Community Central

Visit CentOS and Fedora and many other projects in Community Central, part of Red Hat Central in the expo hall. CentOS experts will be available to talk about CentOS Stream, CentOS SIGs, EPEL, bootc, and image-mode Linux. We'll also have speakers from Community Day at scheduled times for followup conversations.

From Fedora to RHEL: Building the future of enterprise Linux together

Information

Monday, May 11 | 2:15 PM - 2:35 PM EDT | B308-B309 - Level 3

Scott McCarty • Jef Spaleta

The future of Red Hat Enterprise Linux (RHEL) begins in the Fedora community. In this session, Scott McCarty, Lead Product Manager for RHEL, and Jef Spaleta, Fedora Project Lead, will explore how Fedora’s open innovation model directly informs, shapes, and accelerates the evolution of RHEL 11. Attendees will get a behind-the-scenes look at how Fedora provides a transparent path for innovation, feedback, and collaboration across the Linux ecosystem. You’ll learn how ideas move from upstream experiments to enterprise-grade features, and how open communities, contributors, and customers all play a role in shaping what becomes production-ready in RHEL. Together, Scott and Jef will discuss how this upstream-first model benefits everyone: Fedora contributors gain visibility into enterprise needs; RHEL users gain earlier access to innovation and influence; and the Linux ecosystem as a whole becomes stronger through openness, predictability, and shared purpose. This session isn’t just about operating systems—it’s about operating openly. Join us to see how community-driven innovation fuels the next generation of enterprise Linux.

RISC-V enterprise readiness: Technical update

Information

Monday, May 11 | 2:15 PM - 2:35 PM EDT | B311 - Level 3

Jefro Osier-Mixon

How ready is the RISC-V ISA for enterprise use? Is the open source software ecosystem keeping pace with new hardware? What is the real status of datacenter grade RISC-V silicon? This session provides a technical update on the RISC-V ecosystem, answering critical questions from customers evaluating the platform. We will cover the importance of the newly ratified RVA23 ISA specification, which standardizes the architecture for enterprise workloads and provides the foundation for accelerating AI/ML via new vector and matrix extensions. We'll also discuss Red Hat's foundational role in this ecosystem, with contributions spanning from ISA-level working groups to Red Hat Enterprise Linux (RHEL). An architecture is only as strong as its software. We'll discuss open source software porting and hardware enablement in the Linux kernel. We'll explore the integration status in Fedora, and the ongoing enablement in CentOS Stream. These community efforts show the platform's growing maturity. Finally, we connect this upstream innovation directly to enterprise value. Red Hat's engineering leadership in Fedora and CentOS Stream paved the way for the RISC-V Developer Preview on RHEL. This session explores the road ahead for RISC-V—its path toward becoming a fully supported architecture—and how customers can begin evaluating RISC-V for their own workloads.

RISC-V and Red Hat in conversation

Information

Monday, May 11 | 2:40 PM - 3:00 PM EDT | B311 - Level 3

Jefro Osier-Mixon • Brian W Harrington

RISC-V is emerging as a collaboratively developed instruction set architecture, but it faces unique software challenges, from firmware, kernel, and drivers to language runtimes and applications. Red Hat and the open source community have risen to the challenge by co-founding the RISC-V Software Ecosystem (RISE) project, paired with Red Hat's efforts in Fedora, CentOS, and Red Hat Enterprise Linux development. In 2025, Red Hat introduced the first Developer Preview of Red Hat Enterprise Linux running on a RISC-V platform. Join us for a conversation between Jeffrey "Jefro" Osier-Mixon, RISC-V International Board Member (on behalf of Red Hat and IBM) and Ambassador, and Red Hat's Brian "Redbeard" Harrington in a frank and open discussion about the RISC-V software ecosystem. We’ll cover the RISE Project, Red Hat’s development efforts, and what additional work is needed to advance RISC-V as a commercially viable architecture for AI and many other workloads.

