Published onAugust 26, 2025Speeding up PyTorch inference by 87% on Apple devices with AI-generated Metal kernelsKernel-OptimizationPerformanceApple-SiliconOur lab investigated whether frontier models can write optimized GPU kernels for Apple devices to speed up inference. We found that they can: our AI-generated Metal kernels were 1.87x faster across 215 PyTorch modules.