Congratulations to Winners!1st: UTAR Cuda Coders
3rd: UniKL BSE
- This competition is open to everyone.
- This competition is FREE.
- Each team can be composed of up to 3 members.
- The final code benchmarking step will be executed on a MIMOS server, but development will be limited to the participants own computers. Configurations of the MIMOS server will be announced at the end in order for participants to customize the optimization parameters for the compiler.
- A few sample data sets similar to the ones to be used for final benchmarking will be given to the participants for use in their code development.
- The winner shall be the team that produces an error-free shortest path in the fastest time and shows good understanding of GPU architecture in code design.
The CUDA Programming Challenge 2014 is a coding competition that welcomes all types of CUDA programs to solve a shortest path route selection problem. The problem of finding a path between two vertices or nodes, such that the sum of the weights of all the edges in between is minimized, is a problem that can be solved in polynomial time, but involves a huge amount of computation.
The algorithms used to solve this problem can be applied to real-world applications such as finding the shortest routes between two points such as it is done in Google Maps or Waze, and in networking where packets are often routed through the shortest paths to minimize delays. The goal of this year’s challenge is to develop an algorithm to find the shortest path is the least amount of time.
To be added.
May: Sample data sets availability
Sept 15: Judging and presentation (Deadline: Judging session is extended to 30 Sept)
Oct: Prize giving