An embedded soft real time system in Go
summary
An embedded soft‑real‑time sorting system is implemented in Go running on standard Linux (no RT patch) across an i.MX6 quad‑core business logic board and multiple single‑core IO boards. The pipeline classifies up to 100 items/sec, synchronizes devices with PTP hardware timestamps, and meets tight deadlines with measured ~50 μs precision. Performance tuning uses pprof CPU profiles, protobuf messaging, reduced allocations to limit GC pauses, and optimized goroutine/channel patterns; the system runs successfully in production.
AI-written summary · about
chapters (10)
- 0:00Introduction to talk
- 2:17System overview and workflow
- 5:48Hardware architecture choices
- 10:12Synchronization and timing precision
- 13:29Development toolchain advantages
- 15:23Performance monitoring and profiling
- 21:19Optimization metrics and goals
- 24:56Network socket handling challenges
- 31:40Performance comparison insights
- 33:09Conclusion and final remarks
related talks
28mGo in real-time safety critical applications
30mHigh Performance Systems in Go
23mAn Operating System in Go
20mRedefining the Embedded Development Landscape with Software-defined SoCs
28mGo Make Something Real – The Potential for Go on the Factory Floor
26mTuning Real-Time Linux for ARM SoCs
35mWhy your next embedded project should be written in Go
44mGo Profiling and Observability from Scratch