Staff AI/HPC Infrastructure Architect

SanDisk; Milpitas (CA); On-site; USD 154,000 - 261,000; Full time

Job summary

Sandisk is seeking a Staff Engineer AI/HPC Infrastructure in Milpitas, CA to design and operate GPU-based AI platforms for training, inference, and research. The role focuses on AI infrastructure, LLM serving, memory architecture, and storage systems, shaping next-generation AI memory products.

You will build AI testbeds, deploy serving platforms, and characterize workloads while collaborating with academia and industry partners. On-site, full-time role with competitive compensation.

Qualifications

• BS+ 10 years or MS+ 6 years or PhD+ 3 years in relevant AI experience.
• Relevant AI experience required for BS/MS/PhD track.

Responsibilities

• Build and Modify AI/HPC Research Infrastructure: 8-100+ GPU environments and supporting server/network/storage/software.
• Deploy and Optimize LLM Serving Platforms: install and optimize vLLM, NVIDIA Dynamo, TensorRT-LLM, Triton, SGLang.
• AI Workload Characterization: capture workload traces, analyze memory bandwidth, KV-cache growth and storage access.
• AI Memory and Storage Architecture Evaluation: evaluate bottlenecks across SSDs, networking, and GPU interconnects.
• Digital Twin and Simulation: build sim environments and digital twins; model future AI infra configurations.
• Research and Ecosystem Collaboration: collaborate with universities, startups, labs, cloud providers, OEMs.

Skills

Linux

Kubernetes

Docker

Slurm

Python

GPU cluster administration

Distributed systems engineering

Education

BS+10 years or MS+6 years or PhD+3 years in relevant field

Tools

Linux

Kubernetes

Docker

Slurm

Python

Job description

Sandisk is seeking a Staff Engineer AI/HPC Infrastructure in Milpitas, CA to design and operate GPU-based AI platforms for training, inference, and research. The role focuses on AI infrastructure, LLM serving, memory architecture, and storage systems, shaping next-generation AI memory products.

You will build AI testbeds, deploy serving platforms, and characterize workloads while collaborating with academia and industry partners. On-site, full-time role with competitive compensation.