Performance Architect, CPU Cluster
$100k - $500k • Remote • Austin, Texas, United States; Santa Clara, California, United States; Santa Clara, California, United States
Posted 10d ago
Job Location
Austin, Texas, United States; Santa Clara, California, United States; Santa Clara, California, United States
Tech Stack
Remote Work Policy
Fully remote
Categories
Applied AI Engineer
About the job
Tenstorrent is seeking a CPU Cluster Performance Architect to influence the performance and scalability of their next-generation CPUs. This role focuses on the intricate interactions between multiple CPU cores within a cluster, including cache hierarchies, coherent interconnects, memory bandwidth, and overall system behavior. It's a hands-on architecture position that leverages performance models and simulation to identify bottlenecks, evaluate design tradeoffs, and guide architectural decisions early in the development cycle. The architect will collaborate closely with CPU architects, RTL designers, software and compiler teams, and system architects to understand workload characteristics and translate performance insights into tangible architectural enhancements. This is a remote position based in North America, and the company welcomes candidates at various experience levels.
Responsibilities
- Analyze and optimize CPU cluster performance, focusing on cache hierarchies, interconnects, coherence, and memory access patterns.
- Build and utilize performance models and simulation environments (e.g., Gem5) to assess architectural concepts and identify performance improvement opportunities.
- Study real-world workloads to pinpoint bottlenecks related to cache misses, coherence traffic, memory bandwidth, latency, contention, and core-to-core communication.
- Lead architectural tradeoff studies considering performance, scalability, bandwidth, latency, power consumption, and implementation complexity.
- Collaborate with various architecture, software, and system teams to translate performance analysis into concrete design decisions.
Requirements
- Strong foundation in CPU microarchitecture and performance, understanding the interplay of cores, caches, interconnects, and memory systems.
- Proficiency in using modeling, simulation, and workload analysis to understand cluster-level performance bottlenecks and scaling.
- Comfortable addressing both detailed microarchitecture and broader system-level performance aspects like latency, bandwidth, quality of service, utilization, and scalability.
- Analytical and hands-on approach to transforming performance data into clear architectural recommendations.
- Strong communication skills for effective collaboration across diverse engineering teams.
Benefits
- Compensation ranges from $100k - $500k including base and variable compensation targets.
- Highly competitive compensation package and benefits.