Building the best stack for the largest companies
summary
Microsoft describes how a platform layer ties tooling, services and apps so enterprises can move from experiments to production using concrete patterns: data cataloging and deep learning for text, deploying models to edge devices and IoT, and combining telemetry with drone imagery for predictive maintenance. The claim is that pushing inference close to the event and an ongoing hardware arms race — CPUs GPUs FPGAs — will cut latency and cost, and customers are already shipping these integrated solutions.
AI-written summary · about
chapters (5)
- 0:00Introduction
- 0:55How is the AI market shaping up
- 3:17Working with large enterprise customers
- 4:38Do we have enough knowledge
- 6:28Whats next
related talks
29mFrom Experiment to Enterprise: Architecting AI for Stability and Scale
06mIntel’s shift to delivering end-to-end data solutions
09mTackling edge computing with Intel's AI Builders program
26mEngineering Excellence: Platforms and Architectures for Real-Time AI Analytics
08mDesigning Scalable, Secure Data Architectures for AI at Enterprise Scale
04mA Cloud Native Stack From Bare Metal To Tokens for Large-Scale AI Inference
15mEnabling Scalable Data Analytics for Enterprise Success
25mOptimize productivity by meeting next-level demands of the AI era