High-Performance Graphics 2020
Permanent URI for this collection
Browse
Browsing High-Performance Graphics 2020 by Author "Seiler, Larry"
Now showing 1 - 3 of 3
Results Per Page
Sort Options
Item Compacted CPU/GPU Data Compression via Modified Virtual Address Translation(ACM, 2020) Seiler, Larry; Lin, Daqi; Yuksel, Cem; Yuksel, Cem and Membarth, Richard and Zordan, VictorWe propose a method to reduce the footprint of compressed data by using modified virtual address translation to permit random access to the data. This extends our prior work on using page translation to perform automatic decompression and deswizzling upon accesses to fixed rate lossy or lossless compressed data. Our compaction method allows a virtual address space the size of the uncompressed data to be used to efficiently access variable-size blocks of compressed data. Compression and decompression take place between the first and second level caches, which allows fast access to uncompressed data in the first level cache and provides data compaction at all other levels of the memory hierarchy. This improves performance and reduces power relative to compressed but uncompacted data. An important property of our method is that compression, decompression, and reallocation are automatically managed by the new hardware without operating system intervention and without storing compression data in the page tables. As a result, although some changes are required in the page manager, it does not need to know the specific compression algorithm and can use a single memory allocation unit size. We tested our method with two sample CPU algorithms. When performing depth buffer occlusion tests, our method reduces the memory footprint by 3.1x. When rendering into textures, our method reduces the footprint by 1.69x before rendering and 1.63x after. In both cases, the power and cycle time are better than for uncompacted compressed data, and significantly better than for accessing uncompressed data.Item Efficient Adaptive Deferred Shading with Hardware Scatter Tiles(ACM, 2020) Mallett, Ian; Yuksel, Cem; Seiler, Larry; Yuksel, Cem and Membarth, Richard and Zordan, VictorAdaptive shading is an effective mechanism for reducing the number of shaded pixels to a subset of the image resolution with minimal impact on final rendering quality. We present a new scheduling method based on on-chip tiles that, along with relatively minor modifications to the GPU architecture, provides efficient hardware support. As compared to software implementations on current hardware using compute shaders, our approach dramatically reduces memory bandwidth requirements, thereby significantly improving performance and energy use. We also introduce the concept of a fragment pre-shader for programmatically controlling when a fragment shader is invoked, and describe advanced techniques for utilizing our approach to further reduce the number of shaded pixels via temporal filtering, or to adjust rendering quality to maintain stable framerates.Item Quadratic Approximation of Cubic Curves(ACM, 2020) Truong, Nghia; Yuksel, Cem; Seiler, Larry; Yuksel, Cem and Membarth, Richard and Zordan, VictorWe present a simple degree reduction technique for piecewise cubic polynomial splines, converting them into piecewise quadratic splines that maintain the parameterization and C1 continuity. Our method forms identical tangent directions at the interpolated data points of the piecewise cubic spline by replacing each cubic piece with a pair of quadratic pieces. The resulting representation can lead to substantial performance improvements for rendering geometrically complex spline models like hair and fiber-level cloth. Such models are typically represented using cubic splines that are C1-continuous, a property that is preserved with our degree reduction. Therefore, our method can also be considered a new quadratic curve construction approach for high-performance rendering. We prove that it is possible to construct a pair of quadratic curves with C1 continuity that passes through any desired point on the input cubic curve. Moreover, we prove that when the pair of quadratic pieces corresponding to a cubic piece have equal parametric lengths, they join exactly at the parametric center of the cubic piece, and the deviation in positions due to degree reduction is minimized.