Portrait of Gennesaret Tjusila

Gennesaret Tjusila

I’m a PhD student in the iol.INT group at the Zuse Institute Berlin, researching GPU-accelerated algorithms for mixed-integer programming.

Berlin, Germanytjusila (at) zib.de

About me

Background

I completed both my bachelor’s and master’s degrees in mathematics at Technische Universität Berlin. Although my degrees are in mathematics, I’m equally interested in algorithm design and the low-level details that determine performance on modern hardware.

Outside of research, I enjoy reading about programming and computer architecture, reading thrillers, and playing chess.

My research

GPU-accelerated algorithms for MIP

My research focuses on three aspects of GPU-accelerated mixed-integer programming: adapting existing algorithms to exploit GPU parallelism, developing new algorithms designed specifically for GPUs, and integrating GPU-based methods with other components of a MIP solver.

Research

Selected publications

  1. CHAP: A Hybrid GPU-CPU Heuristic for MIP

    Gennesaret Kharistio Tjusila, Alexander Hoen, Nils-Christian Kempke, Gioni Mexi, Timo Berthold, Ambros Gleixner, Thorsten Koch, Sebastian Pokutta

    arXiv preprint · 2026

  2. How many clues to give? A bilevel formulation for the minimum Sudoku clue problem

    Gennesaret Tjusila, Mathieu Besançon, Mark Turner, Thorsten Koch

    Operations Research Letters · 2024

Full academic profile at ZIB — IOL Lab

Awards

First Place, 2026 Land–Doig MIP Competition

Theme: GPU-Accelerated Primal Heuristics for MIP

May 2026

Experience

Research Assistant — Zuse Institute Berlin

Oct 2025 – present
  • PhD research on GPU-accelerated algorithms for mixed-integer programming.

Working Student — Zuse Institute Berlin

May 2025 – Sep 2025
  • Implemented V-polyhedral disjunctive cuts in SCIP via SCIP.jl as part of the master’s thesis.
  • Contributed to the SCIP interfaces for Python (PySCIPOpt) and Julia (SCIP.jl).

Working Student, Cloud Engineer — IVU Traffic Technologies

Aug 2022 – Apr 2025
  • Managed cloud infrastructure for IVU SaaS applications (Linux, AWS, Terraform, Oracle, JBoss) and built Python tooling to integrate IVU systems with customer systems.

Student Project Assistant — IVU Traffic Technologies

Nov 2021 – Aug 2022
  • Participated in the system workstream for the IVU × Lithuanian Railway project (AWS, cloud security, cloud performance, authentication) as the customer’s point of contact.

Education

M.Sc. Mathematics — Technische Universität Berlin

Oct 2022 – Jul 2025 · Grade: 1.1∗

Thesis: “Implementation of V-Polyhedral Disjunctive Cuts in SCIP.”

B.Sc. Mathematics — Technische Universität Berlin

Apr 2019 – Aug 2022 · Grade: 1.7∗

Thesis: “How Many Clues to Give? A Bilevel Formulation for the Minimum Sudoku Clue Problem.” Published in Operations Research Letters (2024).

∗ On the German grading scale, 1.0 is best and 4.0 is worst.