If you are ready for the challenges of research at the intersection of programming languages, compiler construction, and High-Performance Computing and look for an opportunity to work on a challenging PhD with a strong link to applicability in practical tools, feel free to contact me.

Here a list of my PhD students, current and former:

  • Hans Viessmann: “On Effective GPU Programming through Compiler Generated Code”, expected 2021, Radboud University, Nijmegen, Netherlands.
    Hans has shown how sophisticated code generation techniques play a crucial role when aiming at performance portability of HPC codes that run on GPUs. He currently works at Radboud University, Nijmegen, Netherlands.
  • Max Baird: “A Checkpointing Mechanism for GPU Intensive HPC Applications”, expected 2021, Heriot-Watt University, UK.
    Max has tackled the problem of resilience in GPU-accelerated clusters with long running kernels. He invented a novel, light-weight way of interrupting GPU kernels and he integrated his work into the resilience framework FTI. He currently works as Lecturer (Assistant Professor) at University of Guyana, Guyana.
  • Stuart Gordon: “Semi-Automatic Dynamic Program Adaptation for Platform-Independent High-Performance Computing on Heterogenous Systems”, expected 2021, Heriot-Watt University, Edinburgh, UK.
    Stuart has looked at dynamically adapting multi-threaded runtime systems. In particular, he looked into dynamic scaling of thread-pools and redistribution of hard-ware pinning for improved spacial locality. He currently works at Texas Instruments in Munich, Germany.
  • Nilesh Karavadera: “On the Design and Implementation of a Light-Weight Parallel Execution Layer With Power Optimisation for Stream Processing Networks on Distributed Memory”, December 2016, University of Hertfordshire, UK.
    Nilesh developed a runtime system for stream processing on Intel's Single Chip Cloud computer.
  • Artem Shinkarov: “Data Layout Types: a type-based approach to automatic data layout transformations for improved SIMD vectorisation”, May 2015, Heriot-Watt University, UK.
    Artem has pioneered a type-driven approach towards rearranging the layout of multi-dimensional array structures in memory to facilitate improved auto-vectorisation. He currently works as Postdoc at Herriot-Watt University, Edinburgh, UK.
  • Santanu Dash: “Adaptive Constraint Solving for Information Flow Analysis”, December2014, University of Hertfordshire.
    Santanu worked on static analysis of programs using annotated type systems and he proposed novel algorithms for atomic bound constraint solving for Information Flow Analysis. He currently works as Lecturer (Assistant Professor) at the University of Surrey, UK.
  • Jing Guo: “Fully Automated Transformation of Hardware-Agnostic, Data-Parallel Programs for Host-Driven Executions on GPUs”, November 2012, University of Hertfordshire, UK.
    Jing has pioneered fully automated code generation for mapping functional array programs in SaC into CUDA programs that run on GPUs. This work includes several optimisations for minimizing the amount of memory transfers between CPU and GPU. He currently works as freelance programmer in Shanghai, China.
  • Frank Penczek: “Static Guarantees for Coordinated Components: A Statically Typed Composition Model for Stream-Processing Networks”, November 2012, University of Hertfordshire, UK.
    Frank has developed and implemented a type system for safe compositions of streaming networks with tagged data flow interfaces and flow inheritance. He currently works at Intel, Ulm, Germany.
  • Stephan Herhut: “Multi-Facet Subtyping as a Means to Generic Programming with Static Type Guarantees”, June 2010, University of Hertfordshire, UK.
    Stephan has developed a sub typing based approach towards static type guarantees through compiler driven partial evaluation. He currently works at Google, Munich, Germany.
  • Dietmar Kreye: “A Compiler Backend for Generic Programming with Arrays”, March 2003, University of Kiel, Germany.
    Dietmar has invented a code generator for HPC codes on multi-dimensional arrays that supports a hierarchy of array representations with decreasing levels of static information. He currently works at MSG Nexingsure AG, Hamburg, Germany.
  • Clemens Grelck: “Implicit Shared Memory Multiprocessor Support for the Functional Programming Language SaC – Single Assignment C”, January 2001, University of Kiel, Germany.
    Clemens has pioneered highly efficient code generation and runtime systems for data parallel executions on multicore systems. This includes a novel, lock-free memory manager for multi-threaded executions. He currently works as Associate Professor at the University of Amsterdam, Netherlands.

This topic does not exist yet

You've followed a link to a topic that doesn't exist yet. If permissions allow, you may create it by clicking on Create this page.

  • phd_students.txt
  • Last modified: 2021/03/23 07:56
  • by sbscholz