Performance Architect, CPU Cluster

Location: Remote, based in North America. The listed office location is Santa Clara, California, United States.

Compensation: $100,000–$500,000 per year, including base and variable compensation targets.

Company

Tenstorrent develops high-performance AI platforms combining software models, compilers, systems, networking, semiconductors, and RISC-V CPUs.

What you will do

• Analyze and optimize CPU cluster performance across cache hierarchies, interconnects, coherence, and memory access behavior.
• Build and use performance models and simulation environments such as Gem5 to evaluate architectural concepts.
• Study workloads to identify cache, coherence, memory bandwidth, latency, contention, and core-to-core communication bottlenecks.
• Lead tradeoff studies covering performance, scalability, bandwidth, latency, power, and implementation complexity.
• Translate performance analysis into design decisions in collaboration with CPU, cache, interconnect, memory, software, compiler, RTL, and system architects.

Requirements

• Strong foundation in CPU microarchitecture and performance.
• Understanding of cores, caches, interconnects, memory systems, and their interactions.
• Experience or interest in modeling, simulation, and workload analysis for cluster-level performance.
• Ability to analyze performance data and produce clear architectural recommendations.
• Strong communication skills and ability to work across architecture, RTL, compiler, software, and system teams.
• Eligibility to access U.S. export-controlled technology may be required; employment can depend on citizenship, permanent residency, or prior license approval.

Culture & Benefits

• Collaborative environment focused on curiosity and solving technically difficult problems.
• Opportunity to influence CPU cluster architecture from early performance exploration through implementation and silicon.
• Competitive compensation package and benefits.
• Open to candidates at various experience levels, with level determined during the interview process.