This Job Vacancy has Expired!

Tensor Computing Performance Optimization Engineer

Consol Partners

Posted on Sep 8, 2021 by Consol Partners

Paris, France
IT
Immediate Start
€300 - €400 Daily
Contract/Project


Tensor Computing Performance Optimization Engineer
Paris, France
6-12 month contract (extendable)

ConSol Partners client are looking for someone to join the Distributed Parallel Software Lab in their research center and to support its efforts in solving scientific and industrial problems regarding high-performance artificial intelligence computing on heterogeneous architectures.


The main goals of Distributed Parallel Software Lab are:
- Develop automatic parallelization and optimization techniques to improve our clients products.
- Develop and maintain our compilation tool-chain.
- Long-term research in the domain of optimizing compilation and parallelism.

Candidate
The candidate should have PhD in Computer Science with (junior or senior) experience in tensor computing optimization, program optimization and parallelization.


The following skills and experiences are highly desirable:
Excellent programming skills on AI compiler with code-generation toolchain eg, PPCG, TVM, TC, XLA etc.
Excellent code-performance optimization on GPGPU including cuda, opencl, ptx and CPU including SIMD, openmp.
Excellent programming skills in C, C++, and Python, versioning using Git, CI, etc.
Significant knowledge and understanding of compiler construction, compiler optimizations, parallel programming, program parallelization, heterogeneous computing.
Good knowledge of linear algebra, mathematical optimization, computer architecture; familiarity with AI algorithms and software is a plus.
Ability to assist research thanks to experience in participating or driving R&D projects as a student, academic, entrepreneur, or in a large company; publication and achievement record in fields such as compiler, language, parallelism, architecture, AI are highly appreciated.
Familiarity with Open Source development and communities.




Reference: 1313768842

Set up alerts to get notified of new vacancies.