Server Silicon Performance Analysis Engineer

AMD AMD · Semiconductors · Bangalore, India · Engineering

AMD is seeking a Server Performance Analysis Engineer in Bangalore, India, to analyze and optimize the performance of EPYC based servers for cloud, enterprise, HPC, and AI/ML workloads. The role involves characterizing workloads, analyzing performance data, identifying bottlenecks, and working with software teams on optimizations. Requires strong computer architecture expertise and experience with performance profiling tools.

What you'd actually do

  1. Characterize the properties of Cloud, Enterprise, HPC/ML/AI workloads
  2. Analyze performance profiling data, traces and simulation data to quantify and isolate performance bottlenecks on current generation hardware.
  3. Capture traces, run simulations of future designs and highlight key performance sensitivities and bottlenecks to microarchitecture development teams
  4. Work with relevant software teams to drive optimizations based on performance analysis into key server software components (linux, compiler, Java, KVM, etc).
  5. Perform competitive analysis across AMD and other hardware platforms.

Skills

Required

  • B.Tech + 4 years in Performance Engineering, or M.Tech + 2 years, or PhD + 0 years, in Computer Science/Engineering or related field of study

Nice to have

  • Strong foundation in any two of CPU core, memory, storage and I/O subsystems
  • Power user of performance profiling and hardware performance counter tools
  • Familiarity with complex cloud or enterprise software such as database, data analytics, machine learning, and relevant benchmarks such as SPECCPU, SPECjbb15, etc.
  • Familiarity with operating systems concepts (e.g. scheduling, memory management, paging, interrupt handling, etc
  • Good programming skills in any language (C/C++)
  • Data analysis skills (Excel or Python)