From October 15-17, Intel and our partners are participating in value-packed knowledge sessions, presentations, and demos at the Open Compute Project (OCP) Global Summit in San Jose, California. Intel has been a proud member of the Open Compute Project since its founding in 2011. In the spring of 2024, Jeni Barovian took the lead as Intel’s representative on the OCP board to drive the future of open compute.
Intel has long held many leadership positions within OCP projects and workstreams such as Hardware Management, Server (MHS), Cooling Environments (Cold Plate), Sustainability (Carbon Modeling), and more, and has made 59 contributions to OCP in the forms of specifications, whitepapers, usage guides, and reference architectures. We are excited to share the work being driven by our Intel team at the OCP Global Summit through our Keynote and sponsored sessions, technical workshops, and demos in the Intel booth and OCP’s Innovation Village.
Keynote and Sponsored Sessions
At the 2024 OCP Global Summit, themed "From Ideas to Impact," Intel will share its strategic vision on AI systems, emphasizing the pivotal role of open standards in transforming ideas into tangible technologies. As the tech landscape evolves, Intel remains at the forefront, leveraging the OCP community's expertise to drive sustainable, scalable innovation.
Intel's AI strategy harnesses open standards to ensure every step, from conceptualization to implementation, magnifies impact within the data center industry. Our approach, rooted in open collaboration, accelerates development cycles and meets dynamic industry needs. By embracing community-driven innovation, Intel leads in transforming theoretical concepts into practical technologies that redefine efficiency and sustainability. This keynote will delve into how Intel's AI systems strategy aligns with and promotes the OCP's mission. We'll explore case studies where Intel has implemented AI solutions adhering to open standards, highlighting our commitment to transparency and collaborative progress. Join us as we outline future initiatives to continue leveraging and contributing to the OCP community, advancing technology with shared visions and efforts.
Numerous Cloud Service Providers and enterprise customers alike are harnessing the power of an Infrastructure Processing Unit (IPU) to securely accelerate infrastructure workloads, improve CPU utilization, and enable data center innovations. As modern data centers scale, providing open-source infrastructure software combined with high-performance network accelerators is critical for applications such as storage, HPC, AI training clusters, and developing microservice deployment models.
Intel® IPU Adapter E2100 delivers the hardware assistance for an open-sourced next-generation reliable transport protocol that is compatible with existing RDMA and NVME-based application stacks and with widely deployed lossy Ethernet fabrics. The Intel® IPU Adapter demonstrates a robust commitment to open-source infrastructure software by promoting collaboration and innovation within the development community which is underscored by its provision of open-source implementations for well-known use cases under the Open Programmable Infrastructure Project.
Foundation models are great, but without the integration of enterprise-specific and domain-specific data, enterprises can't extract the most value from Gen AI. RAG bridges the gap between enterprise data and AI models by:
- Distilling data into insights easily by improving LLM outputs;
- Enabling enterprises to customize LLMs on their proprietary data without retraining or fine-tuning;
- Connecting LLMs to a proprietary data knowledge base to provide a practical way for businesses to continuously inject fresh data – reducing hallucinations – while keeping their data secure;
- And enabling organizations to customize and launch generative AI applications more quickly and cost-effectively.
To merge the knowledge base of enterprise data with the efficiency of AI, Intel is thrilled to introduce Intel AI for Enterprise RAG to address the highest-demand enterprise customer needs. Uniquely positioned to bring Intel CPU and accelerator advantages to Intel AI for Enterprise RAG, we are solving enterprise RAG deployment challenges through scalability, ease of use, improved TCO, security, and openness. Our complete RAG solutions bring together the entire stack needed for a comprehensive solution. All making it easy for enterprise customers to implement solutions based on Intel AI for Enterprise RAG.
Intel Booth (Location #B4)
Stop by and experience firsthand how Intel is shaping the future of technology. Try our new solutions, network with our experts, and get an exclusive look at how we’re pushing the boundaries of innovation.
If you're on site, participate in Intel's Innovation Pass and earn points to win prizes.
OCP Innovation Village
Station 3: Composable Memory System (CMS)
This station showcases the latest technologies within the CMS project, including Architectural enablement using external PCIe cabling, the cabling ecosystem, and interoperability. CXL-based solutions driven by the OCP CMS subproject demonstrate that these solutions are deployable and that the multi-vendor ecosystem is working to make them scalable. Elastic memory takes pooling to the next level with true automation and management across many servers connected to the memory appliance.
Station 4: Server
At our station, we will showcase advanced Open-Source Firmware solutions, including network-based booting without firmware and secure OpenBMC images. Attendees will see the M-CRPS's ability to combine power supplies from different vendors and customize configurations. We'll also present the first prototype of the DC-SCM 2.1 Type-1 Module and demonstrate DC-SCM 2.x tunneling solutions for multiple nodes. Our display includes both 1U and 2U DC-MHS-inspired chassis, featuring live swaps of DC-HPMs and DC-SCMs. Visitors can explore the security features of the Axiado DC-SCM and new technologies like a Large Memory System layout and E3.S CXL memory.
Station 5: Time Appliances Project (TAP)
The TAP Station at Innovation Village will be a plugfest to test the interoperability of the innovations in this group. Time Appliances Project (TAP) aims to provide a platform to bring together, discuss, standardize, and share technologies and solutions across industries with the data center applications and data center network infrastructure as the main interest. The project aims to bring together the community of data center operators, application developers, and equipment and semiconductor companies together to enable data center time-sensitive applications such as consistency in distributed systems, edge computing, AR/VR, and IoT. These applications will greatly benefit from high accuracy, reliable, and scalable distribution and synchronization of time.
Station 6: OCP Networking/NIC
At this station, participants will showcase AI servers built with open ecosystem components, including CPUs, GPUs, and NICs, connected via Ethernet switches. Attendees will see standard AI benchmarks and applications running on these servers, highlighting how the open ecosystem facilitates scalable AI systems. We’ll feature next-generation servers designed for AI at scale, demonstrating community-driven design principles. Our compact appliance will accommodate OCP SFF and DSFF NICs, with interactive samples to illustrate their versatility. Additionally, we will demonstrate Falcon's full stack performance for HPC and AI/ML using IPU hardware and showcase the new Dual Small Form-Factor (DSFF) NIC specification with physical samples integrated into a DC-MHS motherboard, complemented by a digital presentation highlighting the updates.
Technical Workshops:
Build your agenda and learn and meet Intel experts by searching Intel in the live agenda tool.
Meet with Intel Experts & Leaders
Reach out to your Account Executive to schedule a meeting with Intel in our meeting room in the Marriott hotel.
Intel Leadership Attending
Follow Us on Social Media
Join us on social media @intelbusiness and @Intel and follow the hashtags #OCPSummit24 and #FromIdeastoImpact.
Notices and Disclaimers
Performance varies by use, configuration, and other factors. Learn more on the Performance Index site.
Performance results are based on testing as of dates shown in configurations and may not reflect all publicly available updates. See backup for configuration details. No product or component can be absolutely secure.
Your costs and results may vary.
Intel technologies may require enabled hardware, software, or service activation.
© Intel Corporation. Intel, the Intel logo, and other Intel marks are trademarks of Intel Corporation or its subsidiaries. Other names and brands may be claimed as the property of others.