2025 OCP APAC Summit

2025 OCP APAC Summit

arrow_back View All Dates

9:00am PDT

Keynotes TaiNEX 2 - 701 CD

12:00pm PDT

Lunch Sponsored by ASE Taipei Nangang Exhibition Center Hall 2

1:00pm PDT

Architecting the AI Fabric: Scalable Networking for Next-Generation AI Servers, Racks, and Clusters TaiNEX2 - 701 F Introduction of IOWN Global Forum TaiNEX 2 - 701 E Opening Track Overview / APAC Based Project TaiNEX2 - 701 G

1:20pm PDT

Revisit RoCEv2 issues in large scale deployment and the future that UEC promise TaiNEX2 - 701 F Introduction to OCP for (new) IOWN audience TaiNEX 2 - 701 E

1:40pm PDT

Enhancing Financial Intercommunication through IOWN/APN: The Future of Data Center Connectivity TaiNEX 2 - 701 E Evolving FBOSS to support Generative AI network workloads TaiNEX2 - 701 F Use Cases for CXL RAS Firmware-First Error Handling TaiNEX2 - 701 G

1:55pm PDT

World first international APN and RDMA-over-APN evaluation TaiNEX 2 - 701 E

2:00pm PDT

New paradigm for lossless DCI interconnect in the AI/ML era TaiNEX2 - 701 F

2:10pm PDT

APN-T 400G Muxponder for Cloud Edge Computing TaiNEX 2 - 701 E

2:20pm PDT

Dynamic ECN Threshold Testing Methodology and the Importance of qp-fairness TaiNEX2 - 701 F Architecting AI Storage: A Strategy for High-Performance and Tenant Object Storage TaiNEX2 - 701 G

2:25pm PDT

Operational challenges for realizing a multi-vendor CDI configuration TaiNEX 2 - 701 E

2:35pm PDT

Fast and Accurate Heat Dispersion Modeling for Data Centers TaiNEX2 - 703

2:40pm PDT

Tenant Aware VM and Container Integrated Layer 2 Networks TaiNEX 2 - 701 E e-Tube Technology - Breaking Interconnect Barriers To Accelerate AI Cluster Scale Up TaiNEX2 - 701 F Optimizing I/O for AI Workloads and Boosting Checkpointing with FDP TaiNEX2 - 701 G

3:00pm PDT

Break Taipei Nangang Exhibition Center Hall 2

3:15pm PDT

Remote GPU over APN/Confidential computing TaiNEX 2 - 701 E NVMe TaiNEX2 - 701 G Advancements in 400G/Lane SerDes for High-Speed AI Connectivity TaiNEX2 - 703 Panel: End-to-End Observability Across Network and Compute Layers for AI Workload Optimization TaiNEX2 - 701 F

3:30pm PDT

Ideal AI/ML optical interconnect solution for sustainable data centers TaiNEX 2 - 701 E

3:40pm PDT

Pioneering AI Cooling: High Power Cold Plates with 200W/cm² Heat Flux for Next-Gen AI ASICs TaiNEX2 - 703

3:45pm PDT

KV Cache Sharing over IOWN APN: Building a Sustainable and High-Performance Nation-Wide Distributed AI for LLM Inference TaiNEX 2 - 701 E Proactive Link Management in AI Networks: Lessons from Meta TaiNEX2 - 701 F

4:00pm PDT

Towards Energy-Efficient AI Infrastructure through the Integration of Computing and Networking TaiNEX 2 - 701 E

4:05pm PDT

Adaptive Multi-Tenant Orchestration of AI Fabric TaiNEX2 - 701 F Thermal-Safe Operation for GPU Under Power Constraints Using Reinforcement Learning TaiNEX2 - 703

4:15pm PDT

Optical circuit switching technology for future AI clusters TaiNEX 2 - 701 E Unlocking Next-Generation Data Center Performance: Leveraging CXL Memory for Scalable AI and Database Workloads TaiNEX2 - 701 G

4:25pm PDT

Scale-up AI Networking Alternatives - Comparing UALink, SUE and NVLink TaiNEX2 - 701 F

4:30pm PDT

Reference Implementation of an SDN Controller for Open Optical-circuit-switched AI Clusters TaiNEX 2 - 701 E Tensor Contraction as a First-Class Primitive: Learnings from Building Efficient and Programmable AI Compute TaiNEX2 - 703

4:35pm PDT

Cooling Presentation TaiNEX2 - 701 G

4:45pm PDT

State of the Debate: Pluggable, Linear vs Copackaged Optics TaiNEX 2 - 701 E

5:00pm PDT

Welcome Reception Sponsored by Meta Taipei Nangang Exhibition Center Hall 2

Need help? View Support Guides
Event questions? Contact Event Planner

Powered by Sched Event Planning Software

©2025 Sched • About • Privacy • Terms