SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Graphcore, part of the SoftBank Group, is a leading innovator in AI compute hardware and infrastructure. This Senior Power and Performance Validation Engineer role reports to senior leadership within Architecture and Validation and drives validation strategy and execution for advanced AI compute silicon and systems.
You will be responsible for leading power, thermal, and performance validation activities across pre-silicon and post-silicon environments to ensure products meet efficiency, reliability, and scalability expectations. The role requires strong technical expertise and collaboration across multiple engineering disciplines to deliver robust validation methodologies, scalable automation frameworks, and actionable performance insights.
Key responsibilities include:
- Define validation strategies for power, thermal, and performance characterization of AI compute silicon and platforms
- Contribute to planning and execution of validation activities across multiple projects and milestones
- Develop comprehensive validation plans covering functional, stress, workload, and corner-case scenarios
- Lead post-silicon bring-up and characterization activities for power and performance validation
- Drive validation of CPU, memory, interconnect, and high-speed I/O subsystems under complex workload conditions
- Develop scalable automation frameworks, regression infrastructure, and reporting tools using Python
- Design and execute benchmark workloads, parameter sweeps, and performance experiments to identify optimization opportunities
- Collaborate with architecture, RTL, firmware, software, and systems teams to debug and resolve complex technical issues
- Define validation metrics, pass/fail criteria, and reporting methodologies to ensure repeatable and high-quality analysis
- Guide development of custom workload generators and micro-benchmarks where required
- Analyze power, thermal, and performance data to identify bottlenecks and recommend improvements
- Contribute to continuous improvement of validation processes, tooling, and engineering practices
- Communicate technical findings, project status, risks, and recommendations clearly to stakeholders and engineering leadership
- Support knowledge sharing and contribute to continuous improvement across the validation organization
The Power and Performance Validation team sits within the Architecture and Validation organization and is responsible for validating the performance, efficiency, and thermal behavior of Graphcore silicon and systems. The team supports the full product lifecycle, from early architectural modeling through to first silicon bring-up, characterization, and production readiness. Engineers work closely with cross-functional teams globally to debug complex issues, optimize workloads, and continuously improve validation infrastructure and methodologies.
REQUIREMENTS:
Essential:
- Strong experience in power and performance validation, silicon characterization, or system performance engineering
- Deep understanding of modern SoC architecture, including CPU, memory, interconnect, and high-speed I/O technologies
- Strong Linux systems knowledge and low-level performance analysis experience
- Strong Python programming skills for automation, orchestration, and data analysis
- Experience with benchmarking and profiling tools such as stress-ng, fio, perf, iperf, or equivalent technologies
- Experience debugging complex hardware and software interactions
- Ability to define structured validation methodologies, workload models, and test strategies
- Strong analytical skills with the ability to interpret large datasets and identify system bottlenecks
- Strong communication and cross-functional collaboration skills
- Ability to work independently on complex technical tasks while collaborating effectively across teams
Desirable:
- Experience with AI accelerator, GPU, or high-performance compute architectures
- Experience with pre-silicon modeling or simulation environments
- Knowledge of power management technologies and silicon characterization methodologies
- Programming experience in C/C++ for low-level system or benchmark development
- Familiarity with hardware instrumentation and telemetry systems
- Experience working with high core-count or large-scale compute systems
- Experience contributing to shared validation infrastructure and engineering best practices
