Results 1 
8 of
8
Realtime kdtree construction on graphics hardware
 ACM Transactions on Graphics
, 2008
"... We present an algorithm for constructing kdtrees on GPUs. This algorithm achieves realtime performance by exploiting the GPU’s streaming architecture at all stages of kdtree construction. Unlike previous parallel kdtree algorithms, our method builds tree nodes completely in BFS (breadthfirst se ..."
Abstract

Cited by 82 (8 self)
 Add to MetaCart
We present an algorithm for constructing kdtrees on GPUs. This algorithm achieves realtime performance by exploiting the GPU’s streaming architecture at all stages of kdtree construction. Unlike previous parallel kdtree algorithms, our method builds tree nodes completely in BFS (breadthfirst search) order. We also develop a special strategy for large nodes at upper tree levels so as to further exploit the finegrained parallelism of GPUs. For these nodes, we parallelize the computation over all geometric primitives instead of nodes at each level. Finally, in order to maintain kdtree quality, we introduce novel schemes for fast evaluation of node split costs. As far as we know, ours is the first realtime kdtree algorithm on the GPU. The kdtrees built by our algorithm are of comparable quality as those constructed by offline CPU algorithms. In terms of speed, our algorithm is significantly faster than welloptimized singlecore CPU algorithms and competitive with multicore CPU algorithms. Our algorithm provides a general way for handling dynamic scenes on the GPU. We demonstrate the potential of our algorithm in applications involving dynamic scenes, including GPU ray tracing, interactive photon mapping, and point cloud modeling.
Parallel SAH kD tree construction
 In ACM Conference on High Performance Graphics (HPG
, 2010
"... All intext references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately. ..."
Abstract

Cited by 12 (1 self)
 Add to MetaCart
(Show Context)
All intext references underlined in blue are linked to publications on ResearchGate, letting you access and read them immediately.
SLUSALLEK P.: Object Partitioning Considered Harmful: Space Subdivision for BVHs
 In Proceedings of the Conference on High Performance Graphics 2009
, 2009
"... A major factor for the efficiency of ray tracing is the use of good acceleration structures. Recently, bounding volume hierarchies (BVHs) have become the preferred acceleration structures, due to their competitive performance and greater flexibility compared to KD trees. In this paper, we present a ..."
Abstract

