Silicon Software Lead

Novel silicon without a software stack is a science project. Normal’s ASIC computes with stochastic analog dynamics in memory, and the software layer that makes it programmable and performant for real inference workloads does not yet exist in any standard form. As our Silicon Software Lead, you will lead the team that builds it: the compiler, runtime, kernels, drivers, and hardware abstraction layer that turn our chip into a platform. You will set technical direction, stay hands-on in the stack, and co-design with hardware architects so that software constraints shape the silicon rather than arriving after it.

This is a seat for someone who has built software for hardware that did not exist yet, and wants to do it where the software genuinely changes the chip., * Team Leadership: Lead and grow the silicon software team spanning compiler, runtime, and systems software, staying close enough to the code to review designs and unblock hard problems directly.

• Software Stack Architecture: Own the architecture of the stack from ML framework ingestion through compilation, scheduling, and memory management to execution on Normal hardware.
• Hardware Co-Design: Partner with silicon architects on ISA definition and the hardware abstraction layer, ensuring the chip is compilable and programmable, not just simulatable.
• Runtime & Tooling: Drive development of the runtime, kernels, drivers, profiling, and debugging tools that make the hardware usable for real inference workloads.
• Simulation-to-Silicon Continuity: Keep the software stack running against simulation, FPGA prototypes, and silicon as the hardware matures, so software development never waits on tapeout.
• Roadmap & Hiring: Set the silicon software roadmap, define milestones against the hardware program, and hire the engineers who deliver it.

• Substantial experience building software stacks for accelerators or non-standard hardware targets: compilers, runtimes, kernels, or drivers
• Experience leading engineers as a technical lead or manager while staying hands-on in design and code
• Strong systems programming skills in C++, Rust, or equivalent, with fluency in Python
• Deep understanding of ML inference workloads and the constraints that shape their execution on hardware
• Experience with compiler frameworks such as MLIR or LLVM, or with inference runtimes and kernel development
• Comfort building software for hardware that is still evolving, from simulation through bring-up
• Track record of hiring and developing strong systems engineers

Bonus Points

• Experience taking an accelerator software stack from zero to production at a startup or new hardware program
• Experience with in-memory compute, processing-in-memory, or analog hardware interfaces
• Contributions to open-source compiler or runtime infrastructure
• Experience with hardware-software co-design where software insights shaped ISA or architecture decisions

Normal Computing builds silicon that turns thermal noise from an obstacle into a computational resource. Conventional chips spend most of their energy forcing determinism onto physics; ours compute with it. Stochastic, in-memory, asynchronous: the result is 10-100× more AI inference per dollar, per watt.

We co-design the full stack: AI-native EDA systems in production with the world’s largest semiconductor companies, and the advanced ASICs they make possible. Backed by $85M+ from the world’s leading deep-tech investors and built by scientists, engineers, and operators from the labs that built modern computing.

Normal works as one team across New York, Silicon Valley, London, Copenhagen, and Seoul. We hire people who want the hardest version of their craft, across every discipline, at every seniority.