🚀 High Performance Computing (HPC) Researcher focused on large-scale FEM, sparse systems, and distributed compute. I build systems that don’t fit in RAM—and then make them faster anyway.
A modular, performance-first C++17 utility library designed for scientific computing and HPC workflows.
Core Capabilities
- Thread-safe logging (low overhead, multi-thread aware)
- Scope-based timing + profiling instrumentation
- Throughput benchmarking (compute + memory)
- Resource tracking (memory, execution, scaling)
- Built for MPI/OpenMP environments
- Clean integration with Meson build system
Designed to support real-world HPC workloads, not toy examples.
- Sparse Linear Algebra (CSR / COO / BSR / ELL / VBR)
- Krylov Subspace Methods (GMRES, iterative solvers)
- Finite Element Methods (MFEM)
- Adaptive Mesh Refinement (AMR)
- Distributed Systems & Multi-node Scaling
- Slurm-based HPC environments
- Memory-bound performance optimization
💡 Current emphasis: Bridging numerical methods with real-world scalability constraints (memory, bandwidth, communication overhead).
- Scaling FEM simulations to millions–billions of DOFs
- Sparse matrix performance tradeoffs
- MPI + OpenMP hybrid parallelism
- HPC profiling, bottlenecks, and memory diagnostics
- Linux-based research environments
Started in Manufacturing Engineering → pivoted into High Performance Computing, numerical methods, and large-scale simulation.
- C++ (Modern + Legacy) — 9
- C — 9
- Java (Spring Boot) — 8
- Assembly — 5
Strong focus on performance-critical, low-level systems
- MPI / OpenMP — 8
- MFEM / FEM pipelines — 7
- Slurm / Cluster tooling — 7
- TypeScript / React / Node — 8
- HTML / CSS — 8–9
- SQL (Postgres) — 8
- Shell (Linux) — 6
- Python — 4 (used pragmatically, not religiously)
- R — 4
- Odin
- Zig
- Ladder Logic (Allen-Bradley PLCs)
- CNC Programming (G/M Code)
- NodeJS, Cypress SDET
- C++ on Large Clusters
If your program fits in RAM, you’re playing a different game.
I focus on:
- Problems where memory is the bottleneck
- Systems where performance = survival
- Code that runs on clusters, not laptops