Performance Modeling Engineer - AI Hardware

Normal Computing Corporation; Palo Alto (CA); On-site; USD 150,000 - 210,000; Full time

Job summary

Normal Computing is seeking a Performance Modeling Engineer to build and maintain cycle-accurate TLMs for AI accelerator hardware. You will translate RTL, FPGA prototypes and synthesis results into calibrated latency, bandwidth, and energy parameters used by architecture and DV teams.

You will own the path from RTL to performance results, validate models against real workloads, and plan modeling work against the hardware roadmap, delivering actionable guidance to the team.

Qualifications

• 5+ years of experience in performance modeling at a major semiconductor company.
• Demonstrated proficiency in understanding architectures and transforming that to performant modeling code.
• Use of various modeling techniques and languages.
• Excellent written and spoken communication skills.

Responsibilities

• Transaction-Level Model: Build and maintain a cycle-accurate TLM of compute tiles, memory, and interconnect in SystemC/C++/Python.
• Feeding the Modeling Stack: Translate TLM/RTL results into calibrated latency, bandwidth, energy, etc.
• Verification Reference: Provide TLM as golden model for DV with scoreboards and checks.
• Planning and Performance Feedback: Plan work against hardware roadmap and run sensitivity studies.

Skills

Performance modeling

Architectural understanding

Modeling in SystemC/C++/Python

Communication skills

Tools

AccelForge

Timeloop/Accelergy

CACTI

McPAT

gem5

Job description

Normal Computing is seeking a Performance Modeling Engineer to build and maintain cycle-accurate TLMs for AI accelerator hardware. You will translate RTL, FPGA prototypes and synthesis results into calibrated latency, bandwidth, and energy parameters used by architecture and DV teams.

You will own the path from RTL to performance results, validate models against real workloads, and plan modeling work against the hardware roadmap, delivering actionable guidance to the team.