We are pleased to present the proceedings of High-Performance Graphics 2009. This is the first year of this new conference, but the conference's roots run deep and wide because it is the merger of two existing, successful conferences.
•Graphics Hardware, an annual conference since 1986 focusing on graphics hardware, architecture, and systems, and
•Interactive Ray Tracing, a symposium established in 2006 focusing on the emerging field of interactive ray tracing and global illumination techniques.
The goal of combining these two conferences was to bring to authors and attendees the best of both, while extending the scope of the new conference to cover the overarching field of performance-oriented graphics systems covering innovative algorithms, efficient implementations, and hardware architecture. This broader focus offers a common forum bringing together researchers, engineers, and architects to discuss the complex interactions of massively-parallel hardware, novel programming models, efficient graphics algorithms, and innovative applications.
The merger of these two conferences had several motivations, but we believe that the primary meaning of the merger is something very exciting. We believe it indicates that ray tracing and graphics hardware are truly coming together, providing graphics hardware with the capabilities and generality of the ray tracing approach, while providing the ray tracing community with wide SIMD many-core platforms and better exposure to the enormous interactive graphics community.
In previous years, GH typically received around 25 paper submissions and IRT received around 40. We were delighted to receive 72 submissions to HPG, in spite of a late announcement of the conference to the research community. The submission pool was extremely strong, reflecting the breadth and depth of innovative work in this area.
We chose to have a massively-parallel papers committee, with fifty members. This was quite helpful because of the larger-than- expected number of submissions. We also felt it was important to not load the committee members too heavily, since many were also serving on the EGSR, Vis, or Supercomputing committees at the same time. The committee members were very thoughtful and diligent in choosing external reviewers and in providing useful, detailed reviews of their assigned papers. We greatly appreciate their efforts.
We depended heavily on the reviews, the reviewer discussion, and the scores, but the final decision on all papers rested with the three of us, as had been the GH tradition. The highest-scored rejected paper scored 3.75 (out of 5.0) and the lowest-scored accepted paper scored 3.0. This reflects the great amount of deliberation over the papers and our goals in selecting papers. We ultimately selected 21 of the 72 papers, a 29% acceptance rate. This was quite competitive, and we hope it will serve to bring respect and quality to the conference, though we ached at some of the rejection decisions.
Selecting a set of papers that would satisfy the needs and interests of the GH and IRT communities was our most important task. We tried to find a mix of papers that would have been accepted at the previous conferences. For example, IRT was traditionally friendlier to immediately-actionable practical work than GH, and we tried to preserve that spirit in selecting papers this year. We also especially tried to choose the best of the papers that represented the synergy between the two fields. We thought carefully about what the two fields have to teach each other and selected some papers that we thought would be of particular benefit to the other community than that of the paper's primary topic. Anti-aliasing is one research area where we hope to see the two fields press forward together over the next few years to find improved, unified solutions.
Having seen all the submissions and the accepted papers we are very excited about the merged conference and are confident that the merger will bear good fruit. As you read the papers and attend the conference our hope is that you will agree and will benefit from the merged conference and yourself contribute to the field in years to come.
Proceeding Downloads
Spatial splits in bounding volume hierarchies
Bounding volume hierarchies (BVH) have become a widely used alternative to kD-trees as the acceleration structure of choice in modern ray tracing systems. However, BVHs adapt poorly to non-uniformly tessellated scenes, which leads to increased ray ...
Object partitioning considered harmful: space subdivision for BVHs
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 ...
A parallel algorithm for construction of uniform grids
We present a fast, parallel GPU algorithm for construction of uniform grids for ray tracing, which we implement in CUDA. The algorithm performance does not depend on the primitive distribution, because we reduce the problem to sorting pairs of ...
CFU: multi-purpose configurable filtering unit for mobile multimedia applications on graphics hardware
In order to increase the capability of mobile GPUs in image/video processing, a multi-purpose configurable filtering unit (CFU), which is a new configurable unit for image filtering on stream processing architecture, is proposed in this paper. CFU is ...
Scaling of 3D game engine workloads on modern multi-GPU systems
This work supposes a first attempt to characterize the 3D game workload running on commodity multi-GPU systems. Depending on the rendering workload balance mode used, the intra and interframe dependencies due to render-to-texture require a number of ...
