AlexFiliakov/Ergodic-Insurance-Limits

AI Summary: This issue proposes to significantly speed up the optimization process by leveraging GPUs. It involves batching multiple Monte Carlo simulations for evaluating the optimization objective function and parallelizing multi-start optimization runs across GPU streams. The goal is to achieve a 30-100x speedup for the entire optimization process.

AI Summary: This issue proposes to significantly accelerate the Monte Carlo simulation engine by offloading its core computationally intensive loop to a GPU. The current sequential simulation of thousands of paths will be transformed into a massively parallel GPU kernel, aiming for a 50-200x speedup. This involves modifying existing Python files and introducing GPU-specific configurations and tests.