Struct PerformanceOptimizer

Source

pub struct PerformanceOptimizer { /* private fields */ }

Expand description

Unified performance optimization manager

Combines all performance optimization techniques including SIMD, memory efficiency, and parallel processing to provide optimal performance for neural network operations.

Implementations§

Source §

impl PerformanceOptimizer

Source

pub fn new( _chunk_size: Option<usize>, _max_memory_mb: Option<usize>, num_threads: Option<usize>, enable_profiling: bool, ) -> Result<Self>

Create a new performance optimizer

§Arguments

chunk_size - Chunk size for memory-efficient processing
max_memory_mb - Maximum memory usage in MB
num_threads - Number of threads for parallel processing
enable_profiling - Whether to enable performance profiling

§Examples

use scirs2_neural::performance::PerformanceOptimizer;

let optimizer = PerformanceOptimizer::new(
    Some(256),  // 256 samples per chunk
    Some(1024), // 1GB memory limit
    Some(8),    // 8 threads
    true        // enable profiling
).unwrap();

Examples found in repository ?

examples/new_features_showcase.rs (line 162)

128fn demonstrate_performance_optimizations() -> Result<(), Box<dyn std::error::Error>> {
129    println!("⚡ Performance Optimizations Demonstration");
130    println!("========================================\n");
131
132    // 1. Thread Pool Manager
133    println!("1. Thread Pool Manager:");
134    let thread_pool = ThreadPoolManager::new(Some(4))?;
135    println!(
136        "   Created thread pool with {} threads",
137        thread_pool.num_threads()
138    );
139
140    // Demonstrate parallel matrix multiplication
141    let matrix_a = Array::from_elem((100, 50), 2.0f32).into_dyn();
142    let matrix_b = Array::from_elem((50, 75), 3.0f32).into_dyn();
143
144    let start_time = std::time::Instant::now();
145    let result = thread_pool.parallel_matmul(&matrix_a, &matrix_b)?;
146    let elapsed = start_time.elapsed();
147
148    println!(
149        "   Parallel matrix multiplication: {}x{} * {}x{} = {}x{}",
150        matrix_a.shape()[0],
151        matrix_a.shape()[1],
152        matrix_b.shape()[0],
153        matrix_b.shape()[1],
154        result.shape()[0],
155        result.shape()[1]
156    );
157    println!("   Time elapsed: {:.3}ms", elapsed.as_secs_f64() * 1000.0);
158    println!("   Result sample: {:.1}", result[[0, 0]]);
159
160    // 2. Performance Profiler
161    println!("\n2. Performance Profiler:");
162    let mut optimizer = PerformanceOptimizer::new(Some(1024), Some(512), Some(4), true)?;
163
164    // Simulate some operations with profiling
165    {
166        let timer = optimizer.profiler_mut().start_timer("matrix_setup");
167        let test_matrix_a = Array::from_elem((200, 100), 1.5f32).into_dyn();
168        let test_matrix_b = Array::from_elem((100, 150), 2.5f32).into_dyn();
169        optimizer
170            .profiler_mut()
171            .end_timer("matrix_setup".to_string(), timer);
172
173        let _result = optimizer.optimized_matmul(&test_matrix_a, &test_matrix_b)?;
174    }
175
176    println!("   Performance profile summary:");
177    optimizer.profiler().print_summary();
178
179    // 3. Optimization Capabilities
180    println!("\n3. Optimization Capabilities:");
181    let capabilities = optimizer.get_capabilities();
182    println!("{}", capabilities);
183
184    println!("✅ Performance optimizations demonstration completed!\n");
185    Ok(())
186}

Source

pub fn thread_pool(&self) -> &Arc<ThreadPoolManager>

Get reference to thread pool

Source

pub fn profiler_mut(&mut self) -> &mut PerformanceProfiler

Get mutable reference to profiler

Examples found in repository ?

examples/new_features_showcase.rs (line 166)

