William Power
Principal AI Software Architect at Intel
- Role
- Principal Ai Software Architect at Intel
- Location
- Boulder, CO, US
- LinkedIn followers
- 500 followers
About William Power
Inferencing. AI Inference Scaling, Inference Performance Optimization, Rack-Scale AI Architectures, Disaggregated Prefill/Decode, KV-Cache Optimization, vLLM Runtime Development, NIXL, High-Speed AI Storage Systems (GPU-direct paths), Distributed Systems Performance Engineering (GPU / PCIe / Libfabric / Networking / Storage). D. Content Delivery Networks (CDN), Edge Computing, Cloud Infrastructure (Kubernetes)
Experience
Principal Ai Software Architect
Mar 2024 — Present · Denver, CO, US
AI Inference scaling, disaggregated prefill/decode, kv-cache optimization with vLLM/llm-d• Evaluated GPU Direct Storage for Intel GPUs, assessing throughput and latency tradeoffs. • Ported NIXL to Intel Gaudi GPUs, enabling high-throughput data movement for AI workloads. • Researched KV cache integration with vLLM, optimizing AI inference on Intel Gaudi GPUs. • Led enhancements of the WEKA and VAST storage platform, improving performance and security for multi-tenant AI cloud environments.
Education
University College Dublin
B.S, Computer Science, Mathematical Physics
1989 — 1994
The University of Edinburgh
M.S, Cognitive Science
1995 — 1997
Find verified contacts for anyone on LinkedIn
Unifers gives sales teams verified emails and direct dials, enriched profiles, and outreach that lands in the inbox.
Free plan included · No credit card required
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.