Low-Latency Real-Time (RT) Kernel: The Cornerstone of Arm's 5G ORAN Strategy
summary
Measures real RT latency on Arm Neoverse hardware using cyclictest v2.20 (200 µs interval, 100M loops) and finds surprising long tails at high core counts: 64‑core runs spike over 100 µs while most samples sit near 2 µs. Root cause traces with trace-cmd point to getrandom/CRNG contention from kernel stack randomization exhausting entropy. Disabling stack randomization and applying a CRNG patch drops the max latency to under 10 µs and meets target.
AI-written summary · about
chapters (14)
- 0:00Introduction
- 2:11Real-time latency in 5G
- 7:30Latency requirements and OSI model
- 11:09Benchmark setup and methodology
- 12:54Initial latency data
- 16:11Bug reports and fixes
- 19:21Kernel tuning for real-time
- 24:08Latency results with tuning
- 25:35Kernel stack randomization impact
- 26:40Root cause analysis with Trace-CMD
- 37:11Debugging high latency with function tracing
- 39:43Entropy exhaustion in RNG
- 41:10Final results and patches
- 41:33Q&A
related talks
50mLow-Latency Real-Time (RT) Kernel: The Cornerstone of Arm's 5G ORAN Strategy
26mTuning Real-Time Linux for ARM SoCs
26mReal-time Linux: what is next?
44mReal-Time Linux on Embedded Multicore Processors
54mIntroduction to Realtime Linux
31mBeware of the kernel RTNL mutex
52mEffectively Measure and Reduce Kernel Latencies for Real-time Constraints
40mRT Linux in Safety-Critical Systems: The Potential and the Challenges