Edge AI Performance Modeling Engineer — Cycle-Level

Quadric; Burlingame (CA); On-site; USD 180,000 - 225,000; Full time

Benefits offered by this job

Medical, dental, and vision insurance

Company-paid life Insurance

Equity

Paid Parental Leave

401(k) Retirement Plan

Flexible PTO

Catered lunch

Caltrain reimbursement

Downtown Burlingame office

Job summary

Quadric in Burlingame, California, is seeking an AI Performance Modeling Engineer to build cycle-level models of AI inference on the GPNPU architecture. You will work in Python to guide hardware decisions and optimize mapping of workloads from vision nets to LLMs.

This role welcomes all experience levels with mentorship, and offers a competitive base salary, equity, and a range of benefits; you’ll collaborate with a small, high-impact team in a growth-stage semiconductor company.

Qualifications

• Python skills to write, validate and calibrate numerical models.
• Solid grasp of memory hierarchies, bandwidth, latency trade-offs, and bottlenecks.
• Experience writing clear technical studies with defended conclusions.
• Core depth in NN inference operators or equivalent performance modeling.
• BS/MS/PhD in CS, Electrical/Computer Engineering, or equivalent experience.

Responsibilities

• Build analytical, cycle-level Python models of AI inference on next-gen GPNPU hardware.
• Derive hardware lane bindings and model software pipeline overlaps.
• Model tensor placement, tiling, memory residency, and data movement across memory tiers.
• Calibrate models against ISA simulator and profiling traces; meet accuracy targets.
• Write and defend technical studies that inform architecture and product decisions.
• Balance single-stream latency and throughput across workloads.

Skills

Python

Quantitative modeling

Computer architecture

Technical writing

NN inference operators

Education

BS/MS/PhD in CS/EE/CE

Tools

CUDA

Triton

gem5

Timeloop

MAESTRO

Accel-Sim

Job description

Quadric in Burlingame, California, is seeking an AI Performance Modeling Engineer to build cycle-level models of AI inference on the GPNPU architecture. You will work in Python to guide hardware decisions and optimize mapping of workloads from vision nets to LLMs.

This role welcomes all experience levels with mentorship, and offers a competitive base salary, equity, and a range of benefits; you’ll collaborate with a small, high-impact team in a growth-stage semiconductor company.