AI/ML Systems Engineer - Performance Modeling & Optimization

GlobalFoundries inc.; Richardson (TX); On-site; USD 86,000 - 148,000; Full time

Job summary

GlobalFoundries is seeking an early-career AI/ML Systems Engineer to advance workload characterization and performance modeling across machine learning systems. The role will own quantitative analyses that connect workload behavior to hardware and software design decisions while communicating results across architecture, compiler, runtime, and product teams.

Essential responsibilities include characterizing AI/ML workloads, building models for throughput and latency, and partnering with

Qualifications

• 3.0 GPA minimum and good academic standing.
• Strong mathematical reasoning and numerical analysis for performance estimation.
• Familiarity with computer architecture concepts (cache, memory bandwidth, CPU pipelines).
• Exposure to ML frameworks such as PyTorch or TensorFlow.

Responsibilities

• Characterize AI/ML workloads across inference and training, quantify compute and memory behavior, and identify bottlenecks using methods such as roofline analysis, operational intensity profiling, and bottleneck decomposition.
• Build and validate analytical or code-based models that estimate throughput, latency, utilization, bandwidth, and efficiency for real or projected hardware.
• Partner with hardware, compiler, and runtime engineers to translate workload findings into architectural and software optimization discussions.
• Own benchmarking or modeling efforts end to end, maintain reproducible analysis artifacts, and communicate conclusions through reports and reviews.

Skills

Python

NumPy

Pandas

Matplotlib

PyTorch

TensorFlow

English fluency

Education

Bachelor’s or Master’s in Electrical Engineering or Computer Engineering

Job description

GlobalFoundries is seeking an early-career AI/ML Systems Engineer to advance workload characterization and performance modeling across machine learning systems. The role will own quantitative analyses that connect workload behavior to hardware and software design decisions while communicating results across architecture, compiler, runtime, and product teams.

Essential responsibilities include characterizing AI/ML workloads, building models for throughput and latency, and partnering with