AI Memory Performance & Power Modeling Intern

Micron Technology, Inc.; San Jose (CA); On-site; USD 69,000 - 77,000; Part time

Job summary

Micron Technology, Inc. seeks an AI/ML Workload Analysis Intern to collaborate with system architects and performance engineers. You will analyze AI workloads, collect memory traces, and study memory behavior to inform future memory technologies and AI platforms.

The role emphasizes trace-based analysis, simulation, and data-driven insights into bandwidth, cache efficiency, and data movement across CPUs, GPUs, NPUs, and accelerators.

Qualifications

• Pursuing MS or PhD in a relevant technical field.
• Strong background in computer architecture and ML.
• Experience with performance analysis and race conditions.

Responsibilities

• Analyze AI/ML workloads including LLMs and vision models.
• Collect and analyze memory traces from training/inference.
• Characterize memory access, bandwidth, and latency.
• Develop workload traces and architecture simulations.
• Identify bottlenecks in memory systems and interconnects.
• Present findings via reports and visualizations.

Skills

Python

C/C++

Machine Learning

Parallel Computing

Operating Systems

Data Structures & Algorithms

Performance Analysis

Education

MS or PhD in Computer Engineering, Computer Science, Electrical Engineering, Data Science, or related field

Tools

Nsight

VTune

Perf

rocprof

Gem5

Accel-Sim

Job description

Micron Technology, Inc. seeks an AI/ML Workload Analysis Intern to collaborate with system architects and performance engineers. You will analyze AI workloads, collect memory traces, and study memory behavior to inform future memory technologies and AI platforms.

The role emphasizes trace-based analysis, simulation, and data-driven insights into bandwidth, cache efficiency, and data movement across CPUs, GPUs, NPUs, and accelerators.