The road to Extra Packages for Enterprise Linux (EPEL) 10

Information

Monday, May 11 | 3:15 PM - 3:35 PM EDT | B310 - Level 3

Carl George

The Extra Packages for Enterprise Linux (EPEL) initiative plays a critical role in extending Red Hat Enterprise Linux (RHEL) by delivering thousands of community-maintained packages to enterprise users. Historically, each EPEL followed the corresponding RHEL major release, delaying package availability and adoption of new RHEL versions. EPEL 9 shifted to a new approach, initially building against CentOS Stream 9 to prepare a large number of packages for RHEL 9. EPEL 10 will now also use that approach for all RHEL 10 minor versions. Join this session to explore how EPEL 10 builds on learned insights to deliver a more collaborative, future-ready ecosystem for RHEL users and contributors, including

• How past challenges in EPEL packaging and release timing informed the design of EPEL 10.
• How EPEL 10 uses CentOS Stream 10 to align with the RHEL development lifecycle.
• What technical and organizational changes are needed to make this new approach a reality.
• How this new structure affects packager workflows, user adoption, and ecosystem reliability.

CentOS Stream changes: Konflux, draft builds, and more

Information

Monday, May 11 | 3:40 PM - 4:00 PM EDT | B311 - Level 3

Troy Dawson

During the past year, the CentOS Stream pipeline has undergone some changes. While the principle of being as open and transparent as possible is still the same, we’ve explored the CentOS Stream and Red Hat Enterprise Linux pipelines to find opportunities to enhance the user experience. Join us to learn what’s changing, what’s staying the same—and how these changes might affect you.

From upstream to Day 0: How the CentOS AIE SIG enables Red Hat Enterprise Linux for NVIDIA

Information

Monday, May 11 | 4:15 PM - 4:35 PM EDT | B308-B309 - Level 3

Amnon Ilan • Scott Herold

Preparing enterprise Linux for the latest AI accelerators has required weeks of manual driver integration, NVIDIA CUDA Toolkit configuration, and validation—creating delays between silicon availability and production readiness. The CentOS Accelerated Infrastructure Enablement (AIE) Special Interest Group (SIG) was created to help close this gap. In this session, we'll show how NVIDIA and Red Hat are collaborating upstream in the AIE SIG to build the foundation for Red Hat Enterprise Linux for NVIDIA, a purpose-built Red Hat Enterprise Linux variant delivering Day 0 support for the NVIDIA Vera Rubin platform. We’ll explore:

• How the AIE SIG serves as the upstream build and test pipeline for two-way vendor merge between CentOS Stream and NVIDIA's drivers and CUDA (Compute Unified Device Architecture) Toolkit.
• What Day 0 support for Vera Rubin means in practice: OpenRM kernel modules, CUDA Toolkit, and AI libraries available from the moment hardware ships.
• How upstream open source collaboration translates directly into production-ready downstream AI infrastructure.

Optimize data management for enterprise workloads using Btrfs

Information

Monday, May 11 | 4:40 PM - 5:00 PM EDT | B311 - Level 3

David Duncan • Neal Gompa

Storage needs for modern enterprise workloads are more complex than ever. Container platforms and AI/ML development pipelines present unique, demanding data requirements, and traditional filesystems struggle to keep up—forcing teams to rely on complex, slow, and expensive solutions. In this session, we’ll share how Btrfs, a solution that uses copy-on-write (COW) architecture, can help solve these challenges for your workloads, including:

• Containers: Discover how the native Btrfs storage driver for Podman provides superior, block-level space efficiency for image layers and supports instant container creation.
• AI/ML: Learn about instantaneous snapshots to eliminate the model checkpointing I/O tax, letting GPU-bound training to continue interruption. We’ll also cover reflinks (fast clones) for iterating multi-terabyte datasets and models in seconds.
• Data integrity: Explore how to use Btrfs features, such as support for fsverity, to provide stronger data authentication.

As Btrfs is not supported in Red Hat Enterprise Linux, we’ll also show how to adopt it for Red Hat’s platform using community builds developed by the CentOS Kernel Modules (Kmods) special interest group (SIG). You'll leave with a clear understanding of how to unlock these powerful storage capabilities for your enterprise systems.