SIMD-aware

Checks if SIMD instructions are supported Determines whether the target architecture has Single Instruction, Multiple Data (SIMD) capabilities for parallel processing. This is a compile-time architectural capability check, not a runtime feature detection.

SIMD Support by Architecture

• x86: MMX, SSE family instructions
• x86_64: SSE, AVX family instructions
• ARMv7: NEON SIMD instructions
• ARMv8: Advanced NEON, optional SVE
• Other architectures: May have limited or no SIMD support

Performance Implications

• Vectorization: Enables parallel processing of data arrays
• Mathematical Operations: Accelerated floating-point calculations
• Media Processing: Efficient image/audio/video processing

Returns

true if SIMD instruction support is architecturally available

Note

This function is constexpr and evaluates at compile time

For runtime SIMD feature detection, see features.hpp

See also

hasVectorInstructions() for advanced vector processing

ArchitectureInfo::hasSimdSupport() for member function version

algorithm selection

template<typename T>
void process_array(const T* input, T* output, size_t size) {
    if constexpr (trlc::platform::hasSimdSupport()) {
        // Use SIMD optimized version
        process_array_simd(input, output, size);
    } else {
        // Use scalar fallback
        process_array_scalar(input, output, size);
    }
}

Since

1.0.0