Head of Silicon

Define our silicon direction and build a team for model-specific inference hardware.

About Gradiated

Gradiated is a research lab focused on inference. We co-design software and hardware for efficiently serving open-weight models.
Our goal is to maximize intelligence per joule. We work across model behavior, software, systems, and silicon to find better ways to run models.
We start by thinking about the model as a mathematical function, and map that out onto hardware and then build software on top of that which is custom to each model. This maximises performance and removes wasted abstractions.
We are early. You will help set the technical direction and the way the company works.

This will be our first silicon hire, we do not yet have a fixed micro-architecture. Your first job will be to work with the inference research team to understand the models we want to serve, and the current hardware bottlenecks.
You will build performance models and decide which ideas are worth pursuing. As the direction becomes clear, you will turn that into a silicon roadmap. You will then build and lead the team that takes the architecture through implementation, tape-out, and bring-up

Work with the inference research team to understand open-weight models and current hardware bottlenecks.
Build analytical and simulation models to compare microarchitectures, memory systems, dataflows, and interconnects.
Use those results to set targets for performance, power, area, cost, and reliability.
Decide which capabilities belong in silicon, software, external IP, or partner designs.
Turn the selected architecture into a roadmap through RTL, verification, physical design, tape-out, and bring-up.
Select IP, EDA, foundry, packaging, manufacturing, and design partners.
Hire and lead the initial silicon engineering team.
Review microarchitecture, RTL, verification, PPA, and bring-up results.
Define the hardware-software interface with the inference research team.

Microarchitectures and numerical formats for dense, sparse, and mixture-of-experts models.
Memory hierarchies for weights, activations, and KV-cache state.
Dataflows for attention, expert routing, prefill, and decode.
On-chip and chip-to-chip interconnects for model parallelism.
Performance models that predict latency, throughput, power, and utilization before RTL.
Hardware-software interfaces that can support new model structures without redesigning the chip.
Verification strategies for long-running and highly parallel inference workloads.

You have led architecture or implementation for a complex digital chip and taken at least one design through tape-out and bring-up.
You understand computer architecture, memory systems, interconnects, and power trade-offs.
You can connect workload measurements to architecture decisions.
You can work across architecture, RTL, verification, physical design, software, and systems.
You can evaluate external partners and hold them to a high technical standard.
You have hired and led a silicon engineering team.
You still enjoy getting your hands dirty in technical work and don't shy away from caring about the details.

You are looking for a well defined role with little ambiguity.
You want a management-only role.
You want to copy a general GPU architecture instead of starting with the model definitions.
You treat software as a fixed input to hardware design.
You want a large team before you make progress.
You lean on experience over trying new approaches.

We like to experiment a lot and do the hard thing even if it's likely to fail.
We work in the office on Monday, Wednesday, Friday. We value working in person to solve hard problems, but are also flexible, some of the team work abroad for short periods.

We offer top-of-market compensation, with significant equity.
We pay at the top of the market because talent density is what allows us to build the most efficient inference. We want a small team of exceptional people who have fun solving hard problems together.

A 60-minute call about your goals and the role.
We get together in person and talk through the problems we will work on.
A paid in-person working session, so we can get to know you and you can meet the team.
References and an offer decision.

Apply

If this sounds interesting to you, we'd like to hear from you. Fill in the form below and we'll get back to you within a few days.

Loading the application form