Embedded function composition
A low-level graphics processor is assembled from a collection of hardwired functions of screen coordinates embedded directly in the display. Configuration of these functions is controlled by a buffer containing parameters delivered to the processor on-...
Efficient depth peeling via bucket sort
In this paper we present an efficient algorithm for multi-layer depth peeling via bucket sort of fragments on GPU, which makes it possible to capture up to 32 layers simultaneously with correct depth ordering in a single geometry pass. We exploit ...
Data-parallel rasterization of micropolygons with defocus and motion blur
Current GPUs rasterize micropolygons (polygons approximately one pixel in size) inefficiently. We design and analyze the costs of three alternative data-parallel algorithms for rasterizing micropolygon workloads for the real-time domain. First, we ...
Accelerating shadow rays using volumetric occluders and modified kd-tree traversal
Monte Carlo ray tracing remains a simple and elegant method for generating robust shadows. This approach, however, is often hampered by the time needed to evaluate the numerous shadow ray queries required to generate a high-quality image. We propose the ...
Hardware-accelerated global illumination by image space photon mapping
We describe an extension to photon mapping that recasts the most expensive steps of the algorithm -- the initial and final photon bounces -- as image-space operations amenable to GPU acceleration. This enables global illumination for real-time ...
Image space gathering
Soft shadows, glossy reflections and depth of field are valuable effects for realistic rendering and are often computed using distribution ray tracing (DRT). These "blurry" effects often need not be accurate and are sometimes simulated by blurring an ...
Parallel view-dependent tessellation of Catmull-Clark subdivision surfaces
We present a strategy for performing view-adaptive, crack-free tessellation of Catmull-Clark subdivision surfaces entirely on programmable graphics hardware. Our scheme extends the concept of breadth-first subdivision, which up to this point has only ...
Morphological antialiasing
We present a new algorithm that creates plausibly antialiased images by looking for certain patterns in an original image and then blending colors in the neighborhood of these patterns according to a set of simple rules. We construct these rules to work ...
Selective and adaptive supersampling for real-time ray tracing
While supersampling is an essential element for high quality rendering, high sampling rates, routinely employed in offline rendering, are still considered quite burdensome for real-time ray tracing. In this paper, we propose a selective and adaptive ...
A directionally adaptive edge anti-aliasing filter
The latest generation of graphics hardware provides direct access to multisample anti-aliasing (MSAA) rendering data. By taking advantage of these existing pixel subsample values, an intelligent reconstruction filter can be computed using programmable ...
Efficient ray traced soft shadows using multi-frusta tracing
Ray tracing has long been considered to be superior to rasterization because its ability to trace arbitrary rays, allowing it to simulate virtually any physical light transport effect by just tracing rays. Yet, to look plausible, extraordinary amounts ...
Understanding the efficiency of ray traversal on GPUs
We discuss the mapping of elementary ray tracing operations---acceleration structure traversal and primitive intersection---onto wide SIMD/SIMT machines. Our focus is on NVIDIA GPUs, but some of the observations should be valid for other wide machines ...
Faster incoherent rays: Multi-BVH ray stream tracing
High fidelity rendering via ray tracing requires tracing incoherent rays for global illumination and other secondary effects. Recent research show that the performance benefits from fast packet traversal schemes that exploit high coherence are lost when ...
Efficient stream compaction on wide SIMD many-core architectures
Stream compaction is a common parallel primitive used to remove unwanted elements in sparse data. This allows highly parallel algorithms to maintain performance over several processing steps and reduces overall memory usage.
For wide SIMD many-core ...
Fast minimum spanning tree for large graphs on the GPU
Graphics Processor Units are used for many general purpose processing due to high compute power available on them. Regular, data-parallel algorithms map well to the SIMD architecture of current GPU. Irregular algorithms on discrete structures like ...
Stream compaction for deferred shading
The GPU leverages SIMD efficiency when shading because it rasterizes a triangle at a time, running the same shader on all of its fragments. Ray tracing sacrifices this shader coherence, and the result is that SIMD units often must run different shaders ...
- Proceedings of the Conference on High Performance Graphics 2009
Recommendations
Acceptance Rates
Year | Submitted | Accepted | Rate |
---|---|---|---|
HPG '13 | 44 | 15 | 34% |
Overall | 44 | 15 | 34% |