scirs2-linalg 0.4.0

Linear algebra module for SciRS2 (scirs2-linalg)
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49

//! GPU-accelerated linear algebra operations
//!
//! This module provides a comprehensive set of GPU-accelerated linear algebra operations
//! with automatic CPU/GPU dispatching, performance optimization, and intelligent workload
//! analysis.

// Module declarations
pub mod advanced;
pub mod decompositions;
pub mod dispatcher;
pub mod hardware;
pub mod intelligent;
pub mod kernels;
pub mod matrix_ops;
pub mod metrics;
pub mod optimization;
pub mod profiling;

#[cfg(test)]
mod tests;

// Re-export main public APIs
pub use advanced::AdvancedGpuOperations;
pub use dispatcher::{GpuOperationDispatcher, DEFAULT_GPU_THRESHOLD};
pub use hardware::{DeviceProfile, HardwareCapabilityProfiler};
pub use intelligent::{
    AdvancedIntelligentGpuDispatcher, DataCharacteristics, DispatchDecision,
    GpuPerformancePredictor, MemoryAccessPattern, ModelCoefficients, OptimalChoice,
    PerformancePrediction, WorkloadAnalysis, WorkloadAnalyzer,
};
pub use kernels::{AutoTuneResults, BenchmarkResults, GpuKernelManager, OptimizationLevel};
pub use metrics::{
    EnergyMetrics, MemoryMetrics, MultiDimensionalMetrics, RunningStats, ThroughputMetrics,
    TimeMetrics,
};
pub use optimization::{BatchPerformanceRecord, BatchSizeOptimizer};
pub use profiling::GpuPerformanceProfiler;

// Re-export matrix operations
pub use matrix_ops::{
    BatchedGemmConfig, GemmConfig, GpuMatrixOp, GpuMatrixOpConfig, GpuMatrixOperations,
    NeuralNetworkOps,
};

// Re-export decomposition operations
pub use decompositions::{
    CholeskyDecomposition, EigenDecomposition, GpuDecompositionConfig, GpuDecompositions,
    LuDecomposition, QrDecomposition, SvdDecomposition,
};