NVIDIA GTC Taipei at COMPUTEX 2026

NVIDIA GTC Taipei is NVIDIA's AI conference in Taiwan, held alongside COMPUTEX 2026. NVIDIA describes the event as a gathering for developers, researchers and industry leaders exploring advances across AI infrastructure, models, applications and physical systems.

Unlike a broad technology trade show, GTC Taipei is centred on NVIDIA's accelerated computing platform and the organisations building with it. The event combines a CEO keynote, technical learning, product demonstrations and access to NVIDIA's partner ecosystem.

What usually happens at NVIDIA GTC Taipei

At GTC Taipei, NVIDIA hosts a keynote, technical sessions, full-day workshops, training labs, certification opportunities, demonstrations and developer activities.

For 2026, NVIDIA has confirmed more than 50 sessions covering agentic AI, reasoning AI, AI factories, scaling infrastructure, AI for science, physical AI and robotics. The official schedule also lists more than 70 hands-on training lab sessions, NVIDIA technical certifications, the NVIDIA Demo Showcase and Build-a-Claw.

When is NVIDIA GTC Taipei at COMPUTEX 2026 happening

NVIDIA GTC Taipei at COMPUTEX 2026 is scheduled for June 1 to June 4, 2026, in Taipei, Taiwan.

The event begins with Jensen Huang's keynote at Taipei Music Center on Monday, June 1, at 11:00 a.m. Taipei Time. NVIDIA says Huang will present breakthroughs driving the next generation of AI. Online viewers can watch the keynote livestream for free.

Full-day workshops take place on Tuesday, June 2, followed by conference sessions on Wednesday, June 3, and Thursday, June 4, at the Taipei International Convention Center.

NVIDIA GTC Taipei 2026: Latest Updates

Jensen Huang delivered the GTC Taipei 2026 keynote on June 1 at Taipei Music Center. The two-hour address was broadcast to 70 watch parties across Taiwan. Four major announcements came out of the keynote: Vera Rubin entering full production, the RTX Spark AI PC superchip, a new NVIDIA Agent Toolkit, and the Isaac Groot humanoid robotics reference platform.

Vera Rubin Enters Full Production

NVIDIA announced the Vera Rubin platform is ramping into full production to power agentic AI factories worldwide.

The platform unifies five purpose-built racks into one system for agentic workloads: Vera Rubin NVL72, Vera CPU, Groq 3 LPX, Vera BlueField-4 STX storage and Spectrum-6 SPX Ethernet racks. According to NVIDIA, it delivers 10x agent throughput at scale compared with Grace Blackwell.

The supply chain spans 350-plus factories across 30 countries, with 150 partners in Taiwan. Dell Technologies, HPE, Lenovo, Supermicro, Foxconn, GIGABYTE, ASUS and Wistron are among the confirmed system builders. Mass shipments are scheduled for autumn 2026.

Vera Rubin also introduces Spectrum-X Ethernet Photonics, now in production, delivering 5x better power efficiency and 5x longer AI uptime compared with traditional transceiver networks. CoreWeave, Lambda and Oracle Cloud Infrastructure are among the first adopters.

Huang said at the keynote: "Vera Rubin was built for this moment, an AI factory engine that delivers intelligence at scale."

RTX Spark: NVIDIA Enters the Windows PC Market

NVIDIA announced RTX Spark, a new superchip co-developed with MediaTek and Microsoft for Windows laptops and compact desktops. According to NVIDIA's official announcement, RTX Spark combines a 20-core NVIDIA Grace CPU with a Blackwell RTX GPU featuring 6,144 CUDA cores, connected via NVLink-C2C, with up to 128 GB of unified memory and 1 petaflop of AI performance.

RTX Spark is NVIDIA's first processor designed to serve as the primary chip in a consumer Windows PC. The chip runs local AI agents natively via a new NVIDIA OpenShell runtime, developed in partnership with Microsoft to give users full control over what agents can and cannot do on their devices.

Adobe confirmed it is rebuilding Photoshop and Premiere from the ground up for RTX Spark, targeting 2x faster AI and graphics performance.

Microsoft announced the Surface Laptop Ultra as one of the first RTX Spark devices. It ships this autumn. ASUS, Dell, HP, Lenovo, MSI, Acer and GIGABYTE are all producing RTX Spark laptops and compact desktops for the same window. NVIDIA expects more than 30 laptops and 10 desktops at launch across all partners.

Jensen Huang said at the keynote: "For forty years, you launched apps. Click. Type. With RTX Spark and Microsoft Windows, you ask and the PC does the work."

