F

FluidStack

Builds and runs GPU clusters for AI labs

Controls and Firmware Engineer - Power Electronics

Power Systems EngineeringPosted on Nov 12
Full-time$210k - $270k /yrIn-person

About the Job

Requirements

  • Experience writing and bench-debugging control firmware that operates a power converter at full power.
  • Experience controlling resonant or dual active bridge converters, or balancing series-connected cells.
  • Experience building FPGA or DSP protection logic that trips within microseconds and fails safely.
  • Experience using hardware-in-the-loop or model-based design to identify bugs before they reach hardware.
  • Experience maintaining firmware in version control with reviews and automated tests, and demonstrating the release history of shipped software.

Responsibilities

  • Write control firmware for each solid-state transformer cell and block, including modulation, closed-loop voltage and current control, and voltage sharing across series-stacked cells.
  • Build a fast protection path that detects faults and shuts down the converter within microseconds, from sensing through safe state.
  • Sequence startup, precharge, cell bypass, and output reconfiguration between 800 V and 1,500 V.
  • Connect cell controllers to the block master and site control system across isolation barriers, providing actionable telemetry for operators.
  • Validate firmware on hardware-in-the-loop rigs before deploying it to a high-voltage prototype.

Desired Qualifications

  • Experience with embedded power-control targets such as TI C2000, Xilinx FPGAs, or Intel FPGAs.
  • Experience with model-based design tools such as MATLAB/Simulink or PLECS.
  • Experience with hardware-in-the-loop platforms such as Typhoon HIL or OPAL-RT.
  • Experience with industrial communication technologies such as CAN, EtherCAT, or fiber links.
  • Experience with functional safety practices such as IEC 61508 or MISRA C.

About the Company

FluidStack

Builds and runs GPU clusters for AI labs

Fluidstack provides large clusters of GPU servers to AI labs training advanced models, and it builds custom data centers for customers that outgrow existing facilities. In 2025, Anthropic chose Fluidstack to build data centers in Texas and New York as part of a $50 billion investment in American AI infrastructure. Labs pay for dedicated clusters of Nvidia GPUs, sometimes tens of thousands of chips, housed in partner data centers around the world. They reach the clusters through a web console or API, while Fluidstack runs the hardware, networking, and day-to-day operations. Fluidstack was founded in London in 2017 and has its headquarters in New York.

Company size
501-1,000
Company stage
Late Stage VC
Total funding
$12.4B
Headquarters
New York City, New York
Founded
2017

Recently Posted Jobs

Explore more opportunities at FluidStack

See more jobs