3 results about "Round-off error" patented technology

The application belongs to the field of high-performance computing, and more particularly, a large eddy simulation large-scale parallelization and mixed precision optimization method for a domestic high-performance computing system is designed. Facing the domestic high-performance computing platform supporting multiple floating point precisions, a multi-level parallel optimization strategy is adopted, and the many-core accelerator architecture characteristics of the domestic high-performance computer are deeply integrated. The core scheme includes a multi-level parallel and optimization strategy for the domestic supercomputer general architecture, and an efficient collaborative computing scheme implemented on the CPU+DSP heterogeneous platform. The application faces the Tianhe new generation high-performance computing platform supporting multiple floating point precisions, and according to the numerical calculation rounding error accumulation characteristics in the application field, efficient parallel mixed precision calculation is realized through the programming environment, the memory access amount and the communication amount are effectively reduced, and the large eddy simulation calculation efficiency is improved, thereby providing strong technical support for the domestic transplantation and performance optimization of large-scale numerical simulation.

While the symbol occurrence count is rounded to generate the symbol frequency, the lossless data compressor immediately prevents normalization overflow, allowing the code table generator to generate code table entries without waiting for the symbol frequency table to finish filling. As symbols are normalized, rounding errors accumulate, and are compensated for by reducing the symbol frequency when the symbol frequency is at least 2 and the accumulated error exceeds a threshold. The symbol frequency is also reduced when the number of remaining states in the code table is insufficient to accommodate the number of remaining unprocessed symbols and states for the current code table entry. Because error compensation occurs during symbol normalization, it does not force all symbols in the waiting block to be processed before generating the code table, thus reducing latency. Three pipeline stages operate on three input blocks: symbol counting, normalization / error compensation / code table generation, and data encoding.

An adaptive quantization method considering a distance range for LiDAR-based point cloud compression is provided. According to an embodiment of the present invention, the adaptive quantization method classifies point cloud data on the basis of a distance from an origin, and quantizes the classified point cloud data using different quantization schemes. Accordingly, by performing adaptive quantization that takes the distance range of point cloud data into consideration, rounding errors and clipping errors during quantization are significantly reduced, thereby enabling data compression and integer conversion while minimizing data loss and accuracy degradation.