Sorry, you need to enable JavaScript to visit this website.

SNIA Developer Conference September 15-17, 2025 | Santa Clara, CA

SMTS, Storage Solutions Architect,

Micron Technology

Wes Vaske is a seasoned Storage Solutions Architect with over 15 years of experience in data center storage systems, currently focused on high-performance NVMe solutions optimized for AI workloads. At Micron Technology, Wes plays a pivotal role in shaping next-generation storage architectures that meet the demanding performance and efficiency needs of AI-driven data centers. He is a lead contributor to the MLPerf Storage Working Group, where he helps define industry benchmarks for AI storage performance. Wes has been instrumental in developing automation frameworks for workload execution and system observation, enabling reproducible and insightful performance analysis across diverse environments. Prior to his current role, Wes led the Data Center Workloads Engineering team for a decade, where he pioneered system observation, tracing, and analysis tools that have become foundational to Micron’s workload-first product development strategy. His earlier career includes performance engineering for Oracle RAC database systems at Dell Technologies. Wes holds a B.S. in Physics from Iowa State University and continues to drive innovation at the intersection of storage, AI, and systems performance.

Discussion and Analysis of the MLPerf Storage Benchmark Suite and AI Storage Workloads

Submitted by Anonymous (not verified) on

Storage for AI is rapidly changing: Checkpointing becomes more important as clusters scale to more accelerators, managing large KV-Caches from LLM queries shifts inference bottlenecks to storage, accessing relevant data via VectorDB similarity searches drives small IOs for nearly every query, and future applications may require wildly different storage architectures. The MLPerf Storage v2.0 Benchmark Results were just released and the v2.5 suite is under active development.

Subscribe to Wesley Vaske