Markus Püschel
Computer Science
ETH Zürich




short CV


the pub

How to Write Fast Numerical Code 263-2300 (ETH, CS)

Basic Information

  • Course number: 263-2300, 6 credits
  • Spring 2017, lectures: M 10:15-12:00, HG D3.2; Th 9:15-10:00 CAB G51; occasional substitute lectures: W 13:15-15:00 HG D3.2
  • Instructor: Markus Püschel (CAB H69.3, pueschel at inf, 2-7303)
    • Alen Stojanov (CAB 81.2, astojanov at inf)
    • Georg Ofenbeck (CAB H65, ofgeorg at inf)
    • Gagandeep Singh (CAB H66, gsingh at inf)
    • Only for project supervision: Daniele Spampinato (CAB H65, daniele.spampinato at inf)
  • Office hours:
    • Alen Stojanov: M 3-4pm, CAB H81.2
    • Geoerg Ofenbeck: T 3-4pm, CAB H65
    • Gagandeep Singh: W 3-4pm, CAB H66
    • Markus Püschel: Th 3-4pm, CAB H69.3
  • Mailing lists:
    • For technical questions: (emails to this address go to the lecturer and all TAs)
    • Forum to find project partner: (emails go to all students who have no partner yet and to Alen & Daniele)

Course Description

The fast evolution and increasing complexity of computing platforms pose a major challenge for developers of high performance software for engineering, science, and consumer applications: it becomes increasingly harder to harness the available computing power. Straightforward implementations may lose as much as one or two orders of magnitude in performance. On the other hand, creating optimal implementations requires the developer to have an understanding of algorithms, capabilities and limitations of compilers, and the target platform's architecture and microarchitecture.

This interdisciplinary course aims to give the student an understanding of performance and introduces foundations and state-of-the-art techniques in high performance software development using important functionality such as linear algebra algorithms, transforms, filters, and others as examples. The course will focus on optimizing for the memory hierarchy and special instruction sets, thus complementing courses on parallel programming. Much of the material is based on recent research.

Further, a general strategy for performance analysis and optimization is introduced that the students will apply in group projects that accompany the course. Finally, the course will introduce the students to the recent field of automatic performance tuning.

Prerequisites: solid C programming skills, matrix algebra, Master student or above

Topics Covered

  • Algorithm analysis: Problem versus algorithm, complexity and cost (asymptotic, exact, measured), cost analysis
  • Computer architecture (a software point of view): architecture and microarchitecture, memory hierarchy, special instruction sets
  • Compilers: strengths, limitations, how to use
  • Performance optimization: guide to benchmarking, finding hotspots, code analysis, performance optimization techniques (for memory hierarchy and vector instruction extensions); these techniques are studied using the examples in the next bullet
  • Numerical functionality studied in detail (complexity, algorithms, how to write highest performance code): linear algebra kernels, transforms, filters, sparse linear algebra, others, your research project
  • Automatic Performance Tuning: ATLAS, LAPACK, BeBOP, FFTW, SPIRAL, others

Goals of this Course

  • Obtain an understanding of runtime performance and how to reason about it
  • Learn a guideline how to write fast numerical code and apply it in homeworks and your research project
  • Understand the connection between algorithms, implementations, and computer architecture

Background Material

Academic Integrity

All homeworks in this course are single-student homeworks. The work must be all your own. Do not copy any parts of any of the homeworks from anyone including the web. Do not look at other students' code, papers, or exams. Do not make any parts of your homework available to anyone, and make sure noone can read your files. The university policies on academic integrity will be applied rigorously.

We will be using the Moss system to detect software plagiarism. This system is amazingly good, because it understands the programming language in question (C, in our case).

It is not considered cheating to clarify vague points in the assignments or textbook, or to give help or receive help in using the computer systems, compilers, debuggers, profilers, or other facilities.


