CompArch: Computer Architecture

This course presents the architecture of microprocessor-based systems, from the internals of the processors themselves to the main peripherals, that altogether make a complete computing machine, capable of running operating systems like GNU/Linux, Android, Windows, iOS…

The focus is on the interactions between software and hardware: thanks to lectures, exercises and laboratory sessions students understand what happens between the source code of a software application and its execution on a real hardware computer system, between software concepts and the hardware components that implement them in the end.

Another important aspect is the relationships between performance, safety and security. The course presents various techniques used by computer system designers to improve the performance of their machines, and the impact of these techniques in terms of safety (functional correctness) and security (sensitivity to various types of attacks).

Topics:

History of computer systems
Types of microprocessors, from the simplest to the most powerful
Representation of data in computer systems
Instruction set architectures
Interactions between hardware and software
Pipelined architectures, memory hierarchy and caches, branch prediction, hardware support for operating systems, peripherals…
Interactions between performance, safety and security
Labs on RISC-V architectures: observation of a processor during the execution of software applications, understanding of its interactions with its environment, design of a tiny operating system

Learning outcomes:

Understand the relationships between software concepts (functionality, performance, parallelism, security…) and the underlying hardware machines (processors, peripherals…).
Discover the most important techniques used by the processor and/or computer vendors to improve the performance of their products (pipelining, caches, branch prediction…) and the related issues (pipeline hazards, cache coherence, speculative execution…).
Understand the necessary hardware support for operating systems: processor modes, Memory Management Units (MMU), timers, interrupts, atomic operations…
Become familiar with the classical trade-offs between performance, cost, and complexity in computer systems.

Attending the lectures is not mandatory but highly recommended: there is no real textbook and the provided material does not contain the entirety of what is discussed in the classroom.

Attendance to the lab sessions is mandatory, the reports and the produced source code account for 1/3 of the overall mark. It is perfectly possible, and even recommended, to work in cooperation with others but lab reports and source codes are 100% personal works. Identical reports or source codes will be considered as void.

A two hours written exam with documents accounts for the remaining 2/3 of the final grade. The questions and problems correspond to all labs and lectures that have been given. Connected devices (laptops, smart phones, tablets…) are not allowed. Example past exams:

2024
2023

Recommendations for the lab sessions

All labs are distributed and managed using git and GitLab.

If you never used the EURECOM GitLab instance, first visit it and log in using your EURECOM credentials (LDAP). This will activate your GitLab account.
Visit the ca project and request access (top right of web page, "More actions" menu).
As soon as you will be granted access you will receive a confirmation email. Visit the ca project again, read the home page and follow the instructions.
All lab instructions are given for a GNU/Linux environment (Debian, Ubuntu…) If you work on your personal computer under another Operating System you will have to adapt the instructions.

You read one of these books? Or another that is not in the list and should? Please drop me note and give me some feedback.

The most relevant parts of the Computer Organization and Design, RISC-V Edition (The Hardware Software Interface) book for this course (do not forget the corresponding exercises at the end of each chapter) are:

Section 1.3 Below Your Program
Section 1.4 Under the Covers
Section 1.6 Performance
Sections 2.1 to 2.10 of Chapter 2 Instructions: Language of the Computer
Sections 3.1 and 3.2 of Chapter 3 Arithmetic for Computers
Sections 4.1 to 4.9 of Chapter 4 The processor
Sections 5.1 to 5.4, 5.8 and 5.10 of Chapter 5 Large and Fast: Exploiting Memory Hierarchy

Plus, optionally:

Section Historical Perspective and Further Reading of Chapters 1, 2 and 3
Section 2.13 A C Sort Example to Put it All Together
Section 3.5 Floating Point