By 
The eBPF Foundation today announced $100,000 in unrestricted research grants to support projects aimed at strengthening eBPF programmability, runtime safety, and datacenter energy efficiency. Two universities—University of Michigan and University of California, Riverside—were each awarded $50,000 following a competitive review of 27 proposals from 23 institutions.
The selected projects tackle key challenges in the evolution of eBPF, a Linux kernel-originated technology now powering production systems across billions of devices. Research includes:
📹 Going on record for 2026? We're recording the TFiR Prediction Series through mid-February. If you have a bold take on where AI Infrastructure, Cloud Native, or Enterprise IT is heading—we want to hear it. [Reserve your slot
Verifier-Cooperative Instrumentation for eBPF Safety – Led by Ryan Huang (University of Michigan), this project introduces EPASS, a runtime enforcement framework that extends static verification. Early results resolved 91% of previously rejected safe programs and mitigated 14 known vulnerabilities with minimal performance overhead.
QoS-Aware Power Management for Datacenters – Led by Daniel Wong (UC Riverside), this project develops eBPF-based power governors to dynamically optimize CPU performance and energy use. Initial findings show up to 19% power savings in volatile workloads without sacrificing latency-critical performance.
“The proposals selected this year highlight how eBPF continues to advance as a foundation for the next generation of infrastructure software,” said Bill Mulligan eBPF Foundation Board Member. “By improving safety guarantees and extending programmability into areas like datacenter energy efficiency, these projects show how eBPF is not just making the kernel safer and easier to program, but also enabling entirely new capabilities to address the infrastructure challenges of today and tomorrow. We are excited to support these projects and look forward to the contributions they will make to the eBPF ecosystem.”
The eBPF Foundation Research Grant program funds university-led innovation to improve eBPF’s functionality, security, and programmability while offering hands-on experience for students and researchers. Results from the 2025 projects will be published in 2026 alongside a new funding call.