  • 40% research project
    • Topic: Very fast, ideally adaptive implementation of a numerical problem
    • Team up in pairs
    • March 6: find a partner, find a problem or I give you one (tip: look at the prior courses linked above for examples)
    • Complete "milestones" during semester and enter them into the online check list
    • Write 6 page standard conference paper (template will be provided)
    • Give short presentation end of semester
  • 25% midterm
  • 35% homework
    • Exercises on algorithms analysis
    • Implementation exercises
      • study the effect of program optimizations, compilers, special instructions, etc.
      • write and submit C code & create runtime/performance plots
    • Some templates will be provided
    • All homeworks are single-student homeworks
  • There is no final Exam

Research Project

  • All projects have to be registered at This site is also used later for updates.
  • How it works:
    • Weeks without homeworks should be used to work on the project
    • You select a numerical problem and create a correct (verified) implementation in C
    • You determine the arithmetic cost, measure the runtime and performance
    • You profile the implementation to find the parts in which most the runtime spent
    • Focussing on these you apply various optimization techniques from this class
    • You repeat the previous steps to create various versions with (hopefully) continuously better runtime
    • You write a paper about your work and give a presentation
  • Paper:
    • Maximal 6 pages (hard limit), conference style, template and instructions below
    • Everybody reads this: report.pdf
    • For latex use: (start with reading the README file)
    • For Word (discouraged) use this: report-word.doc
    • Due date: Friday, June 16 (as final-report.pdf in your svn)
  • Presentation
    • Last week of classes
    • Template (the use is totally optional) and some guidelines (ppt is 2007 and later): presentation-template.pptx , presentation-template.pdf
    • The order will be determined randomly right before class
    • Who talks will be determined randomly right before class
  • Projects (each one has a supervisor shown in brackets):
    1. Eliza W, Hui Z, Jingwei T, Yiqing Z: Ant-inspired edge detection
    2. Alberto M, Andreas B, Marc R F, Marko P: t-Distributed stochastic neighbor embedding
    3. Gaurav P, Jonathan Me, Luca A, Marc J F: Marching cubes
    4. Alexey K, Jonathan Ma, Jonathan R: Locality sensitive hashing
    5. David H, Magdalena K, Pirmin V, Stephanie C: PatchMatch algorithm
    6. Anton P, Jinank J, Manuel R, Milan P: Fractal compression
    7. Marcin D, Marius F, Patrick S, Thomas S: Fast ray tracing for TSDFs
    8. Ankit S, Jingxuan H, Lidia CdF, Gokula S: Binary convolutional neural network
    9. Andreas H, David S, Simon F, Till E: A robust descriptor for line matching
    10. Benjamin G, Cristina C, Frédéric L, Saurav S: Latent Dirichlet Allocation
    11. David S, Hasan H, Kaan K, Konstantin T: Ray tracing
    12. Fabio L, Henri R, Mahamadou B, Roffler C: Medial axis transform
    13. Ladislas JdN, Luca C, Matteo T, Yeyao Z: Quantized neural networks
    14. Aaker OE, Bian W: Matrix multiplication over GF(2)
    15. Lars B, Michal W, Samuel M: Non-linearly coupled elliptic BVPs
    16. Daan N, Julien L, Linus H, Stefano P: GP-UCB
    17. Acharya D, Li C, Victor C, Yu-chen T: Online dictionary learning for sparse coding

Tips & Tricks


26.4., 13:15-15:00. Location: Hönggerberg! HPH G 3


Late policy: No deadline extensions, but you have 3 late days. You can use at most 2 on one homework. For example, submitting 7 hours late costs one late day.

We will be using Moodle for the homeworks.

Lectures (including pdfs)

1 M 20.02. Course motivation, overview, organization link    
2 Th 23.02. Cost analysis and performance link    
3 M 27.02. Intel Haswell architecture and microarchitecture, memory- and compute-bound link   Intel Haswell, Intel software optimization manual, Agner Fog's instruction tables
4 W 01.03. Instruction-level parallelism, compiler limitations link, link    
5 M 06.03 Benchmarking, SIMD (SSE, AVX) overview link    
6 M 13.03. SIMD (SSE, AVX) intrinsics link   Intel intrinsics guide
7 Th 16.03. SIMD (SSE, AVX)      
8 W 22.03. Locality, caches link  
9 Th 23.03. Caches, analysis of blocked MMM      
10 M 27.03. Roofline model link link roofline paper
11 Th 30.03. Linear algebra libraries, LAPACK, BLAS, ATLAS link    
12 M 03.04. Fast MMM (model-based ATLAS)   link fast MMM paper
13 Th 07.04. Fast MMM, register renaming      
14 M 10.04. Virtual memory system   link  
15 W 12.04. Memory bound computations, sparse MVM link    
16 M 24.04.. sparse MVM, linear transforms link  
  W 26.04. Exam in HPH G3, on Hönggerberg      
  Th 27.04. cancelled