Keynote: Network Transformation: Are We Ready for 5G and Edge Clouds?
summary
Claims 100+ field trials and concrete platform moves: DPDK hits about 40 million pps (≈600 Gbps on a two‑socket Xeon), XDP work aims to fold those optimizations into the kernel, and NIC features like application device queuing provide container‑level low‑latency paths. Second‑gen Xeon -N SKUs report 58% performance gain versus prior gen and new persistent memory cuts TCO ~34%. Prototypes show AI‑driven closed‑loop automation for EPC and CMTS within ~2% of static policies; ONAP, telemetry (collectd/PMF) and an “openness” API bundle target cloud‑native edge deployments.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 3:11Intel 5G Contributions
- 4:47Network Infrastructure Transformation
- 6:22Challenges
- 7:58Cloud Native Network Functions
- 11:33Closed Loop Automation
- 13:20Open Source
- 14:57Digital Transformation
- 22:04Intel Announcements
- 23:25Summary
related talks
19mKeynote: Network of Tomorrow - 5G and Edge
30mKeynote: 5G is Now
30mKeynote: Cloud Native Networking
25mDPDK performance optimization for edge computing scenario
01h 25mDay 2 Part 1 Sessions
22mEmpowering Edge Software Stack on Arm with High Performance Container Networking Tech
22mKeynote: Accelerating the Pace of Innovation
20mKeynote