Precomputed sub-pixel coverage maps and scalar lookups speed custom product rendering while preserving quality for complex geometry and transparency.
Uses MSAA intermediate buffers and pretrained blending weights to upscale images with high quality and low runtime overhead on standard GPUs.
Uses MSAA intermediate buffers and pretrained blending weights to upscale images with edge detail on low-power GPUs and low runtime cost.
Segmenting anti-aliased samples into distinct memory pools reduces bandwidth congestion by keeping frequently accessed data in faster graphics memory.
Sequential pixel processing computes visibility via polygon edge equations, eliminating memory buffers and reducing hardware area for surveillance devices.
A graphic rendering system calculates pixel boundary coverage rates to determine display colors.
A multi-sample anti-aliasing load instruction loads data elements into a vector register based on sampling rate and storage format.
Adaptive interpolation circuitry reduces aliasing artifacts at texel boundaries while maintaining desired visual effects in graphics processing.
Selective anti-aliasing reduces computational load by applying super-sampling only to boundary pixels, improving rendering speed and quality.
Preprocessing cubic Bezier splines into implicit polynomial forms enables accurate anti-aliased rendering while reducing CPU processing time and memory usage.
Adaptive sampling partitions the field of view into depth regions to allocate sample rays based on distance.
A path building engine processes un-parameterized user inputs to dynamically modify or extend existing paths in a user interface.
A kernel splatting network aligns temporal samples to generate supersampled images via mixed precision convolution.
Alpha blending attenuates pixel colors based on coverage values to reduce memory usage during back-to-front rendering.
A graphics rendering method decouples shading from rasterization to enable asynchronous processing across multiple processors.
A compressed vertex buffer system reduces CPU-to-GPU data transfer by storing unique vertices with indices.
Programmable sample pattern tables enable dynamic sample locations for compressed graphics data storage.
A parallel scanline algorithm rasterizes vector contour lines and filled areas on many-core hardware.
Calculating transparency values from partial derivatives of mathematical functions to render sharp graphical edges.
Level-set methods compute electrical behavior from geometric descriptions to eliminate manual mesh generation and simulate full technology suites.
GPU instancing control orders primitives within instances to resolve visual artifacts from incorrect overlaps while reducing CPU-GPU communication overhead.
A technique selects between multisampling and supersampling per pixel based on geometry coverage.
Linear Efficient Antialiased Normal mapping captures specularity changes at different magnifications to resolve aliasing artifacts on shiny surfaces.
Accumulating jittered samples with variable fractional alpha values in the screen buffer reduces GPU memory consumption while maintaining high image quality.