x

conferenceimageresize

Poster

The course focuses on highly scalable parallel computing for helping research users to effectively use GPU and multiple processing resources on UBDA platform. It covers various topics from concepts of parallel computing, a suite of packages of RAPIDS – using the NVIDIA GPUs, cuDF, cuPy, conversions of cuDF and cuPy, conducting analysis of bootstraps and machine learning with GPUs, and DASK. Some Python programs and exercises will be conducted with Google Colaboratory.  There will also be demos to illustrate the use of the UBDA system. Please find the arrangements below.

Course

Scalable Parallel Computing (Poster)

Date

7 Jul 2021 to 21 Jul 2021

Session activities

  • Parallel Computing with RAPIDS (1st online session: 7 Jul 2021, 2:30 p.m.- 4:00 p.m.)
  • Bootstraps and Machine Learning with GPU (2nd online session: 14 Jul 2021, 2:30 p.m. - 4:00 p.m.)
  • Overview UBDA Services and DASK (3rd online session: 21 Jul 2021, 2:30 p.m. - 4:00 p.m.)
  • Consultation Hours (Face-to-face sessions, P505, 12 Jul, 19 Jul & 26 Jul 2021, 11:00 a.m. - 12:00 p.m.)

Registration/Quota

All staff and research students

60 (first come, first served via online registration)

Confirmation emails will be sent to the registrants on 5 Jul 2021

Course Topics

  • Parallel Computing with RAPIDS
    • Introduction to Parallel and GPU Computing
    • RAPIDS – cuDF and cuPy
    • Profiling and conversions of cuDF and cuPy
  • Bootstraps and Machine Learning with GPU
    • Bootstrap basics
    • Bootstrap regression models and dependent data
    • cuML – Regression, Clustering, Dimensionality reduction
  • Overview UBDA Services and DASK
    • Introduction to UBDA Facility
    • cuSignal, BlazingSQL, and DASK

Should you have any enquiry, please contact Dr Vincent Ng (Email: This email address is being protected from spambots. You need JavaScript enabled to view it.) or Dr Heung Wong (Email: This email address is being protected from spambots. You need JavaScript enabled to view it.)