NVIDIA is offering GTC attendees the opportunity to create and deploy personalized, always-on AI assistants, dubbed “claws,” through the OpenClaw project, currently the fastest-growing open source initiative. From March 16-19 within the GTC Park, participants can customize these proactive agents, defining their personalities and granting access to necessary tools with support from NVIDIA experts. These AI assistants, which can run on cloud compute or local accelerated computing via NVIDIA DGX Spark or GeForce laptops, are designed for a wide range of tasks, from scheduling to coding. NVIDIA suggests that these assistants are like “the personal assistant you’ve always wanted, reachable via your preferred messaging app,” highlighting their potential for continuous learning and proactive assistance. NVIDIA will also be selling DGX Spark and Jetson modules at the event.
OpenClaw Agents Deployed with NVIDIA DGX & GeForce
The rapidly expanding OpenClaw project has reached a new stage of accessibility with the deployment of customizable AI agents alongside NVIDIA’s DGX and GeForce hardware. NVIDIA is facilitating this process with expert support, allowing participants to define an agent’s personality and grant it access to necessary tools, regardless of their technical expertise. These agents are versatile, capable of handling tasks ranging from calendar management and vacation planning to workout recommendations and even software coding; users can tailor them as specialists or generalists, enabling continuous learning and proactive suggestions. Individuals can utilize cloud compute resources provided onsite or leverage the acceleration capabilities of their own NVIDIA DGX Spark or GeForce laptops to power these agents. NVIDIA is also making DGX Spark and Jetson modules available for purchase, further supporting the development and deployment of these local-first AI agents. Developers can consult the newly released OpenClaw Playbook, a step-by-step guide designed to run OpenClaw on DGX Spark, enabling the creation of AI assistants that directly interact with files, applications, and workflows without reliance on cloud connectivity.
GTC Keynote & Panel Discussions with Industry Leaders
The convergence of artificial intelligence and practical application is a central focus at the upcoming GTC event, extending beyond theoretical discussions into hands-on agent creation. The aim is to create an “always-on” assistant, accessible through messaging apps, capable of tasks ranging from scheduling to suggesting workout routines. The event’s keynote, delivered by NVIDIA founder and CEO Jensen Huang on March 16 at 11 a.m. PT, will address the entire AI stack, from hardware to applications, and is expected to draw thirty thousand attendees from 190 countries. The pregame show, featuring the CEOs of Perplexity, LangChain, Mistral, Skild AI and OpenEvidence, starts at 8 a.m. PT on Monday, two hours before Huang takes the stage.
A subsequent panel on March 18 will focus on open versus closed AI models, with Harrison Chase, cofounder and CEO of LangChain, joining leaders from A16Z, AI2, Cursor, and Thinking Machines Lab. The event description states that “What it means for everyone building on top of them” will be a central theme. Beyond the keynotes, GTC will host discussions on AI’s impact on climate research, featuring Dario Gil, Undersecretary at the U.S. Department of Energy, and explore the intersection of music and AI with Universal Music Group chairman and CEO Sir Lucian Grainge.
NVIDIA GTC Event Details: Registration, Locations & Activities
From 8 a.m. to 5 p.m. daily within the GTC Park, excluding times during the Monday keynote, NVIDIA experts will guide individuals in customizing and deploying these proactive AI tools, regardless of their technical expertise. NVIDIA provides cloud compute resources or allows attendees to utilize local accelerated computing with DGX Spark or GeForce laptops to facilitate this hands-on experience. The event features over 700 sessions covering topics ranging from physical AI to inference, and Jensen Huang’s keynote on March 16 at 11 a.m. PT will outline the complete technological stack. Thirty-thousand attendees representing 190 countries are expected, and the keynote will also stream freely from nvidia.com for virtual participation. Registration for the event, running through Thursday afternoon, is available at nvidia.com/gtc, with a 20% discount offered using code GTC26-20.
It’s a buildout measured in gigawatts.