128fn demonstrate_performance_optimizations() -> Result<(), Box<dyn std::error::Error>> {
129    println!("⚡ Performance Optimizations Demonstration");
130    println!("========================================\n");
131
132    // 1. Thread Pool Manager
133    println!("1. Thread Pool Manager:");
134    let thread_pool = ThreadPoolManager::new(Some(4))?;
135    println!(
136        "   Created thread pool with {} threads",
137        thread_pool.num_threads()
138    );
139
140    // Demonstrate parallel matrix multiplication
141    let matrix_a = Array::from_elem((100, 50), 2.0f32).into_dyn();
142    let matrix_b = Array::from_elem((50, 75), 3.0f32).into_dyn();
143
144    let start_time = std::time::Instant::now();
145    let result = thread_pool.parallel_matmul(&matrix_a, &matrix_b)?;
146    let elapsed = start_time.elapsed();
147
148    println!(
149        "   Parallel matrix multiplication: {}x{} * {}x{} = {}x{}",
150        matrix_a.shape()[0],
151        matrix_a.shape()[1],
152        matrix_b.shape()[0],
153        matrix_b.shape()[1],
154        result.shape()[0],
155        result.shape()[1]
156    );
157    println!("   Time elapsed: {:.3}ms", elapsed.as_secs_f64() * 1000.0);
158    println!("   Result sample: {:.1}", result[[0, 0]]);
159
160    // 2. Performance Profiler
161    println!("\n2. Performance Profiler:");
162    let mut optimizer = PerformanceOptimizer::new(Some(1024), Some(512), Some(4), true)?;
163
164    // Simulate some operations with profiling
165    {
166        let timer = optimizer.profiler_mut().start_timer("matrix_setup");
167        let test_matrix_a = Array::from_elem((200, 100), 1.5f32).into_dyn();
168        let test_matrix_b = Array::from_elem((100, 150), 2.5f32).into_dyn();
169        optimizer
170            .profiler_mut()
171            .end_timer("matrix_setup".to_string(), timer);
172
173        let _result = optimizer.optimized_matmul(&test_matrix_a, &test_matrix_b)?;
174    }
175
176    println!("   Performance profile summary:");
177    optimizer.profiler().print_summary();
178
179    // 3. Optimization Capabilities
180    println!("\n3. Optimization Capabilities:");
181    let capabilities = optimizer.get_capabilities();
182    println!("{}", capabilities);
183
184    println!("✅ Performance optimizations demonstration completed!\n");
185    Ok(())
186}

Source

pub fn profiler(&self) -> &PerformanceProfiler

Get reference to profiler

Examples found in repository ?

examples/new_features_showcase.rs (line 177)

128fn demonstrate_performance_optimizations() -> Result<(), Box<dyn std::error::Error>> {
129    println!("⚡ Performance Optimizations Demonstration");
130    println!("========================================\n");
131
132    // 1. Thread Pool Manager
133    println!("1. Thread Pool Manager:");
134    let thread_pool = ThreadPoolManager::new(Some(4))?;
135    println!(
136        "   Created thread pool with {} threads",
137        thread_pool.num_threads()
138    );
139
140    // Demonstrate parallel matrix multiplication
141    let matrix_a = Array::from_elem((100, 50), 2.0f32).into_dyn();
142    let matrix_b = Array::from_elem((50, 75), 3.0f32).into_dyn();
143
144    let start_time = std::time::Instant::now();
145    let result = thread_pool.parallel_matmul(&matrix_a, &matrix_b)?;
146    let elapsed = start_time.elapsed();
147
148    println!(
149        "   Parallel matrix multiplication: {}x{} * {}x{} = {}x{}",
150        matrix_a.shape()[0],
151        matrix_a.shape()[1],
152        matrix_b.shape()[0],
153        matrix_b.shape()[1],
154        result.shape()[0],
155        result.shape()[1]
156    );
157    println!("   Time elapsed: {:.3}ms", elapsed.as_secs_f64() * 1000.0);
158    println!("   Result sample: {:.1}", result[[0, 0]]);
159
160    // 2. Performance Profiler
161    println!("\n2. Performance Profiler:");
162    let mut optimizer = PerformanceOptimizer::new(Some(1024), Some(512), Some(4), true)?;
163
164    // Simulate some operations with profiling
165    {
166        let timer = optimizer.profiler_mut().start_timer("matrix_setup");
167        let test_matrix_a = Array::from_elem((200, 100), 1.5f32).into_dyn();
168        let test_matrix_b = Array::from_elem((100, 150), 2.5f32).into_dyn();
169        optimizer
170            .profiler_mut()
171            .end_timer("matrix_setup".to_string(), timer);
172
173        let _result = optimizer.optimized_matmul(&test_matrix_a, &test_matrix_b)?;
174    }
175
176    println!("   Performance profile summary:");
177    optimizer.profiler().print_summary();
178
179    // 3. Optimization Capabilities
180    println!("\n3. Optimization Capabilities:");
181    let capabilities = optimizer.get_capabilities();
182    println!("{}", capabilities);
183
184    println!("✅ Performance optimizations demonstration completed!\n");
185    Ok(())
186}

