Intel just archived a batch of open-source GitHub projects, and buried in the list are tools that mattered for AI inference work. The Intel Edge AI Performance Evaluation Toolkit, used for qualifying platforms running deep learning inference on Intel hardware, is now frozen. So is Predictive Assets Maintenance, a time-series AI solution that demonstrated Intel Xeon capabilities. The code stays readable in archived form, but active development and support are finished. The moves stem from Intel's brutal August 2024 cost-cutting. CEO Pat Gelsinger ordered over 15,000 layoffs after posting a $1.6 billion net loss in Q2. The target: $10 billion in savings by 2025, with spending redirected toward manufacturing and core silicon. Community programs and evangelism teams don't generate direct revenue, so they got axed. Open-source evangelist Katherine Druckman departed back in July, and the entire Open Ecosystem Community and Evangelism program is now formally shut down. As Michael Larabel at Phoronix reported, Intel has been archiving projects month after month as part of this restructuring. The Edge AI toolkit was an actual tool for benchmarking deep learning performance on Intel hardware. Without maintenance, it'll fall behind as the hardware changes. The company is pulling back from the software ecosystem that made its chips easier to adopt, right when AMD and NVIDIA are investing heavily in developer tools and community. Intel is betting that silicon matters more than the software layer around it.