COURSES

DTU Courses in parallel programming

• 02614 High-Performance Computing (with CUDA content)
• 02616 Large-scale modelling

PhD summer school in Nodal Discontinuous Galerkin Finite element Methods 2012

Course material (2011):

• Web page
• Course flyer

Workshops

• August 17, 2011. Workshop on "Easy, Effective, Efficient: GPU Programming in Pyhton with PyOpenCL and PyCUDA" will be given by Andreas Kloeckner, Courant Institute of Mathematical Sciences, New York University, USA.
• Lecture slides 1
• Lecture slides 2
• Problem set 1
• Problem set 2

GPU Computing Today and Tomorrow

August 18, 2011. Workshop on "GPU Computing Today and Tomorrow" with short research highlights from GPUlab and partners. Invitation.

Slides:

• Allan P. Engsig-Karup: Very fast simulation of nonlinear water waves in very large numerical wave tanks on affordable graphics cards
• Stefan L. Glimberg: A High Performance GPU-based Framework for PDE Prototyping
• Nicolai Fog Gade-Nielsen: Accelerating interior point methods with GPUs for smart grid systems
• Hans Henrik Sørensen: Tuning of BLAS level 1 and 2
• Thomas Schrøder: A molecular dynamics code optimized for GPUs

PhD School in Iterative Methods for Large Linear Systems

Course material (2011):

• Web page
• ITMAN and DCAMM flyer

PhD School in Scientific GPU Computing

Course material (2011):

• Web page
• ITMAN and DCAMM flyer

Examples of student projects completed as a part of the PhD School in 2011

• CUDA and OpenCL implementations of 1D Flexible Finite Dierence Methods
• Similarity Calculations in clustering of streaming data
• Performance of parallel sum reduction in CUDA and OpenCL
• Cupic Spline Interpolation on the GPU
• Non-bonded energy calculation on the GPU
• LDDKBM integration on GPUs
• An application of GPU programming to block triangular matrices
• GPU-based Forward Projection in Cone-Beam CT
• GPU accelerated calculation of the effective electron-phonon density of states
• GPU implementation of Maximum Likelihood Localization of Multiple Sources by Alternating Projection
• Non-linear Least Squares Problems
• FFT strategy exploration for hybrid solver on GPU

Course material (2010):

• Web page
• ITMAN and DCAMM flyer

Examples of student projects completed as a part of the PhD School in 2010

• Flexible-order finite difference computations
• The Poisson problem
• Implementation of the Lattice Boltzmann Method on GPU
• Segmented line extraction
• Large-scale primal SVM training in CUDA
• Gaussian process regression using GPUs
• Performing measurements for comparing 3D observations with a generative human model
• Sparse matrix-vector techniques for finite difference operations using CUDA
• Realtime-ish ray tracer in CUDA
• Sparse octree computation on the GPU
• CudaVox: a voxel terrain renderer
• Implementing a feature detector in CUDA
• k-means clustering on a GPU