Where compute, models, and hardware meet. A curated calendar of the major events shaping how the industry measures, understands, and optimizes the AI stack — workloads × models × hardware infrastructure.
AI infrastructure's ultimate stage — 8,000 attendees, 400 speakers, and 8 stages spanning Data & Models, Compute, Data Movement, AI Data Center, and Physical AI. Where inference meets influence, and where efficiency becomes the deciding metric.
ai-infra-summit.com →OCP's Korea community day — memory, storage, and open hardware from the region that supplies much of the silicon underneath the AI stack.
opencompute.org →The symposium on high-performance chips, where the accelerators, memory, and interconnect that run AI get unveiled — the silicon whose real-world energy efficiency gets measured downstream.
hotchips.org →OCP's inaugural Southeast Asia tech day, hosted inside Data Centre World Asia — open infrastructure meeting the region's fast-growing AI data center build-out.
opencompute.org →Two days on the AI frontier — models, agents, and the infrastructure that runs them, where deployment cost and efficiency increasingly set the agenda.
aiconference.com →The Linux Foundation's flagship open source gathering — the projects and communities, from Linux to Kubernetes, that form the software substrate of AI infrastructure.
events.linuxfoundation.org →The Open Compute Project's flagship for open data center hardware — 800 VDC power, liquid cooling, and interconnect standards. The open foundation beneath efficient, measurable AI infrastructure.
opencompute.org →The Linux Foundation's summit for agentic AI and the Model Context Protocol — the agent workloads whose token and compute cost define the next efficiency frontier.
events.linuxfoundation.org →CNCF's flagship conference, and the platform layer that schedules AI workloads across clusters — where cloud native meets the infrastructure behind modern AI systems.
events.linuxfoundation.org →The international conference for high-performance computing, networking, storage, and analysis — where compute at extreme scale, and the power it draws, sets the frontier for AI training and inference.
sc26.supercomputing.org →AWS's flagship cloud and AI conference — the instances, accelerators, and services on which a vast share of AI workloads run, and where their cost gets measured.
reinvent.awsevents.com →The Linux Foundation's premier open source event in Japan — open infrastructure and community across APAC powering modern AI systems.
events.linuxfoundation.org →The premier machine learning research conference — the models and methods whose deployment cost, in joules per token, Aitra exists to attribute and optimize.
neurips.cc →NVIDIA's accelerated-computing gathering, shaping the GPUs and rack-scale systems that define modern AI — the hardware side of every workload × model × infrastructure equation.
nvidia.com/gtc →The Open Compute Project's European gathering — open data center hardware, power, and cooling standards brought to the operators building efficient, measurable AI infrastructure across EMEA.
opencompute.org →CNCF's Japan gathering, where we ran Aitra Meter live — metering AI inference from PDU to pod. The cloud native platform layer that schedules AI workloads, read our recap.
events.linuxfoundation.org →OCP's Asia-Pacific summit at the heart of the hardware supply chain — the silicon, systems, and data center standards that make AI infrastructure efficient and measurable.
opencompute.org →A curated selection focused on AI infrastructure — compute, data center, power, and the models that run on them. Dates are confirmed against organizer sources; some 2026–27 details may shift. Spot something we should add? Find us in #soda-aitra on the SODA Foundation Slack.