Sven Widmer

Fast dynamic memory allocator for massively parallel architectures

Dynamic memory allocation in massively parallel systems often suffers from drastic performance decreases due to the required global synchronization. This is especially true when many allocation or deallocation requests occur in parallel. We propose a method to alleviate this problem by making use of the SIMD parallelism found in most current massively parallel hardware. More specifically, we propose a hybrid dynamic memory allocator operating at the SIMD parallel warp level. Using additional constraints that can be fulfilled for a large class of practically relevant algorithms and hardware systems, we are able to significantly speed-up the dynamic allocation. We present and evaluate a prototypical implementation for modern CUDA-enabled graphics cards, achieving an overall speedup of up to several orders of magnitude.

An adaptive acceleration structure for screen-space ray tracing

We propose an efficient acceleration structure for real-time screen-space ray tracing. The hybrid data structure represents the scene geometry by combining a bounding volume hierarchy with local planar approximations. This enables fast empty space skipping while tracing and yields exact intersection points for the planar approximation. In combination with an occlusion-aware ray traversal our algorithm is capable to quickly trace even multiple depth layers. Compared to prior work, our technique improves the accuracy of the results, is more general, and allows for advanced image transformations, as all pixels can cast rays to arbitrary directions. We demonstrate real-time performance for several applications, including depth-of-field rendering, stereo warping, and screen-space ray traced reflections.

Visual-interactive Exploration of Interesting Multivariate Relations in Mixed Research Data Sets

The analysis of research data plays a key role in data‐driven areas of science. Varieties of mixed research data sets exist and scientists aim to derive or validate hypotheses to find undiscovered knowledge. Many analysis techniques identify relations of an entire dataset only. This may level the characteristic behavior of different subgroups in the data. Like automatic subspace clustering, we aim at identifying interesting subgroups and attribute sets. We present a visual‐interactive system that supports scientists to explore interesting relations between aggregated bins of multivariate attributes in mixed data sets. The abstraction of data to bins enables the application of statistical dependency tests as the measure of interestingness. An overview matrix view shows all attributes, ranked with respect to the interestingness of bins. Complementary, a node‐link view reveals multivariate bin relations by positioning dependent bins close to each other. The system supports information drill‐down based on both expert knowledge and algorithmic support. Finally, visual‐interactive subset clustering assigns multivariate bin relations to groups. A list‐based cluster result representation enables the scientist to communicate multivariate findings at a glance. We demonstrate the applicability of the system with two case studies from the earth observation domain and the prostate cancer research domain. In both cases, the system enabled us to identify the most interesting multivariate bin relations, to validate already published results, and, moreover, to discover unexpected relations.

Is your permutation algorithm unbiased for n ≠ 2 m ?

Many papers on parallel random permutation algorithms assume the input size n to be a power of two and imply that these algorithms can be easily generalized to arbitrary n. We show that this simplifying assumption is not necessarily correct since it may result in a bias. Many of these algorithms are, however, consistent, i.e., iterating them ultimately converges against an unbiased permutation. We prove this convergence along with proving exponential convergence speed. Furthermore, we present an analysis of iterating applied to a butterfly permutation network, which works in-place and is well-suited for implementation on many-core systems such as GPUs. We also show a method that improves the convergence speed even further and yields a practical implementation of the permutation network on current GPUs.

Efficient Heuristic Adaptive Quadrature on GPUs: Design and Evaluation

Numerical integration is a common sub-problem in many applications. It can be solved easily in CPU-based applications using adaptive quadrature such as the adaptive Simpson’s rule. These algorithms rely, however, on error estimation yielding a significant computational overhead. In addition, they require recursive function evaluations, which are not well suited for parallel computation on graphics processing units (GPUs) due to warp divergence issues. In this paper, we introduce heuristic forward quadrature as an alternative that is not only more efficient than traditional methods, but also better suited for accelerated massively-parallel calculation on GPUs. Additionally, we will give an error estimate for our method and demonstrate performance results for 1D and 2D integral applications which show that the algorithm leverages quadrature for the efficient implementation on GPUs.

Using graphics processing units to investigate molecular coevolution

We present a massively parallel implementation of the computation of (co)evolutionary signals from biomolecular sequence alignments based on mutual information (MI) and a normalization procedure to neutral evolution. The MI is computed for two‐point and three‐point correlations within any multiple sequence alignment. We meet the high computational demand in the normalization procedure efficiently with an implementation on Graphics Processing Units (GPUs) using NVIDIAs CUDA framework. In particular, the normalization of the MI for three‐point ‘cliques’ of amino acids or nucleotides requires large sampling numbers in the normalization, which we achieve by using GPUs. GPU computation serves as an enabling technology here insofar as MI normalization is also possible using traditional computational methods [1] or cluster computation, but only GPU computation makes MI normalization for sequence analysis feasible in a statistically sufficient sample and in acceptable time given affordable commodity hardware. We illustrate (i) the computational efficiency and (ii) the biological usefulness of two‐point and three‐point MI by applications to the well‐known protein calmodulin and the variable surface glycoprotein (VSG) of Trypanosoma brucei, which are subject to involved evolutionary pressure. Here, we find striking coevolutionary patterns and distinct information on the molecular evolution of these molecules that question previous work that relied on non‐efficient MI computations. Copyright

Decoupled space and time sampling of motion and defocus blur for unified rendering of transparent and opaque objects

We propose a unified rendering approach that jointly handles motion and defocus blur for transparent and opaque objects at interactive frame rates. Our key idea is to create a sampled representation of all parts of the scene geometry that are potentially visible at any point in time for the duration of a frame in an initial rasterization step. We store the resulting temporally‐varying fragments (t‐fragments) in a bounding volume hierarchy which is rebuild every frame using a fast spatial median construction algorithm. This makes our approach suitable for interactive applications with dynamic scenes and animations. Next, we perform spatial sampling to determine all t‐fragments that intersect with a specific viewing ray at any point in time. Viewing rays are sampled according to the lens uv‐sampling for depth‐of‐field effects. In a final temporal sampling step, we evaluate the predetermined viewing ray/t‐fragment intersections for one or multiple points in time. This allows us to incorporate all standard shading effects including transparency. We describe the overall framework, present our GPU implementation, and evaluate our rendering approach with respect to scalability, quality, and performance.

Analysis of cache behavior and performance of different BVH memory layouts for tracing incoherent rays

With CPUs moving towards many-core architectures and GPUs becoming more general purpose architectures, path tracing can now be well parallelized on commodity hardware. While parallelization is trivial in theory, properties of real hardware make efficient parallelization difficult, especially when tracing incoherent rays. We investigate how different bounding volume hierarchy (BVH) and node memory layouts as well as storing the BVH in different memory areas impacts the ray tracing performance of a GPU path tracer. We optimize the BVH layout using information gathered in a pre-processing pass applying a number of different BVH reordering techniques. Depending on the memory area and scene complexity, we achieve moderate speedups.