Cited by 11 (0 self)
 Add to MetaCart
A major factor for the efficiency of ray tracing is the use of good acceleration structures. Recently, bounding volume hierarchies (BVHs) have become the preferred acceleration structures, due to their competitive performance and greater flexibility compared to KD trees. In this paper, we present a study on algorithms for the construction of optimal BVHs. Due to the exponential nature of the problem, constructing optimal BVHs for ray tracing remains an open topic. By exploiting the linearity of the surface area heuristic (SAH), we develop an algorithm that can find optimal partitions in polynomial time. We further generalize this algorithm and show that every SAHbased KD tree or BVH construction algorithm is a special case of the generic algorithm. Based on a number of experiments with the generic algorithm, we conclude that the assumption of nonterminating rays in the surface area cost model becomes a major obstacle for using the full potential of BVHs. We also observe that enforcing space subdivision helps to improve BVH performance. Finally, we develop a simple space partitioning algorithm for building efficient BVHs.
Parallel SAH kD Tree Construction for Fast Dynamic Scene Ray Tracing
"... The kD tree is a wellstudied acceleration data structure for ray tracing. It is used to organize primitives in a scene to allow efficient execution of intersection operations between rays and the primitives. The highest quality kD tree can be obtained using greedy cost optimization based on a sur ..."
Abstract
 Add to MetaCart
(Show Context)
The kD tree is a wellstudied acceleration data structure for ray tracing. It is used to organize primitives in a scene to allow efficient execution of intersection operations between rays and the primitives. The highest quality kD tree can be obtained using greedy cost optimization based on a surface area heuristc (SAH). While the high quality enables very fast ray tracing times, a key drawback is that the kD tree construction time remains prohibitively expensive. This cost is unreasonable for rendering dynamic scenes for future visual computing applications on emerging multicore systems. Much work has therefore been focused on faster parallel kD tree construction performance at the expense of approximating or ignoring SAH computation, which produces kD trees that degrade rendering time. In this paper, we present new, faster multicore algorithms for building precise SAHoptimized kdtrees. Our best algorithm makes a tradeoff between worse cache performance and higher parallelism to provide up to 7X speedup on 16 cores, using two different kinds of parallelism models, without degrading tree quality and rendering time. 1.
Parallel SAH kD Tree Construction
"... The kD tree is a wellstudied acceleration data structure for ray tracing. It is used to organize primitives in a scene to allow efficient execution of intersection operations between rays and the primitives. The highest quality kD tree can be obtained using greedy cost optimization based on a sur ..."
Abstract
 Add to MetaCart
(Show Context)
The kD tree is a wellstudied acceleration data structure for ray tracing. It is used to organize primitives in a scene to allow efficient execution of intersection operations between rays and the primitives. The highest quality kD tree can be obtained using greedy cost optimization based on a surface area heuristc (SAH). While the high quality enables very fast ray tracing times, a key drawback is that the kD tree construction time remains prohibitively expensive. This cost is unreasonable for rendering dynamic scenes for future visual computing applications on emerging multicore systems. Much work has therefore been focused on faster parallel kD tree construction performance at the expense of approximating or ignoring SAH computation, which produces kD trees that degrade rendering time. In this paper, we present two new parallel algorithms for building precise SAHoptimized kD trees, with different tradeoffs between the total work done and parallel scalability. The algorithms achieve up to 8 × speedup on 32 cores, without degrading tree quality and rendering time, yielding the best reported speedups so far for preciseSAH kD tree construction. 1.
SRDH: Specializing BVH Construction and Traversal Order Using Representative Shadow Ray Sets
"... We derive the Shadow Ray Distribution Heuristic (SRDH), an accurate cost estimator for shadow ray traversal through a bounding volume hierarchy (BVH). The SRDH leverages upfront knowledge of the distribution and intersection results of previously traced shadow rays to construct a shadowrayspecial ..."
Abstract
 Add to MetaCart
(Show Context)
We derive the Shadow Ray Distribution Heuristic (SRDH), an accurate cost estimator for shadow ray traversal through a bounding volume hierarchy (BVH). The SRDH leverages upfront knowledge of the distribution and intersection results of previously traced shadow rays to construct a shadowrayspecialized BVH and choose an associated traversal order policy which together promote early termination by quickly finding occlusions. In scenes containing large amounts of occlusion, SRDH reduces the number of BVH node traversal steps needed for shadow computations between 22 % and 56 % compared to averagecase traversal through SAHconstructed trees. Evaluating the SRDH using a sparse shadow ray set recorded from a 16×16 pixel rendering of the scene consistently produces BVHs whose traversal cost is within 6 % of those built when all shadow rays are available to the metric at the time of construction. The benefits of the SRDH come at the cost of storing an additional BVH in memory and a 2.4 × increase (on average) in BVH construction time.
Edited by:
, 2014
"... doi: 10.3389/fninf.2014.00068 1D3D hybrid modeling—from multicompartment models to full resolution models in space and time ..."
Abstract
 Add to MetaCart
(Show Context)
doi: 10.3389/fninf.2014.00068 1D3D hybrid modeling—from multicompartment models to full resolution models in space and time
Tree Build Heuristics for Spatial Partitioning Trees of 3D Games
, 2013
"... Spatial partitioning trees are needed for processing collision detections efficiently. In order to select split planes for spatial partitioning trees, the tree balance and the number of polygons overlapped with the split plane should be considered. In this paper, the heuristic algorithm controlling ..."
Abstract
 Add to MetaCart
(Show Context)
Spatial partitioning trees are needed for processing collision detections efficiently. In order to select split planes for spatial partitioning trees, the tree balance and the number of polygons overlapped with the split plane should be considered. In this paper, the heuristic algorithm controlling weight values of tree build criteria is proposed for spatial partitioning trees of 3D games. As the weight values are changed, tree build time, Tjunction elimination time which can cause visual artifacts in splitting polygons overlapped with the split plane, rendering speed (frame per second: FPS) according to tree balance are analysed under 3D game simulations.