SmartFlow v1.0.1

Summary

This release adds an update to the SLURM job submission parameters:

• Added --gres=gpu:0 to srun options.

Details

This modification is necessary to run on the CPU nodes of the Leonardo Booster and DCGP partitions. Without specifying --gres=gpu:0, jobs may fail to allocate the proper resources on these partitions.

SmartFlow v1.0.0 – First Stable Release

Initial release of smartflow, a modular framework for multi-agent deep reinforcement learning (DRL) applied to turbulence modeling, flow control, and numerical algorithm development.

✨ Highlights

• 🧠 Multi-Agent DRL Framework
  Designed for advanced control and modeling in turbulent flows.

• 🌊 CFD Solver Integration

  • ✅ Supports Fortran 90/95/2008 solvers
  • 🔜 Support for C++ solvers coming soon

• ⚡ GPU Acceleration
  Compatible with GPU-accelerated CFD solvers for high-performance training and evaluation.

• 🖥️ Deployment Flexibility
  Runs on both standalone workstations and HPC clusters.

---

smartflow v1.0.0 is ideal for researchers and engineers aiming to bring AI into CFD workflows, enabling robust, scalable, and intelligent modeling strategies.