PhD Intern: AI Memory Systems Architectures

1000 Micron Technology, Inc.; Folsom (CA); On-site; USD 80,000 - 98,000; Full time

Benefits offered by this job

Medical, dental and vision plans

Paid time off

Paid holidays

Job summary

Micron Technology, Inc. invites a PhD intern to join the HBM Architecture team to research memory hierarchies and interconnect choices for AI accelerators.

You will build performance models, analyze AI workloads, and push innovations across GPUs, TPUs, and custom accelerators in a collaborative environment. The role focuses on memory system optimization, power efficiency, and scalability for next‑gen AI infrastructure, offering opportunities to work with seasoned engineers and publish impactful

Qualifications

• Pursuing a PhD in a related technical field.
• 0-2 years of relevant research or internships in computer architecture or AI hardware.
• Experience with memory systems: HBM, DRAM, caches.
• Programming in Python or C++.

Responsibilities

• Research and evaluate advanced memory and system architecture concepts for AI accelerators.
• Develop and analyze performance models to assess bandwidth, latency, power, scalability.
• Analyze AI training, inference, and HPC workloads to identify memory bottlenecks.
• Investigate memory hierarchy, HBM, DRAM, interconnects to improve performance.
• Collaborate with cross-functional teams and explore AI/LLM tools to aid research workflows.

Skills

Python

C++

Performance modeling

AI hardware

Analytical thinking

Education

PhD candidate in Computer Architecture

Tools

SystemC

CUDA

GPU profiling tools

Job description

Micron Technology, Inc. invites a PhD intern to join the HBM Architecture team to research memory hierarchies and interconnect choices for AI accelerators.

You will build performance models, analyze AI workloads, and push innovations across GPUs, TPUs, and custom accelerators in a collaborative environment. The role focuses on memory system optimization, power efficiency, and scalability for next‑gen AI infrastructure, offering opportunities to work with seasoned engineers and publish impactful