Expand description
Optimized sparse matrix implementations with SIMD acceleration and buffer pooling.
This module provides high-performance matrix storage formats optimized for sublinear-time algorithms with focus on minimizing memory allocation overhead and maximizing cache efficiency.
Structs§
- Buffer
Pool - High-performance buffer pool for reducing allocation overhead.
- Buffer
Pool Stats - Buffer pool statistics.
- OptimizedCSR
Storage - Optimized CSR storage with SIMD acceleration and buffer pooling.
- Optimized
Matrix Stats - Performance statistics for optimized matrix operations.
- ParallelCSR
Storage - Parallel CSR storage for multi-threaded operations.
- Streaming
Matrix - Memory-efficient matrix representation for extremely large problems.
Functions§
- get_
global_ buffer - Get a buffer from the global pool.
- return_
global_ buffer - Return a buffer to the global pool.