The Thermofluids R&D TeamExamples of key problems the team is working on- Reject more than 10 gigawatts of heat, starting now. Fluidstack is targeting 10 GW of compute in 2027 and 30 GW the year after. Every watt of it becomes heat that has to go somewhere, and no cooling team has ever been asked to open at that number rather than grow into it.
- Design and build the machines, because nobody can supply them. The merchant compressor market cannot deliver at this rate, and the vendor base behind today's chillers will not scale to gigawatts. We are taking compressor and chiller design in house, working on machine architectures nobody sells today, and going from blank sheet to production faster than this industry thinks is possible.
- Own every thermofluids problem behind the fleet, not just the chiller. Working fluid and cycle architecture, heat exchangers at a scale that dominates the cost of the plant, two-phase heat transfer at the chip, and thermal energy storage. All of it is open, and whatever this team lands gets built thousands of times.
Role Scope- Own the aerodynamic design of the centrifugal compressor, from mean-line and 1D through CFD to released impeller, diffuser and return-channel geometry.
- Set the machine architecture, including stage count, shaft speed and diameter, while the working fluid decision is still open and moves the answer substantially.
- Design the expansion turbine that recovers shaft work from the high-side pressure drop, sharing a shaft with the compressor and the motor instead of throwing that energy away across a valve.
- Carry the rotor through rotordynamics, magnetic bearing integration and stress with the structural and mechanical engineers, at the shaft speeds a compact high-speed machine demands.
- Take the machine onto the test stand, and close the loop from a measured performance map back into the aerodynamic design.
What We're Looking ForThe below is a starting point. We always make space for exceptional people, so if you don't fit this role exactly, tell us where you would.- You've designed a radial or axial turbomachine that got built and ran, and you were standing next to it when it did.
- You've taken a stage from mean-line through CFD and back, and you know the conditions under which your CFD is lying to you.
- You've worked a rotordynamics problem through to resolution on a real shaft with real bearings, not just in a report.
- You've made an architecture call early with incomplete data, stage count or speed or bearing type, and lived with it through build and test.
- You read a test map and change the design, rather than explaining why the map is wrong.
- You've worked somewhere the next article was already in the shop before the last one came off test, and you preferred it that way.
- Bonus: Turbopumps, gas turbines and cryogenic turbomachinery (launch vehicles, APUs, aero engines). High-speed motor-driven compressors (magnetic bearings, hermetic, oil-free machines). Transcritical and supercritical working fluids (R744, sCO2 cycles). Real-gas property libraries (CoolProp, REFPROP). Vapor-compression machine design (chillers, heat pumps, industrial refrigeration).
We are committed to pay equity and transparency.
