- 3-4 years experience in C/C++ development and debugging on multicore systems with expertise in CPU profiling and performance optimization
- Analyze and optimize performance bottlenecks in CPU software stack for AMD EPYC processors on multi-core, multi-node systems
Job Summary
This is a great opportunity to work as part of a highly regarded team to deliver leading-edge solutions.
Responsibilities
- Performance engineering of the CPU software stack for AMD EPYC server processors: find and fix performance and scalability bottlenecks of C/C++ applications and libraries on multi-core, multi-node systems (HPC, MPI, data center).
Skills
Must have
- 3-4 years of relevant experience
- C/C++ development and debugging on multicore systems
- Solid data structures and algorithms
- Performance analysis and optimization of bottlenecks
- Understanding of cache subsystem, ISA and CPU pipeline
- CPU profiling tools
- Experience: 3 to 4 years relevant; onsite in Bangalore 5 days, two F2F rounds on the same day
Nice to have
- HPC / MPI application performance
- x86 or ARM optimization
- Intel MKL, linear algebra, FFT
- x86/ARM assembly