The Hardware Systems Architect position at Anthropic involves owning the system-level architecture of compute fleets, spanning from package boundaries to boards, racks, interconnects, power, cooling, and datacenter interfaces. Located in San Francisco, CA, New York City, NY, or Seattle, WA, this role is ideal for experienced hardware professionals who want to shape the foundational infrastructure supporting advanced AI research and deployment.
Role Overview
Anthropic trains and serves frontier artificial intelligence models on massive accelerator fleets. The Hardware Systems team oversees compute infrastructure to directly influence capability, cost, and reliability. As a Hardware Systems Architect, the professional works broadly across the hardware stack while diving deep when necessary to make directional decisions on building, buying, or partnering. Responsibilities include defining system architecture, writing specifications for internal teams and partners, and leading validation strategies.
Key Responsibilities
Own system path-finding and architecture for a specific domain across the breadth of compute deployed by Anthropic, acting as a reviewer and thought partner for others.
Author high-level requirements and specifications for systems, boards, interfaces, and racks that internal teams and external partners build against.
Evaluate partner designs against requirements, make build, buy, or co-design decisions alongside supply chain teams, and identify schedule risks early.
Drive interconnect and networking architecture, including scale-up and scale-out fabrics, Network Interface Cards, switches, optics, and network topology.
Formulate board, chassis, and system-level architecture with partners, covering component placement, printed circuit board architecture, connectors, cables, power, and thermal budgeting.
Develop performance simulations, hardware models, and scenario evaluations to select candidate hardware architectures.
Define validation, bring-up, and qualification strategies for new platforms, including telemetry and reliability targets.
Collaborate with machine learning performance and infrastructure software teams to ensure hardware decisions align with running workloads.
Shape hardware engineering practices by utilizing AI-assisted design, specification review, and validation.
Requirements
Minimum of 8 years of professional experience in hardware systems architecture or engineering for hyperscale, high-performance computing, artificial intelligence, machine learning, or high-end networking platforms.
Deep, hands-on expertise in at least one core hardware systems domain, such as interconnect, power, or thermal management for large-scale compute or networking systems.
Demonstrated experience owning hardware system architecture at the machine, rack, row, and cluster levels while authoring technical specifications.
Proven background taking hardware from initial architecture through bring-up and high-volume production deployment.
Experience working alongside external hardware vendors and partners to review designs and enforce specifications.
Strong capacity to reason about trade-offs across adjacent hardware domains like interconnect, power, thermal, and mechanical systems.
Track record of technical leadership, taking ownership of directional decisions and aligning cross-functional teams and partners.
Comfortable operating with high autonomy and limited process within a fast-moving, ambiguous environment.
Strong written communication skills for creating comprehensive specifications and technical reviews.
A bachelor’s degree in Electrical Engineering, Computer Engineering, Mechanical Engineering, Computer Science, or a related field, or equivalent practical experience.
Nice to Have
Experience with artificial intelligence accelerator systems including graphics processing units, tensor processing units, or custom application-specific integrated circuit platforms and their scale-up or scale-out fabrics.
Familiarity with chip-package-system interfaces and co-design.
Experience establishing new engineering functions, programs, or practices.
Curiosity and comfort utilizing artificial intelligence tools as part of an engineering workflow.
Specialized depth in high-speed SerDes, PCIe, CXL, Ethernet, InfiniBand, optical interconnects, baseboard management controllers, platform firmware, hardware root of trust, secure boot, 48V or high-voltage direct current power delivery, direct liquid cooling, or fleet-scale reliability engineering and reliability, availability, and serviceability architecture.
Skills and Tools
Interconnect and networking architecture
Power and thermal budgeting
Performance simulations and hardware modeling
System validation, bring-up, and qualification
High-speed SerDes, PCIe, CXL, Ethernet, InfiniBand, and optics
Baseboard management controllers and secure boot frameworks
Direct liquid cooling and high-voltage power delivery
AI-assisted engineering and design review workflows
Salary and Benefits
Annual base salary range of $320,000 to $485,000 USD.
Competitive compensation and benefits package.
Optional equity donation matching.
Generous vacation and parental leave.
Flexible working hours.
Location-based hybrid policy requiring staff to be in an office at least 25% of the time.
Visa sponsorship available for qualified candidates.
About the Company
Anthropic is a public benefit corporation headquartered in San Francisco dedicated to building reliable, interpretable, and steerable artificial intelligence systems that are safe and beneficial for society. Operating like big science, the team works as a single cohesive unit on large-scale research efforts, viewing AI research as an empirical science akin to physics and biology.
How to Apply
Interested candidates should submit an application directly on the official Anthropic careers page.
