Native Parity for Development & Testing of SOAFEE Software-Defined Vehicle
summary
Argues that running build/test on the same ARM architecture as ECUs cuts time, energy and bugs by removing cross‑compilation and emulation; shows Apex AI moving CI from x86 emulation to Ampere Ultra servers and AWS Graviton instances, using native toolchains (mold/GCC12), containers and KVM to run tens of thousands of unit tests and many virtual ECUs in parallel, and reports a surprising ~100x throughput gain and large carbon‑footprint reductions.
AI-written summary · about
chapters (11)
- 0:00Native Parity for Development & Testing of SOAFEE Software-Defined Vehicle
- 0:54“Native Parity” for Developer Productivity
- 1:39Motivation for Native Party Apex.AI
- 3:08Build Quality, Time, and Energy Use
- 4:03Test Quality, Time, and Energy Use 1 of 3
- 5:03Test Quality, Time, and Energy Use 2 of 3
- 6:12Test Quality, Time, and Energy Use 3 of 3
- 6:35Eliminate Cross-Compilation
- 7:20End Cross-Platform Debugging
- 7:32Simplify the CI Pipeline
- 7:57How to start the Native parity?
related talks
20mACES: Automatic Compartments for Embedded Systems
01h 03mApplied Best Practices
13mAccelerating Grid Digitalisation With Linux-based SEAPATH and Qu...
10mLightning Talk: Using Zephyr to Power the Sustainable Cloud
27mAutomated Testing using Travis and Shippable
21mEmpowering Developers to Thrive from Cloud to Edge
20mIntegrate RISC V to build Open Common Automotive Platform
20mDriving Intelligence: The Software Shift Shaping Tomorrow’s Cars
