Director, Sustaining Engineering - Graphcore

OpenTalent; Milpitas (CA); Hybrid; USD 140,000 - 190,000; Full time

Job summary

Graphcore is seeking a Principal Engineer, Power Engineering to define the architecture of power delivery from grid to chip. You will collaborate with hardware, firmware, and datacenter teams to ensure performance, reliability, and scalability across the AI infrastructure.

The role is based in the US (Austin or San Jose) with hybrid/on-site options and potential remote work within the US. Candidates should drive improvements in power efficiency and system-level design.

Responsibilities

• Provide leadership and design for power delivery from datacenter power drop to systems and chips.
• Establish power roadmap; align vendor technologies with our system engineering roadmap and drive innovation for power components.
• Define rack-level power shelves hardware/firmware specs; design for resiliency.
• Set datacenter power policies (buffer/oversubscription, job scheduling by power) and monitor power quality; resolve telemetry for policies.
• Ensure IT power robustness during grid disturbances and compatibility with grid requirements.
• Guide board/system power regulation (VRM/TLDR/POL), improve efficiency, consider manufacturing constraints.
• Advise manufacturers/ODM on testing and validation procedures for power hardware/firmware.
• Develop and test hardware.

Job description

Location: Milpitas, CA - Please note this a hybrid role with a minimum of 3 days onsite.

About Us

Graphcoreis a globally recognised leader in Artificial Intelligence computing systems. The company designs advanced semiconductors and data centre hardware that provide the specialised processing power needed to drive AI innovation, while delivering the efficiency required to support its broader adoption.

As part of the SoftBank Group,Graphcoreis a member of an elite family of companies responsible for some of the world’s most transformative technologies. We are opening a new AI Engineering Campus in Austin, which will playa central roleinGraphcore'swork building the future of AI computing.

Job Overview:

Graphcoreis seeking a Principal Engineer, Power Engineering to aid in defining the architecture of the power delivery components from grid to chip. You will work closely with multiple cross-functional teams, including hardware engineers, firmware developers, and datacenteroperations, to ensure the infrastructure we build meets customer requirements as well asattainingstandards of performance, reliability, and scalability.

This role will be based in Austin or San Jose and could potentially be performed onsite or remotely within the US.

Responsibilities:

• Provide thought leadership, planning, and design for power delivery from datacenterpower drop through delivery to systems & chips
• Establish power roadmap. Work with vendors to ensure their planned future technologies are pulled into our system engineering roadmap AND ensure that our (system engineering)requiredtechnological innovation is being driven into their (power vendor) roadmap. Work to improve reliability and efficiency of power engineeringcomponentover time.
• Rack-level power components. Competency with specification of hardware and firmware for rack-level power shelves. Competency with design for resiliency.
• Set power policies in the datacenterfor: power buffer/oversubscription, scheduling of jobs based on power, and monitoring (power quality faults) / accommodation faults. Resolve telemetry and data needed to support power policy algorithms.
• Ensure robustness of IT power in the presence of grid disturbances and make sure that IT power loads are compatible with grid power requirements (“good grid power citizen”)
• Board and system-level power regulation components. Understand and able to apply current VRM, TLDR, and POL technologies. Help to developnew technologiesto increase efficiency. Understand manufacturing capabilities and limitations ofnew technology. Drive
• Provide guidance to the power equipment manufacturers, ODM and engineering teams on the required testing and validation procedures for power hardware and firmware components.
• Develop and test har