ASIC Architect

Job Title

ASIC Architect

Role Summary

As an ASIC Architect on the architecture team, you will translate high-level architecture specifications into micro-architecture requirements, build and validate performance and power models, and guide PPA trade-offs for new features. You will profile workloads, identify bottlenecks, and collaborate with software teams to model application-level behavior at cluster scale.

Experience Level

Senior — 10+ years of relevant experience in performance analysis and modeling of GPUs, CPUs, or accelerator products.

Responsibilities

The role focuses on architecture-to-microarchitecture translation, modeling, and performance-driven feature definition.

• Translate high-level architecture specifications into micro-architecture feature requirements.
• Develop and validate performance and power models; bring up new features in models.
• Perform PPA (performance, power, area) trade-off analysis for architectural features.
• Profile workloads, identify performance bottlenecks, and benchmark against competitors.
• Collaborate with software teams for end-to-end application and cluster-level modeling.
• Identify opportunities for kernel-level hardware acceleration and micro-code optimizations.

Requirements

Must-have skills and experience:

• 10+ years in performance analysis and modeling for GPUs, CPUs, or accelerator products.
• Strong background in computer architecture and high-level architectural trade-offs.
• Experience creating performance models from scratch using Python or similar tools.
• Experience with profiling and diagnosing micro-code/kernel performance bottlenecks.
• Ability to extract insights on power efficiency and project performance for benchmarking.
• Effective collaboration with software teams for system-level evaluation.

Nice-to-have:

• Familiarity with ML workload profiling and neural network architectures.
• Experience with large-scale accelerator or cluster-level performance modeling.

Education Requirements

Master's or PhD in Electrical Engineering or Computer Engineering (as listed by the employer).

About the Company

Company: Cerebras

Headquarters: Sunnyvale, CA, USA

Developer of wafer-scale AI accelerators, Cerebras designs the Wafer Scale Engine (WSE)—one of the world’s largest AI chips—to deliver high-speed training and inference solutions for model labs, enterprises, and AI-native startups.