Founding Lead Silicon Engineer

About Zettascale

Energy efficient chips for AI

Tech

Our XPU chips are state-of-the-art AI compute engines, highly programmable and efficient enough to support AGI, and eventually ASI, without requiring massive power infrastructure. The major bottleneck in computing is memory movement (often referred to as the "memory wall"). Moving large amounts of data from point A to B is **very expensive** in both time and energy. A good rule of thumb is 1 bit = 1 pJ, and modern AI models are >100 GB in parameters. Our **polymorphic architecture** is more programmable, which lets us fuse kernels more aggressively. That keeps more of the model's computation in-place, so we move less data, route it more efficiently, and write back to memory less often.

The role

## The Mission**Zettascale Computing is building the highest-performance, most energy-efficient AI chips for artificial superintelligence to accelerate discovery and further humanity's understanding of the universe.**We are a high-agency, deeply technical, and intensely mission-driven founding team. We value first principles thinking, intellectual honesty, speed, and full ownership.## The RoleLead Zettascale Computing's silicon efforts from FPGA prototype through tapeout and bring-up. You will set technical direction across architecture, RTL, verification, tooling, and timing closure, build the silicon team as the company grows, and partner closely with the PCB, interconnects, racks and pods, and software teams.## What You Will Build* Lead development and optimization of the FPGA prototype of our XPU* Design and review the RTL that becomes our first ASIC* Build the lint, simulation, synthesis, timing, and verification flows* Make architecture decisions across compute, memory, on-chip interconnect, and control* Drive the design to tapeout and bring-up## What You Bring* Strong digital design fundamentals, including RTL, pipelining, clocks, resets, latency, throughput, and clean microarchitecture* Hands-on FPGA experience taking real designs from RTL to a working board* Synthesizable SystemVerilog or Verilog you are proud of* Fluency with simulation, linting, synthesis, SDC constraints, timing closure, and CDC* Experience automating hardware flows with Python, Tcl, Make, Nix, or similar tools* Comfort designing compute datapaths and memory systems around bandwidth and latency* Experience with AI accelerators or low-level AI kernels* A proven autodidactTapeout experience, large FPGA prototypes, high-speed interface integration, formal verification, UVM, cocotb, firmware, drivers, and systems programming are strong advantages.## Expectations* This is a full-time, in-person role at our San Francisco HQ. It is not remote or hybrid* You move quickly, take broad ownership, and prefer working hardware over perfect specifications* We do not track hours. We care about output, quality, and throughput* Critical tapeout, bring-up, and delivery milestones sometimes require long days and weekends. We expect you to lean in when they do* You are ready to go all-in and do the work of your life## Compensation and Support* $200,000 to $300,000 base salary* 1.00% to 3.00% equity* Medical, dental, and vision coverage* 401(k), equipment, and gym support* Relocation assistance to San Francisco* Visa sponsorship where feasible

Skills

• Tcl
• Verilog
• FPGAs