Staff Engineer, Product Validation Engineering

Job Summary

We are looking for an experienced Post-Silicon Performance Measurement Benchmarking Engineer to drive performance characterization, bottleneck analysis, and optimization of next-generation SoCs. The candidate will work closely with Architecture, Design, Firmware, Driver, Validation, and Software teams to ensure silicon achieves its targeted performance, power, and efficiency.

The candidate brings strong expertise in performance measurement methodologies, hardware performance counters, profiling tools, benchmarking, and root-cause analysis across CPU, GPU, NPU/AI accelerators, memory subsystem, and I/O.

Key Responsibilities

• Develop and execute post-silicon performance validation plans.
• Measure and analyze silicon performance against architectural KPIs.
• Build automated benchmarking and performance characterization frameworks.
• Analyze hardware performance counters (PMU/PMC) to identify bottlenecks.
• Tune firmware, drivers, BIOS, OS, and platform configurations for optimal performance.
• Perform workload characterization across AI, Storage, Networking, Multimedia, and Compute workloads.
• Correlate pre-silicon simulation results with post-silicon measurements.
• Debug CPU, Memory, Cache, Interconnect, PCIe, DDR, Storage, and AI accelerator performance issues.
• Analyze bandwidth, latency, throughput, utilization, and QoS metrics.
• Work with architecture teams to recommend micro-architectural improvements.
• Create dashboards and performance trend reports.
• Develop Python automation for benchmark execution, log collection, and report generation.
• Collaborate with firmware teams to optimize scheduling, cache policies, memory allocation, and interrupt handling.
• Support silicon bring-up and characterization activities.
• Exposure to AI tools and methodologies.

Required Technical Skills

Performance Measurement

• Performance benchmarking methodologies
• Throughput and latency analysis
• Memory bandwidth measurement
• Cache utilization analysis
• Hardware counter analysis
• Power-performance tradeoff analysis
• Thermal throttling characterization

Hardware Knowledge

• ARM/x86 Architecture
• Cache hierarchy
• DDR/LPDDR subsystem
• PCIe
• USB
• Ethernet
• NoC (Network-on-Chip)

Performance Tools

• Linux perf
• PMU/PMC
• VTune
• Perfetto
• Trace32
• JTAG
• Lauterbach
• ETM/PTM Trace
• ftrace
• eBPF (Preferred)

Programming

• Python
• C/C++
• Shell scripting

Operating Systems

• Linux Kernel
• Android (Preferred)
• Embedded Linux

Benchmarking Experience

Experience with one or more of the following:

• MLPerf
• SPEC CPU
• FIO
• IOZone
• STREAM
• Geekbench
• CoreMark
• AI inference benchmarks
• Storage performance benchmarks
• Multimedia benchmarks

Preferred Skills

• Experience with AI accelerators (NPU/DSP/GPU)
• Performance tuning for TensorFlow, PyTorch, ONNX Runtime, TensorRT
• Understanding of compiler optimizations
• Firmware optimization
• Linux kernel performance tuning
• DVFS and power management
• Memory subsystem optimization
• Storage performance optimization
• Knowledge of virtualization and hypervisors

Soft Skills

• Strong analytical and debugging skills
• Excellent problem-solving ability
• Cross-functional collaboration
• Effective communication and presentation skills
• Ability to drive performance optimization initiatives independently

Education

Bachelor's or Master's degree in Computer Science, Electronics, Electrical Engineering, or a related field.

Disclaimer: This job posting has been aggregated from external sources. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.