Difference between revisions of "User:Vasco.costa"
From BRL-CAD
Vasco.costa (talk | contribs) (Created page with "Hello. I am interested in ray-tracing, and GPGPU. I plan to work on OpenCL acceleration of the BRL-CAD librt rendering pipeline. =Preliminary Schedule= * 1 week - ellipsoid (...") |
|||
| (40 intermediate revisions by 3 users not shown) | |||
| Line 1: | Line 1: | ||
| − | Hello. I am interested in ray-tracing, and GPGPU. I | + | Hello. I am interested in ray-tracing, and GPGPU. I work on the [[User:Vasco.costa/GSoC15/logs|OpenCL acceleration of the BRL-CAD librt rendering pipeline]]: |
| + | * [https://www.google-melange.com/gsoc/project/details/google/gsoc2015/vasc/5657382461898752 GSoC 2015 Abstract] | ||
| + | * [[User:vasco.costa/GSoC15/proposal|GSoC 2015 Project Proposal]] | ||
| + | * [[User:vasco.costa/GSoC15/logs|GSoC 2015 Development Logs]] | ||
| − | = | + | =OCL librt TODO= |
| − | * | + | * Split <code>struct hit</code> in common.cl into two structs. One for <code>shot()</code> results and another for <code>norm()</code> results. This will reduce the amount of memory used to store temporary results between stages. |
| − | * | + | * Execute prefix sums and reductions in <code>clt_frame()</code> on the device to eliminate round-trips. |
| − | * | + | * Refactor code so single-hit and multi-hit don't require recompiling all the sources twice. |
| − | * | + | * Cache the compiled binaries so things aren't recompiled on every launch. |
| − | * | + | * Quadrics Primitives: HYP, SUPERELL. |
| − | * | + | * Grids Primitives: '''DSP''', <s>EBM</s>, VOL. |
| − | * | + | * Composite Primitives: '''PIPE''', '''NMG''', METABALL, EXTRUDE, REVOLVE. |
| + | * Instancing Primitive: SUBMODEL. | ||
| + | * Make Phong shader results look 100% similar to ANSI C librt. | ||
| + | * Support lights in the shader. | ||
| + | * Don't intersect primitives twice. This requires a dynamic memory allocator. | ||
| + | * Smarter kernel scheduling for reduce thread divergence. For example coallesce all the quadric intersections. | ||
| + | * Spatial partitioning (perhaps a hybrid Kd-tree) in order to have early exit on scenes with high depth complexity. | ||
| + | * UV routines for all OCL implemented primitives. | ||
| + | * BREP support. | ||
| + | * BOT plate mode. | ||
| + | * FASTGEN support (both CLINE primitive and bool.cl code). | ||
| − | + | Deprecated primitives: | |
| + | * ARS is converted to BOT. | ||
| + | * HF is converted to DSP. | ||
| + | |||
| + | =OCL librt Development Status= | ||
| + | * Refactored dispatcher, shoot, optical renderer to process many rays in parallel when rendering an image or block. | ||
| + | * HLBVH object partitioning builder in C. traversal in OCL. | ||
| + | * GPU side database storage of OCL implemented primitives. | ||
| + | * Ported compute intensive or critical parts of the dispatcher, boolean evaluation, optical renderer to OCL. | ||
| + | * OCL dispatcher that performs the shot routines for a whole frame. | ||
| + | * OCL rasterizer that does the pixel pushing for a whole frame. | ||
| + | * OCL lighting modes: Phong, Diffuse, Surface Normals. | ||
| + | * OCL lighting modes: Multi-hit transparent (Single Stage Render Pipeline only). | ||
| + | * ARB8, EBM, EHY, ELL, EPA, ETO, PART, REC, RHC, RPC, SPH, TGC, TOR shot routines in OCL. | ||
| + | * Surface normal routines for all OCL implemented primitives. | ||
| + | * CPU HLBVH BOT shot construction with OCL traversal and interpolated per pixel normals. | ||
Latest revision as of 19:57, 24 October 2017
Hello. I am interested in ray-tracing, and GPGPU. I work on the OpenCL acceleration of the BRL-CAD librt rendering pipeline:
OCL librt TODO[edit]
- Split
struct hitin common.cl into two structs. One forshot()results and another fornorm()results. This will reduce the amount of memory used to store temporary results between stages. - Execute prefix sums and reductions in
clt_frame()on the device to eliminate round-trips. - Refactor code so single-hit and multi-hit don't require recompiling all the sources twice.
- Cache the compiled binaries so things aren't recompiled on every launch.
- Quadrics Primitives: HYP, SUPERELL.
- Grids Primitives: DSP,
EBM, VOL. - Composite Primitives: PIPE, NMG, METABALL, EXTRUDE, REVOLVE.
- Instancing Primitive: SUBMODEL.
- Make Phong shader results look 100% similar to ANSI C librt.
- Support lights in the shader.
- Don't intersect primitives twice. This requires a dynamic memory allocator.
- Smarter kernel scheduling for reduce thread divergence. For example coallesce all the quadric intersections.
- Spatial partitioning (perhaps a hybrid Kd-tree) in order to have early exit on scenes with high depth complexity.
- UV routines for all OCL implemented primitives.
- BREP support.
- BOT plate mode.
- FASTGEN support (both CLINE primitive and bool.cl code).
Deprecated primitives:
- ARS is converted to BOT.
- HF is converted to DSP.
OCL librt Development Status[edit]
- Refactored dispatcher, shoot, optical renderer to process many rays in parallel when rendering an image or block.
- HLBVH object partitioning builder in C. traversal in OCL.
- GPU side database storage of OCL implemented primitives.
- Ported compute intensive or critical parts of the dispatcher, boolean evaluation, optical renderer to OCL.
- OCL dispatcher that performs the shot routines for a whole frame.
- OCL rasterizer that does the pixel pushing for a whole frame.
- OCL lighting modes: Phong, Diffuse, Surface Normals.
- OCL lighting modes: Multi-hit transparent (Single Stage Render Pipeline only).
- ARB8, EBM, EHY, ELL, EPA, ETO, PART, REC, RHC, RPC, SPH, TGC, TOR shot routines in OCL.
- Surface normal routines for all OCL implemented primitives.
- CPU HLBVH BOT shot construction with OCL traversal and interpolated per pixel normals.
