Product Manager, Hardware
Normal Computing | Build with Us
Normal Computing is building a new paradigm of AI hardware to power human progress. We build inference servers and rack-scale systems designed to increase efficiency and interactivity. We use AI to accelerate silicon co-design and product development.
Our primary offices are in New York City, London, and Mountain View.
The Role
As the Product Manager for Hardware, you'll help shape the roadmap for our silicon: what goes in the chip, and the execution that gets it to tapeout. We are building architectures targeting LLM inference. Your job is to turn target workloads, customer signal, and supply chain constraints into a chip definition the team commits to, then be accountable for milestones from architecture spec through RTL, verification, physical design, and signoff to tapeout.
You will support the Head of Product in driving alignment across teams by justifying technically how the silicon roadmap supports the company's higher level product vision in discussions with the CPO, CTO, and CSO. On any given day, you might sit with a chip architect weighing an architecture tradeoff against a workload assumption, pressure-test a performance model, rework the milestone plan after a late-breaking DV bug, or sit with a prospective customer to understand their cost-per-token and latency targets. This is a highly collaborative role as you work with the Silicon, Partnerships, Strategy, and Business teams to refine and execute on the roadmap.
What You'll Own
Work with the Silicon engineering team to decide what goes in each silicon generation. Prioritize features against target workloads, PPA tradeoffs, and what customers will pay for.
Work with the Ecosystem and Partnerships team to translate workload assumptions and customer requirements into architecture requirements and success criteria the design team builds against.
Own the milestone plan from spec to tapeout. Track the dependency chain across architecture, RTL, verification, DFT, and physical design closure.
Organize and facilitate milestone alignment and refinement meetings for both internal stakeholders and external customers
What Makes You a Great Fit
A degree in EE, CS, Material Science/Engineering, or similar, or the equivalent taught to yourself.
4+ years in product management, program management, or engineering roles in semiconductors or adjacent hardware.
You've been close to a tapeout. You know the lifecycle from spec through RTL, verification, physical design, DFT, and silicon bring-up.
Fluency with coding agents. You already use Claude Code, Codex, Cursor, or something similar as a normal part of how you work.
Clarity of communication in both speaking and writing.
Comfort with ambiguity. You should be able to take a vague problem, frame it, and come back with a proposal
Bonus Points
Experience defining or shipping AI accelerators, or deep familiarity with the mechanisms underlying LLM inference.
Exposure to SystemVerilog, design verification, or hardware performance modeling.
Experience working directly with foundries or memory vendors.
Understanding of ASIC based inference server design, IP blocks, packaging, and the implications of choosing specific process nodes.
We encourage you to apply even if you don't meet every qualification listed.
Equal Employment Opportunity Statement
Normal Computing is an Equal Opportunity Employer. We celebrate diversity and are committed to creating an inclusive environment for all employees. All qualified applicants will receive consideration for employment without regard to race, color, religion, sex, sexual orientation, gender identity, national origin, disability, veteran status, or any other legally protected status.
Accessibility Accommodations
Privacy Notice
By submitting your application, you agree that Normal Computing may collect, use, and store your personal information for employment-related purposes in accordance with our Applicant Privacy Notice.
Required Languages
🇬🇧 English