Cpu Performance Architect

Intel Intel · Semiconductors · Bangalore, India

This role focuses on the architecture of CPUs, specifically on improving methodologies and infrastructure for power and performance modeling, analysis, and workload bring-up for next-generation client products. The individual will research and drive ideas to enhance SoC power and performance modeling, collaborate with design teams, and analyze bottlenecks to propose solutions.

What you'd actually do

  1. Research and improve methodologies and infrastructure for power and performance modeling, analysis and for workload bring up, analysis.
  2. Own and drive research ideas to improve Soc power and performance modeling including modeling these ideas in power and performance models and collaborating with wider arch and design teams to productize ideas.
  3. Model proposed micro-arch in power and performance models, analyze bottlenecks and propose solutions for them.
  4. Work with validation teams to ensure power and performance features bring expected benefits in the product

Skills

Required

  • Computer architecture expertise
  • C++ or object-oriented programming
  • C++/System C TLM modeling
  • Performance modeling of complex SoCs
  • Analysis of bottlenecks
  • Collaboration with design and validation teams

Nice to have

  • Traditional data structures and algorithms
  • Deep learning techniques
  • Scripting
  • RTL validation environments
  • Bringing up/integrating performance/functional simulators
  • Workload analysis

What the JD emphasized

  • Computer arch expertise
  • strong understanding of either traditional data structures and algorithms or deep learning techniques
  • Strong C++ or object oriented programming skills
  • Experience in C++/System C TLM modeling
  • Experience in specific sub-systems of Soc like cache, memory or fabric
  • handling performance model/performance analysis of complex Socs with latency/bandwidth/power expectations