PHAS0102: Techniques of High-Performance Computing (2022)
Timetable
- Lecture: Fridays 10–11, Cruciform Building B404 - LT2
- Practical class (group 1): Mondays 10–11, Euston Road 222 G01
- Practical class (group 2): Mondays 11–12, Chadwick Building 2.18
- Virtual drop-in hour: Wednesdays 11:30–12:30, link on Moodle
Assessment
Note that these assignments are from 2022
This course will be assessed through four coursework assignments:
- Coursework assignment 1 (20%)
- Deadline: Thursday 20 October 5pm
- View assignment 1
- Coursework assignment 2 (20%)
- Deadline: Thursday 3 November 5pm
- View assignment 2
- Coursework assignment 3 (30%)
- Deadline: Thursday 1 December 5pm
- View assignment 3
- Coursework assignment 4 (30%)
- Deadline: Thursday 15 December 5pm
- View assignment 4
Lecture Notes & slides
The lecture notes for this course can be found at tbetcke.github.io/hpc_lecture_notes.
These slides were used in lectures:
- Lecture 0: introduction
- Lecture 1: Numpy
- Lecture 2: Numba
- Lecture 3: GPUs
- Lecture 4: Sparse matrices
- Lecture 5: GMRES
- Lecture 6: CG
- Lecture 7: LU
- Lecture 8: PETSc and multigrid
- Lecture 9: Time-dependent problems and wider context
Practical class tasks
- Class 1: Numpy, timing and matplotlib
- Class 2: Numba
- Class 3: GPUs and Cuda
- Class 4: Sparse storage formats
- Class 5: GMRES
- Class 6: CG
- Class 7: LU for a tridiagonal matrix
Getting help
Questions about course content can be asked on the PHAS0102 (22/23) Zoom chat. To join this chat:
- Open the Zoom app (not the website)
- Click "SSO" then sign on via ucl.zoom.us
- Click the "Chat" tab
- Search for PHAS0102 (22/23)