Responsible for performance bench-marking and characterization of SOC at system level– Power and Performance Bench-marking: Characterization of system using compute, memory and GPU benchmarks and system analysis using collected data correlating with estimated KPIs.– Memory benchmarking with BW and Latency analysis, memory fabric, NOC and DDR sub-system profiling.– Bench-marking and characterization on Android and Windows based platforms including UX performance measurement– Develop and own SOC/IP performance models and analysis tools in C++ and Python, collaborating with Micro architects to identify and address performance bottlenecks and work across teams with a growth mindset to deliver world-class SOC products.

Designing, Development and Deployment of device certification framework.– Edge Device Certification Suite: Led the planning and execution for framework comprise of smoke and functional test cases to execute feature validation of base platform in customer environment and provide report/certification.

Performance Measurements and Bench-marking.– Developed benchmark as a service framework, streamlining performance evaluation processes.– Collaborated with product management to define bench-marking requirements, to align with product goals and performance metrics.– Key skills: competitive bench-marking and reporting, Kubernetes, Microservices, REST API, CPU bench-marking (speccpu17, Passmark, Geekbench), AI bench-marking (OpenVino, TensorFlow, PyTorch, oneDNN) Infrastructure-as-a-service– Led a platform validation for an infrastructure-as-a-service program and developed a validation micro-service and deployments.– Key skills: Microservices, Kubernetes (with NFD), CV workload, Gstreamer, DLstreamer, AI (object detection and classification) Use-case Validation and Video Analytics– Led the execution of performance, use-case & functional validation on client and server programs. Developed and executed a video analytics workload to assess CPU/GPU performance.– Key skills: Computer vision, video processing, performance optimization and benchmarking. Gstreamer, DLstreamer, AI topologies (Yolov3, Mobilenet, Resnet34, Resnet50) Functional Validation and Automation– Implemented automated testing procedures, to reduce testing time by 90%.– Optimization of platform firmware performance through iterative testing, code profiling, and collaboration with development teams.– Key skills: Firmware profiling and optimization, FUSA, ISO26262, Platform HW validation, ITP (in target probe), memory array tests

Functional validation– Conducted Functional Safety (FUSA) validation for ISO 26262 compliance, ensuring platform adheres to safety standards for autonomous driving applications.