Registration for this course is open until Wednesday, 18.12.2024 23:59.

News

Reminder: Register for Project Git Repo

Written on 28.10.24 (last change on 28.10.24) by Tejas Harith

As mentioned in the previous lecture (whose recording will soon be uploaded), please remember to use the CMS "registration" > "Project Git Repo" so that we can launch a project repository for you with all the basic material. We will go over the project in the first project session which is this… Read more

As mentioned in the previous lecture (whose recording will soon be uploaded), please remember to use the CMS "registration" > "Project Git Repo" so that we can launch a project repository for you with all the basic material. We will go over the project in the first project session which is this week! You will get an email when the repository has been generated. 

Reminder: Lecture Today in Room 029 Building E1 5

Written on 25.10.24 (last change on 25.10.24) by Tejas Harith

Walk into the MPI SWS building, all the way to the back. Take a right, and walk to the door at the end of the hall, past the bathrooms.

Lecture zoom link: https://shlink.mpi-sws.org/os-24-lectures

Reminder: No Project Session this week!

Written on 23.10.24 (last change on 23.10.24) by Tejas Harith

Project sessions start from next Wednesday 30.10. See you at lecture this Friday :). Note the location: E 1 5, the MPI-SWS building Rm 029. Starts at 12.15.

Reminder: Kickoff Lecture Today!

Written on 18.10.24 by Tejas Harith

Welcome to the inaugural lecture of SwiSH W24/25!

 

We look forward to seeing you all in Rm 024, MPI-INF (E1 4). We will start at quarter past 12 and will go over course logistics as well as an introduction to Post-Moore systems!

Systems with Specialized Hardware


As we move into a computational era beyond the scaling of Moore's law, our computing systems are forced to evolve to keep pace with the ever-growing demands of nascent fields such as that of AI/ML, scientific computing, and blockchain. The unprecedented explosion in computational requirements has given rise to bespoke computing platforms which tradeoff flexibility, for excellence in a specific application. Complementing concepts introduced in W23/24's hit seminar: Accelerating Applications with Specialized Hardware, this advanced lecture will introduce you to the world that is designing and building those specialised systems.
 
 
The goal of this advanced lecture is to provide students with the tools and expertise to identify and characterize the weaknesses of a system design, empirically. Then to implement and evaluate their own improvement based on a characterized weakness. Successful students will be prepared for undertaking potential Master Thesis projects on these topics (e.g. with The OS Group)

Format


course type / weekly hours: 
      2 h lectures 
    + 2 h project (this is a hands-on session) 
    + 1 h tutorials
    = 5 h (weekly) 

total workload: 
       75 h of classes 
    + 195 h private study 
    = 270 h (= 9 ECTS)

This course is comprised of lectures and project sessions. The lectures will introduce key concepts of accelerator design and project sessions will walk students through using open source tools to evaluate a rudimentary specialised system which they will later improve on. At the end of the course, students will work on an open-ended task: designing, implementing, and evaluating an improvement of their own choosing. This task will be accompanied with a light report explaining design choices and results. 

Contents


The course is divided into three sections.

  • Part A: Intro to Post Moore Acceleration

    • Lectures: fundamentals of why accelerators matter in modern computing systems
    • Labs: prerequisites for designing accelerators in this course: Verilog overview, tools summary
  • Part B: System Bottleneck & Resource Evaluation

    • Lectures: theory behind efficient accelerators, and how integration can either make or break a complete system
    • Labs: hands-on experience with tools used to quantify performance & utilisation
  • Part C: Design Tradeoffs & Optimisation

    • Open-ended design exploration and improvement project.
    • Lectures: techniques for shifting an accelerator's design along one or more axes, in order to achieve specific target metrics.
    • Labs: Open office hours for guidance on the project

The projects will use the SimBricks full system simulation framework, providing in-depth control and visibility of the complete system, without the need for special hardware.

Kick-Off Meeting: Fri Oct 18, 12-14, E1 4 Room 024 MPI-INF Building

Lectures: Fri, 12-14, E1 5 Room 029 MPI-SWS Building
Projects: Wed, 12-14, E1 5 Room 029 MPI-SWS Building

This advanced lecture is organised by the OS Group at MPI-SWS.


Lectures

Part Lecture Project Session Lecture Date Project Session Date MS Resources
A Intro to Post-Moore Systems    18.10      
  System Design Terminologies Verilog & Project Primer 25.10 30.10 1 Dark Silicon
  Holiday: All Saints Intro to Hardware Design Tools -- 06.11 1  
  Accelerator Classification  Devices, Drivers & Interfacing 08.11 13.11 1  
B Hardware Resource Utilisation Evaluating Synthesis Output 15.11 20.11 2  
  Software Resource Utilisation DMA 22.11 27.11 2  
  Simbricks: Full System Evaluation Using Simbricks 29.11 04.12 3  
  Specialising for Workloads Memory Utilisation, Buffer Fill 06.12 11.12 3  
C Parallelism & Batching  Discuss Project Ideas 13.12 18.12 Fin  
  Pipelining Additional Simbricks/Blueprint Features 20.12 08.01 Fin  
  Resource Sharing From Design to ASIC 10.01 15.01 Fin  
  Dynamic Parallelism & Beyond CMOS Special Topics: TBA 17.01 22.01 Fin  
  Guest Lecture: TBA Special Topics: TBA 24.01   Fin  
  TBA -- 31.01   Fin  
End -- -- 07.02      

Milestones

MS Topic Release Due Pts
1 Measuring Baselines 30.10 19.11 10
2 Design Tradeoffs 20.11   10
3 Orchestrating Data      10
Final System Improvement     30

Disclaimer

The projects in this advanced lecture are relatively implementation-heavy compared to what you may be used to from other courses. You should also expect to engage in some hairy debugging (e.g. segmentation faults, and memory corruption) typical of low-level systems code. Please plan accordingly.

 

Requirements


This advanced lecture is open to senior Bachelor, Masters, and doctoral students. Ideally, students should have already taken courses on system architecture and operating systems. Bachelor students must have passed the basic courses on Programming 2 or equivalent. Proficiency in C/C++ programming (including low-level aspects such as pointers and memory management) and UNIX development tools (e.g., shell, make, gcc, gdb) is strictly required to pass this course. Basic Verilog knowledge will also be required for the projects, but this can be acquired or refreshed as part of the project. Please contact Antoine with questions about these requirements.

The language of the advanced lecture is English. All meetings and communication with the course staff will be conducted exclusively in English.

Privacy Policy | Legal Notice
If you encounter technical problems, please contact the administrators.