I created an advanced Smoothed Particle Hydrodynamics (SPH) simulation utilizing C++ and OpenGL, meticulously engineered to model and visualize realistic fluid dynamics in real-time. This project showcases my expertise in sophisticated programming techniques, parallel computing, and high-performance graphics rendering.

Key Features

• Fully Multithreaded Architecture:
  • Physics Computations: Leveraging C++’s multithreading capabilities, I parallelized core physics calculations to efficiently manage and simulate thousands of particles concurrently. This ensures smooth and responsive simulations even under heavy computational loads.
  • Spatial Hash Creation and Update: Developed a multithreaded spatial hashing system to swiftly identify and manage particle neighbors. This significantly reduces computational overhead and enhances the scalability of the simulation.
• Optimized Neighbor Searching with Spatial Hashing:
  • Implemented a spatial hash to partition the simulation space, enabling rapid retrieval of neighboring particles. This approach reduces the complexity of neighbor searches from O(n^2)to approximately O(n), where n is the number of particles. Consequently, this facilitates efficient large-scale fluid simulations.
• High-Performance Mathematical Computations with SIMD:
  • Incorporated SIMD (Single Instruction, Multiple Data) instructions to accelerate critical mathematical operations within SPH algorithms, such as density and pressure calculations. This optimization harnesses data-level parallelism, substantially boosting computational throughput and overall simulation performance.
• Efficient Rendering with Instanced OpenGL:
  • Employed instanced rendering in OpenGL to draw thousands of particles with minimal draw calls. This technique not only enhances rendering efficiency but also maintains high frame rates, providing a seamless visual experience even with a large number of particles.