Assignment

Assignment

• 06/16, Hadoop Workshop. Deadline: 5 pm, jun 22, 2020.
  • Please follow the tutorial and finish two workshops.
  • Run your code on server (no online judge) and demo your result to corresponding TA.
• 06/09, CUDA Workshop. Deadline: 5 pm, jun 15, 2020.
  • Solve the problem 6. Labeling String (CUDA)
    • [Basic] You can use Thrust API (doxygen) to easily implement an parallel algorithm in GPU.
    • [Advanced] Implement your algorithm without Thrust API and improve performance.
    • Small Talk: CUDA Something.
  • Solve the problem 7. Matrix Operation (CUDA)
    • Using multiple devices to speedup this computing with at least two GPU devices.
• 06/02, OpenCL Advanced Workshop. Deadline: 5 pm, Jun 08, 2020.
  • Solve problem 5.Cooperative/Separate Dot Product
  • The GPU spec of new judgegirl: GPU: NVidia GTX 1080Ti ( X2 ).
  • The GPU spec of old judgegirl: GPU NVidia GTX 1080, NVidia GTX 980Ti, NVidia Tesla K20c.
  • Note that the GPU workload of Morris might be heavier than Miwa's, so you could try to submit your answer to old judgegirl while getting TLE on new judgegirl.
• 05/26, OpenCL Intermediate Workshop. Deadline: 5 pm, Jun 01, 2020.
  • Solve problem 4. Very Fast Dot Product (OpenCL)
• 05/17, OpenCL Basic Workshop. Deadline: 5 pm, May 25, 2020.
  • Solve problem 1. OpenCL Build Program Debug and 2. Dot Product (OpenCL)
• 05/12, OpenMP Advanced Workshop. Deadline: 5 pm, May 18, 2020.
  • Solve problem 10081. Fast Game of Life and 10203. Fast Matrix Chain Multiplication Basic.
• 05/05, OpenMP intermediate Workshop. Deadline: 5 pm, May 11, 2020.
  • Solve the problem 10094. Fast 0/1 Knapsack Problem and 10202. Fast N-Queens Basic.
• 04/28, OpenMP basic Workshop. Deadline: 5 pm, May 04, 2020.
  • Solve the problem 10022. Fast Matrix Multiplication and 10025. Fast Image Match.
• 04/21/2020 In this assignment you need to do the followings. The deadline is 2 pm, Aril 28, 2020.
  • Get the code from https://github.com/joaquintides/usingstdcpp2015 You need to use git clone to do this.
  • Compile the code.
  • Run the code.
  • Measure the performance of the code with perf and valgrind. Note that you need to compare the performance among several ways to access the data, so you may need to modify the code to isolate the effect. For example, you may need to modify the SOA-AOS comparison code into two programs, one for SOA and one for AOS, so that you can measure their performance (e,g., cache misses) separately.
  • Explain the performance difference you observed.
  • You need to start from scratch and demonstrate the process to the corresponding TA.
• 04/14 lab1: 10084. The dealine is 5pm, 04/17/2020. Please demo your homework to the corresponding TA.
• 04/07/2020 The assignment 3 The deadline is 5pm, 04/10/2020, grade.
• 04/05/2020 The grade of assignment 2 The answer is at the bottom.
• 03/24/2020 Parallel Programming Principle The deadline is 5pm, 03/27/2020. grade
• 03/31/2020 Parallel algorithm (1)