Source

pub fn get_capabilities(&self) -> &OptimizationCapabilities

Get optimization capabilities

Examples found in repository ?

examples/new_features_showcase.rs (line 181)

128fn demonstrate_performance_optimizations() -> Result<(), Box<dyn std::error::Error>> {
129    println!("⚡ Performance Optimizations Demonstration");
130    println!("========================================\n");
131
132    // 1. Thread Pool Manager
133    println!("1. Thread Pool Manager:");
134    let thread_pool = ThreadPoolManager::new(Some(4))?;
135    println!(
136        "   Created thread pool with {} threads",
137        thread_pool.num_threads()
138    );
139
140    // Demonstrate parallel matrix multiplication
141    let matrix_a = Array::from_elem((100, 50), 2.0f32).into_dyn();
142    let matrix_b = Array::from_elem((50, 75), 3.0f32).into_dyn();
143
144    let start_time = std::time::Instant::now();
145    let result = thread_pool.parallel_matmul(&matrix_a, &matrix_b)?;
146    let elapsed = start_time.elapsed();
147
148    println!(
149        "   Parallel matrix multiplication: {}x{} * {}x{} = {}x{}",
150        matrix_a.shape()[0],
151        matrix_a.shape()[1],
152        matrix_b.shape()[0],
153        matrix_b.shape()[1],
154        result.shape()[0],
155        result.shape()[1]
156    );
157    println!("   Time elapsed: {:.3}ms", elapsed.as_secs_f64() * 1000.0);
158    println!("   Result sample: {:.1}", result[[0, 0]]);
159
160    // 2. Performance Profiler
161    println!("\n2. Performance Profiler:");
162    let mut optimizer = PerformanceOptimizer::new(Some(1024), Some(512), Some(4), true)?;
163
164    // Simulate some operations with profiling
165    {
166        let timer = optimizer.profiler_mut().start_timer("matrix_setup");
167        let test_matrix_a = Array::from_elem((200, 100), 1.5f32).into_dyn();
168        let test_matrix_b = Array::from_elem((100, 150), 2.5f32).into_dyn();
169        optimizer
170            .profiler_mut()
171            .end_timer("matrix_setup".to_string(), timer);
172
173        let _result = optimizer.optimized_matmul(&test_matrix_a, &test_matrix_b)?;
174    }
175
176    println!("   Performance profile summary:");
177    optimizer.profiler().print_summary();
178
179    // 3. Optimization Capabilities
180    println!("\n3. Optimization Capabilities:");
181    let capabilities = optimizer.get_capabilities();
182    println!("{}", capabilities);
183
184    println!("✅ Performance optimizations demonstration completed!\n");
185    Ok(())
186}

Source

pub fn optimized_matmul( &mut self, a: &ArrayD<f32>, b: &ArrayD<f32>, ) -> Result<ArrayD<f32>>

Optimized matrix multiplication using all available optimizations

Automatically selects the best optimization strategy based on matrix size, available features, and system capabilities.

