Facebook pixel

Senior Staff Performance Architect
Posted on 11/1/2022
INACTIVE
Locations
San Francisco, CA, USA
Experience Level
Entry
Junior
Mid
Senior
Expert
Desired Skills
CUDA
C/C++/C#
Linux/Unix
Requirements
  • 10+ years of experience in the field of design and development of accelerated HPC systems, and/or system (real-time) performance analysis and optimization
  • Extensive experience with programming different HW accelerators (e.g. GPUs, NNAs, DSPs)
  • Extensive experience with HPC development ecosystem, e.g. profiling and optimizing software through parallelization techniques for complex multi-device compute nodes
  • Extensive experience with linux-based performance profiling and debugging tools (for CPUs and accelerators)
  • Strong expertise in writing production quality C++ code and setting standards for code quality across engineering teams
  • Experience in driving technical strategy and vision for engineering teams and organizations
  • Leadership experience for planning and execution of cross-functional initiatives and projects
  • In-depth understanding of software development lifecycle (SDLC) and best practices - CI/CD, coding, debugging, optimization, testing, integration and deployment
  • PhD in CS/CE/EE, or equivalent, in industry experience
Responsibilities
  • Define strategic vision and execution roadmap that innovates on performance optimizations of heterogeneous AV workloads on HW-constrained compute platforms
  • Partner with cross-functional teams to define the vision and plan to scale the performance testing and learning cycle, from characterization and analysis, to diagnosis and optimization, and to gating and prevention
  • Collaborate effectively with leaders across AV hardware and software orgs to drive consensus through data-driven studies and tradeoffs
  • Define execution processes that strive for streamlined engineering development with robust quality and excellence
  • Lead engineering teams through the SDLC lifecycle, communicate effectively with senior leadership and optimize investment with impact in mind
  • Closely follow industry and academic developments in the SOTA for performance optimization, real-time execution and ML hardware domains and adopt technology that is the best fit for Cruise's needs
Desired Qualifications
  • Experience with ASIC program and design process (from a compute software perspective)
  • Experience with performance tooling, modeling, simulation and architecture design for modern HW microarchitectures
  • Experience with modern parallel programming: GPU programming (CUDA, OpenCL), SIMD (avx, neon/SVE), multi-process and multi-threaded designs
  • Familiarity with HW vendors' deep learning stacks (e.g., cuDNN, cuBLAS, AMD MIOpen, TensorRT, OpenVino, ARM Compute Library, etc)
  • Experience with model optimization techniques like pruning, quantization, NAS, etc
  • Experience with real-time scheduling frameworks
  • Experience with open-source deep learning stacks (TVM, XLA, etc)
Cruise

1,001-5,000 employees

Self-driving car service
Company Overview
Cruise's is building self-driving vehicles to improve life in our cities. The company makes autonomous, sustainable, self-driving EVs.
Benefits
  • Flexible vacation
  • Paid holidays
  • Paid parental leave
  • Fertility & family expansion benefits
  • 401k matching program
  • Monthly social events
  • Community volunteering programs
  • Healthy meals & snacks for onsite employees
  • Quarterly offsites & working retreats
  • Monthly wellness stipend
  • Mental health support
  • Professional development programs
  • On-site gym in SF HQ
  • Commuter benefits for onsite employees
  • Medical, dental & vision coverage
Company Core Values
  • Stay safe
  • Stay focused
  • Own it
  • Seek truth
  • Work together
  • Be humble