Project Title: Dynamic Graphs on the GPU

Name: Jimmy Lord

E-Mail: jimmylord@gmail.com

My project this semester will involve both learning the CUDA language for GPUs and the implementation and verification of the results of existing techniques developed for the parallelisation of dynamic graphs using a hash table on the GPU.

Startup Paper(s):

• Dynamic Graphs on the GPU

Deliverables:

• Literature Review (PDF file created from LATEX template)
• Presentation Outline (PDF file created in google slides)
• Slide Presentation (PDF File created in google docs) incl. Question Sheet
• Final Paper (PDF file created from LATEX template)
• Code and Data (zip file containing source code and all reasonably sized relevant data)

Relevant References:

• A Dynamic Hash Table for the GPU
• STINGER: Spatio-Temporal Interaction Networks and Graphs (STING) Extensible Representation
• Static graph challenge on GPU
• Solving path problems on the GPU
• Hornet: An Efficient Data Structure for Dynamic Sparse Graphs and Matrices on GPUs
• STINGER: High performance data structure for streaming graphs
• cuSTINGER: Supporting dynamic graph algorithms for GPUs
• networkrepository.com/misc.php
• Accelerating dynamic graph analytics on GPUs
• faimGraph: High Performance Management of Fully-Dynamic Graphs Under Tight Memory Constraints on the GPU
• Autonomous, independent management of dynamic graphs on GPUs