Examples found in repository ?

examples/new_features_showcase.rs (line 173)

128fn demonstrate_performance_optimizations() -> Result<(), Box<dyn std::error::Error>> {
129    println!("⚡ Performance Optimizations Demonstration");
130    println!("========================================\n");
131
132    // 1. Thread Pool Manager
133    println!("1. Thread Pool Manager:");
134    let thread_pool = ThreadPoolManager::new(Some(4))?;
135    println!(
136        "   Created thread pool with {} threads",
137        thread_pool.num_threads()
138    );
139
140    // Demonstrate parallel matrix multiplication
141    let matrix_a = Array::from_elem((100, 50), 2.0f32).into_dyn();
142    let matrix_b = Array::from_elem((50, 75), 3.0f32).into_dyn();
143
144    let start_time = std::time::Instant::now();
145    let result = thread_pool.parallel_matmul(&matrix_a, &matrix_b)?;
146    let elapsed = start_time.elapsed();
147
148    println!(
149        "   Parallel matrix multiplication: {}x{} * {}x{} = {}x{}",
150        matrix_a.shape()[0],
151        matrix_a.shape()[1],
152        matrix_b.shape()[0],
153        matrix_b.shape()[1],
154        result.shape()[0],
155        result.shape()[1]
156    );
157    println!("   Time elapsed: {:.3}ms", elapsed.as_secs_f64() * 1000.0);
158    println!("   Result sample: {:.1}", result[[0, 0]]);
159
160    // 2. Performance Profiler
161    println!("\n2. Performance Profiler:");
162    let mut optimizer = PerformanceOptimizer::new(Some(1024), Some(512), Some(4), true)?;
163
164    // Simulate some operations with profiling
165    {
166        let timer = optimizer.profiler_mut().start_timer("matrix_setup");
167        let test_matrix_a = Array::from_elem((200, 100), 1.5f32).into_dyn();
168        let test_matrix_b = Array::from_elem((100, 150), 2.5f32).into_dyn();
169        optimizer
170            .profiler_mut()
171            .end_timer("matrix_setup".to_string(), timer);
172
173        let _result = optimizer.optimized_matmul(&test_matrix_a, &test_matrix_b)?;
174    }
175
176    println!("   Performance profile summary:");
177    optimizer.profiler().print_summary();
178
179    // 3. Optimization Capabilities
180    println!("\n3. Optimization Capabilities:");
181    let capabilities = optimizer.get_capabilities();
182    println!("{}", capabilities);
183
184    println!("✅ Performance optimizations demonstration completed!\n");
185    Ok(())
186}

Source

pub fn optimized_conv2d( &mut self, input: &ArrayD<f32>, kernel: &ArrayD<f32>, bias: Option<&[f32]>, stride: (usize, usize), padding: (usize, usize), ) -> Result<ArrayD<f32>>

Optimized convolution using all available optimizations

Source

pub fn get_performance_stats(&self) -> PerformanceStats

Get comprehensive performance statistics

Source

pub fn reset_stats(&mut self)

Reset all performance tracking

Source

pub fn benchmark_strategies( &mut self, a: &ArrayD<f32>, b: &ArrayD<f32>, iterations: usize, ) -> Result<BenchmarkResults>

Benchmark different optimization strategies

Auto Trait Implementations§

§

impl UnwindSafe for PerformanceOptimizer

Blanket Implementations§

Source §

impl<T> Any for T
where T: 'static + ?Sized,

Source §

fn type_id(&self) -> TypeId

Gets the TypeId of self. Read more

Source §

impl<T> Borrow<T> for T
where T: ?Sized,

Source §

fn borrow(&self) -> &T

Immutably borrows from an owned value. Read more

Source §

impl<T> BorrowMut<T> for T
where T: ?Sized,

Source §

fn borrow_mut(&mut self) -> &mut T

Mutably borrows from an owned value. Read more

Source §

impl<T> From<T> for T

Source §

fn from(t: T) -> T

Returns the argument unchanged.

Source §

impl<T, U> Into for T
where U: From<T>,

Source §

fn into(self) -> U

Calls U::from(self).

That is, this conversion is whatever the implementation of From<T> for U chooses to do.

Source §

impl<T> IntoEither for T

Source §

fn into_either(self, into_left: bool) -> Either<Self, Self>

Converts self into a Left variant of Either<Self, Self> if into_left is true. Converts self into a Right variant of Either<Self, Self> otherwise. Read more

Source §

fn into_either_with<F>(self, into_left: F) -> Either<Self, Self>
where F: FnOnce(&Self) -> bool,

Converts self into a Left variant of Either<Self, Self> if into_left(&self) returns true. Converts self into a Right variant of Either<Self, Self> otherwise. Read more

Source §