NVIDIA Portfolio Optimization
GPU-accelerated portfolio optimization toolkit
Description
With many assets, Mean-CVaR optimization and backtests take hours on CPUs. This NVIDIA blueprint builds, backtests and scales portfolio optimization on GPUs with cuOpt, an order of magnitude faster.
It supports Mean-Variance and Mean-CVaR with CUDA-X data science libraries, delivered as Jupyter notebooks.
GPU solver:cuOpt.
Models:Mean-Variance and Mean-CVaR.
Backtesting:Validate strategies.
Notebooks:Ready to run.
It supports Mean-Variance and Mean-CVaR with CUDA-X data science libraries, delivered as Jupyter notebooks.
Features
GPU solver:cuOpt.
Models:Mean-Variance and Mean-CVaR.
Backtesting:Validate strategies.
Notebooks:Ready to run.
