PhD student in theoretical physics specializing in computational many-body methods (Tensor Networks, Exact Diagonalization, Monte Carlo). My work focuses on moiré systems and large-scale numerical simulations, and I enjoy building high-performance Julia libraries for scientific computing, quantitative modeling, and industry applications.
- Researcher at the University of Würzburg working on moiré materials and strongly correlated electrons using tensor-network and Monte Carlo methods.
- Interested in software development at the intersection of computational physics, scientific software engineering, and quantitative modeling-especially where numerical methods, HPC, and rigorous, data-driven decisions matter.”
- Maintaining and developing many-body methods and generic Julia packages in the open-source organization ManyBodyLab:
- Infinite tensor-network algorithms (e.g. iDMRG, MPO-compression) based on ITensors.jl and TensorKit.jl
- MPI-parallelized tensor-network algorithms in MPSKitParallel.jl
- High-performance exact diagonalization tools in native Julia as well as in DiagHamInterface.jl
- Allocation-free Monte Carlo algorithms for quantum many-body systems in WaveMC.jl (based on the Carlo.jl framework)
- and many more (MPILargeCounts.jl, PeriodicArrays.jl, ...)
- Designing modular backends to support multiple tensor-network libraries, improving flexibility and reuse across projects and reducing duplication in downstream codes.
- Contributing to MPSKit.jl, abstracting code and improving performance and maintainability.
- Implemented a low-latency C++ matching engine (price-time priority) achieving 15+ million orders/s on real and synthetic LOBSTER data using a single CPU core
- Reduced p99 latency to <200 ns by replacing pointer-based linked lists with a pre-allocated contiguous memory pool (intrusive list), minimizing cache misses and allocator overhead during bursty workloads
- Designed a deterministic testing suite using fuzzing and edge cases to verify correctness against a reference model
- Engineered a binary parser for the NASDAQ TotalView-ITCH 5.0 protocol to efficiently ingest raw market data
- Email: andreas.feuerpfeil@manybodylab.com
- LinkedIn: Andreas Feuerpfeil
- Website: manybodylab.com