PhD Position F/M Predictable cache memories for Real-Time Embedded Multicore Systems

1000scholars; Villeneuve-d'Ascq; On-site; EUR 25,000 - 35,000; Full time

Job summary

1000scholars invites applications for a PhD student position focusing on memory placement in caches for multi-core embedded real-time systems. The candidate will study state-of-the-art algorithms and static analysis of binaries, then develop a novel memory-placement algorithm.

The role requires very good C/C++ expertise, and knowledge of compilers, linkers, and optional areas such as Real-Time scheduling and WCET analysis. English language proficiency (B2) is expected.

Qualifications

• Strong C/C++ skills with experience in compilers and linkers.
• Optional: knowledge of Real-Time scheduling, multicore systems, and WCET analysis.

Responsibilities

• Study the state of the art in memory placement algorithms.
• Study static analysis of program binaries to determine memory locations of data and code.
• Develop a novel memory placement algorithm.

Skills

C/C++

Compiler

Linker

Real-Time scheduling

Multicore systems

Worst-case execution time analysis

Job description

Context

The PhD student will work in the context of the ANR PRCI CoDaLi project, with Inria Rennes and the RPTU (https://rptu.de/).

The general objective of the CoCoDaLi project is to design and implement offline and runtime techniques that allow precise control cache contents in multi-core architectures for embedded real-time systems.

The main idea is to take advantage of the content lifespan of cache blocks, in order to:

• prevent the hardware replacement algorithm from removing useful cache blocks;
• reduce the number of cache errors by placing the application memory in the appropriate cache lines.

The PhD student will spend some time in Kaiserslautern and Rennes to collaborate with the project partners.

Assignment

Missions:

1. Study the state of the art in algorithms for memory placement in cache
2. study algorithms for the static analysis of program binary code, so to obtain the memory location of data an code
3. develop a novel algorithm for memory placement.

For a better knowledge of the proposed research subject :

Scientific references of the research team are available at the followings URLs :

• https://www.fil.univ-lille.fr/~lipari/research/phd-call.html
• www.cristal.univ-lille.fr/equipes/sycomores

Main activities

Research and development of algorithms and code for optimizing the memory placement.

Skills

Technical skills :

We require a very good extertise in C/C++ language, compiler and linker

Optional expertise :

• Real-Time scheduling, and scheduling analysis
• Multicore systems,
• worst-case execution time analysis.

Languages :

English B2