
to
15
| /// chunk index | ||
| size_t index; |
Contributor
There was a problem hiding this comment.
do we need to store index in chunk?
Contributor
Author
There was a problem hiding this comment.
It is useful for logging and progress callbacks.
source/MRMesh/MRChunkIterator.cpp
Outdated
|
|
||
| const auto size = totalSize - overlap; // otherwise the last chunk's size may be smaller or equal to the overlap i.e. fully in the previous chunk | ||
| const auto step = chunkSize - overlap; | ||
| return ( size / step ) + !!( size % step ); // integer variant of `std::ceil( a / b )` |
Contributor
There was a problem hiding this comment.
return (size + step - 1) / step;
Comment on lines
39
to
42
| const Dipole* __restrict__ dipoles; | ||
| const Node3* __restrict__ nodes; | ||
| const float3* __restrict__ meshPoints; | ||
| const FaceToThreeVerts* __restrict__ faces; |
Contributor
There was a problem hiding this comment.
should we initialize it with nulltprs?
|
|
||
| const auto q = ( meshPoints[face.verts[0]] + meshPoints[face.verts[1]] + meshPoints[face.verts[2]] ) / 3.0f; | ||
| processPoint( q, resVec[index], dipoles, nodes, meshPoints, faces, beta, index ); | ||
| processPoint( q, resVec[index], dipoles, nodes, meshPoints, faces, beta ); |
Contributor
There was a problem hiding this comment.
should we pass faceIndex instead of index here (please compare to cpu version)
| res.resize( size ); | ||
| CUDA_LOGE_RETURN_UNEXPECTED( data_->cudaPoints.fromVector( points ) ); | ||
| // TODO: allow user to set the upper limit | ||
| const auto maxBufferBytes = getCudaAvailableMemory(); |
Contributor
There was a problem hiding this comment.
lets use ~70-80% of max available memory just in case
Comment on lines
+93
to
+94
| DynamicArray<float3> cudaPoints; | ||
| CUDA_LOGE_RETURN_UNEXPECTED( cudaPoints.resize( bufferSize ) ); |
Contributor
There was a problem hiding this comment.
in this case we use 3 floats, while bufferSize was counted for one float
Grantim
approved these changes
Feb 25, 2025
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.