IIT Bombay Efficient AI IIT Bombay · Autumn 2026

CS 6106 Autumn 2026 · Dept. of Computer Science & Engineering

Efficient AI

Algorithmic and hardware-aware approaches to making modern AI cheaper to train and faster to serve — pruning, quantization, parameter-efficient fine-tuning, speculative decoding, KV-cache compression, and the systems that carry them.

Slot TBA · Venue TBA · 6 credits · jump to schedule

Updates

Course staff

Prof. Aditya Desai

Instructor

instructor@cse.iitb.ac.in
KReSIT Building, Room KR 208 · office hours by appointment

TA Name 1

Teaching assistant

ta1@cse.iitb.ac.in
Office hours TBA

TA Name 2

Teaching assistant

ta2@cse.iitb.ac.in
Office hours TBA

TA Name 3

Teaching assistant

ta3@cse.iitb.ac.in
Office hours TBA

Logistics

Lectures
Slot TBA — two sessions a week. Recordings, if any, will be linked from the schedule.
Venue
TBA
Credits
6
Prerequisites
A first course in machine learning. Familiarity with computer architecture helps but is not required. Others by consent of the instructor.
Grading
  • Course project40%
  • Paper presentations25%
  • Assignments20%
  • Reviews & participation15%
Indicative — final weights are announced in the first lecture.
Communication
Announcements on this page and Moodle. The discussion-forum link will be shared in the first lecture.

Contents

Five arcs, ordered roughly as the semester runs.

  1. I. Compression & quantization

    Model compression · knowledge distillation · quantization · hardware support for quantization

  2. II. Sparsity & pruning

    Pruning · the lottery ticket hypothesis · hardware support for sparsity · sparse training

  3. III. Architectures & adaptation

    Neural architecture search · low-rank structure (butterfly & monarch matrices) · PEFT — LoRA and variants, SketchTune · elastic parameter memory · hardware-aware hashing

  4. IV. Efficient attention & inference

    Sparse attention (prefill & decode, training-time) · speculative decoding · KV-cache compression · semantic prompt caching · attention alternatives — SSMs & linear attention

  5. V. Systems & hardware

    IR techniques in LLM inference (SLIDE and successors) · upcoming GPU IR languages · context extension & compaction

Schedule

Dates are tentative until the academic calendar is finalised. Slides and readings land in the resources column after each week.

No. Date Description Resources
1Jul 28 Why efficient AI — the cost of training and serving; roofline model, memory hierarchy; course logistics. MIT 6.5940 — the closest course elsewhere; good introductory slides.
2Aug 04 Model compression & knowledge distillation — the compression landscape; teacher–student training.
3Aug 11 Pruning & sparsity — magnitude and structured pruning; the lottery ticket hypothesis. Frankle ’23 — the thesis behind the lottery ticket hypothesis.
4Aug 18 Hardware support for sparsity — sparse formats, structured sparsity on accelerators. Sky-Light — Berkeley project; see its reading list.
5Aug 25 Quantization I — post-training quantization and quantization-aware training.
6Sep 01 Quantization II — hardware support; low-precision formats (INT4, FP8) and kernels.
7Sep 08 Elastic parameter memory & hardware-aware hashing — parameter sharing at extreme scale. Desai ’24, Chen ’20 — the two theses this lecture draws on.
Sep 14 – 20 · mid-semester examinations, no lecture
8Sep 22 Neural architecture search — search spaces, hardware-aware objectives, once-for-all networks.
9Sep 29 Low-rank structure — butterfly and monarch matrices; structured matrices as compute savers. Quang ’23 — hardware-aware algorithms, including monarch matrices.
10Oct 06 Parameter-efficient fine-tuning — LoRA and its variants, SketchTune.
11Oct 13 Sparse attention — inference-time and training-time sparsity; prefill versus decode.
12Oct 20 Speculative decoding — draft models, verification, acceptance schemes. SpecDec papers — living list of speculative-decoding papers.
13Oct 27 KV-cache compression & semantic prompt caching — eviction, quantized caches, cache reuse across requests. kvpress — NVIDIA’s library and paper list for KV-cache compression.
14Nov 03 Attention alternatives & systems — SSMs and linear attention; IR techniques (SLIDE), GPU IR languages; context extension & compaction. Wrap-up.

References

Foundational theses and living paper lists. Weekly readings are linked from the schedule.

  1. Desai, A. (2024). Elastic Parameter Memory for Efficient Machine Learning. Doctoral dissertation, Rice University.
  2. Frankle, J. (2023). The Lottery Ticket Hypothesis: On Sparse, Trainable Neural Networks. Doctoral dissertation, Massachusetts Institute of Technology.
  3. Quang, T. D. P. (2023). Hardware-Aware Algorithms for Efficient Machine Learning. Doctoral dissertation, Stanford University.
  4. Chen, B. (2020). Locality Sensitive Sampling for Extreme-Scale Optimization and Deep Learning. Doctoral dissertation, Rice University.
  5. TinyML and Efficient Deep Learning Computing — MIT 6.5940, a close cousin of this course.
  6. Sky-Light, UC Berkeley — see the list of papers.
  7. Speculative decoding paper list — curated on GitHub.
  8. NVIDIA kvpress — KV-cache compression papers and code.