Link Search Menu Expand Document

CSE 234: Data Systems for Machine Learning

Instructor: Hao Zhang, UC San Diego, Winter 2025

Announcements

Week 6 Announcements

Feb 11 · 0 min read
  • We have just released PA2, due March 4th 2025. Please start early!

Week 1

Jan 7
1 Introduction
Slides
Survey Beginning of Quarter Survey (Due: End of Week 2 - 1/19)
Readings (Due 1/14)
Jan 9
2 Basics: Modern DL, computational graph, frameworks
SlidesRecordingScribe Note

Week 2

Jan 14
3 Basics: autodiff, ML system architecture overview
SlidesRecordingScribe Note
Readings (Due 1/21)
Jan 16
4 Tensor format, matmul deep dive, accelerators
SlidesRecordingScribe Note

Week 3

Jan 21
5 GPUs and CUDA
SlidesRecordingScribe Note
Readings (Due 1/28)
Jan 23
6 GPU matmul, operator compilation
SlidesRecordingScribe Note

Week 4

Jan 28
7 Triton, graph optimization and compilation
SlidesRecordingScribe Note
Readings (Due 2/4)
Jan 30
8 Memory
SlidesRecordingScribe Note

Week 5

Feb 4
9 Quantization
SlidesRecordingScribe Note

Readings (Due 2/11)

Feb 6
10 Guest Lecture by Tianqi Chen
SlidesRecordingScribe Note

Week 6

Feb 11
11 Parallelization
SlidesRecordingScribe Note
Readings (Due 2/18)
Feb 13
12 Parallelization - 2, collective communication
SlidesRecordingScribe Note

Week 7

Feb 18
13 Parallelization - 3, data, inter- and intra-op parallelism
SlidesRecordingScribe Note
Readings (Due 2/25)
Feb 20
14 Guest Lecture by Hector Liu
SlidesRecording; Scribe Note

Week 8

Feb 25
15 Parallelization - 4
SlidesRecordingScribe Note
Readings (Due 3/4)
Feb 27
16 LLM - 1: transformers, attention
SlidesRecordingScribe Note

Week 9

Mar 4
17 LLM - 2: Scaling law, MoE
SlidesRecordingScribe Note
Readings (Due 3/11)
Mar 6
18 Guest Lecture by Hongyang Zhang
SlidesRecordingScribe Note

Week 10

Mar 11
19 LLM - 3: Continuous batching, paged attention, Disaggreged prefill and decoding
SlidesRecordingScribe Notes
Reading:
Mar 13
20 LLM - 4: flash attention, deepseek-v3 review, ending words
SlidesRecordingScribe Notes