NVIDIA Agent Toolkit

NVIDIA announced a revamped Agent Toolkit at GTC Taipei. The toolkit gives enterprises a standard way to build, deploy and run AI agents across cloud, on-premises, PC and robotics environments. According to NVIDIA's blog coverage of the keynote, the computing pattern is the same regardless of whether the agent runtime lives in the cloud, on premises, on a PC or in a robot.

Isaac Groot Humanoid Reference Platform

NVIDIA introduced the Isaac Groot reference platform for humanoid robotics at the event. According to Blockspace reporting on the keynote, the platform is a 6-foot, 150-pound humanoid with 31 degrees of freedom and 25 degrees of freedom per hand, running on NVIDIA's Thor computer. Huang said at the keynote: "The age of robotics starts here."

NVIDIA AI for Media

NVIDIA announced new AI for Media capabilities at GTC Taipei, targeting broadcasters, streaming platforms and developers building real-time video pipelines. New tools include multilingual LipSync with French, German and Spanish support, an enhanced Active Speaker Detection capability, NVIDIA RTX Video Super Resolution, and the NVIDIA Synthetic Video Detector, a NIM microservice that identifies AI-generated video with roughly 92% accuracy in as little as 22 milliseconds.

Nemotron Sovereign AI Expansion

NVIDIA announced new additions to the Nemotron open model family, alongside new regional datasets developed with partners in Vietnam and El Salvador. According to NVIDIA's official update, NAVER Cloud, Viettel AI, Sarvam and ABEJA are among the organisations using Nemotron models and NeMo libraries to build sovereign AI for local languages and regulated industries.

Taiwan Investment and Constellation Campus

Huang confirmed during the week preceding the keynote that NVIDIA expects to spend $150 billion per year in Taiwan. According to ServeTheHome's event coverage, construction is set to begin on NVIDIA's Constellation campus in Taiwan. The announcement underlines Taiwan's role as the centre of NVIDIA's manufacturing ecosystem.

NVIDIA received COMPUTEX 2026 Best Choice Awards

NVIDIA reported that Vera Rubin NVL72, Jetson Thor and Alpamayo received COMPUTEX 2026 Best Choice Awards.

According to NVIDIA, Vera Rubin NVL72 received a Golden Award and the Sustainable Tech Special Award. Jetson Thor received a Golden Award, while Alpamayo received the Vehicle Technology and Smart Cockpit Category Award.

Jensen Huang visited Taiwan's developer community before the event

NVIDIA reported that Jensen Huang visited Meet-a-Claw ahead of GTC Taipei, where developers attended demonstrations, technical talks and networking sessions focused on autonomous agents and OpenClaw.

During the visit, Huang also said that he expected to share an update on NVIDIA's planned Taipei headquarters during his time in Taiwan.

Main topics at NVIDIA GTC Taipei 2026

Agentic AI and reasoning AI

NVIDIA says the event will cover long-running autonomous agents and the systems needed to help them plan, reason and complete tasks over longer periods.

One listed session, Long-Running AI Agents: The Next Breakthrough in Enterprise Work, focuses on the model infrastructure, tools and AI factory foundations needed to support production-ready agents.

AI factories and AI infrastructure

AI infrastructure is a major focus of the event. NVIDIA says its sessions will cover how businesses can move from individual clusters to larger AI factories that build, refine and deploy intelligence across organisations.

Confirmed sessions include AI Factories for Self-Evolving Agents: Powering OpenClaw With NVIDIA and Networking Innovations for Gigascale AI Systems.

Physical AI and robotics

NVIDIA has confirmed sessions on robotic factories, autonomous machines and physical AI.

A featured session led by Deepu Talla, Vice President of Robotics and Edge AI at NVIDIA, focuses on how physical AI can support the design, construction and operation of robotic factories.

AI for science and healthcare

NVIDIA lists AI for science as another major topic, including the use of accelerated computing, simulation and data in biology, climate and materials.

Healthcare sessions include discussions on clinical AI agents, intelligent medical devices and systems that combine AI with physical operations in care environments.

Event experiences

Jensen Huang's keynote

The keynote takes place on June 1 at 11:00 a.m. Taipei Time at Taipei Music Center.

Before the keynote, NVIDIA will run the GTC Live Keynote Pregame beginning at 9:00 a.m. Taipei Time. NVIDIA's official schedule lists conversations on the AI server boom, embedded AI, agentic AI and robotics in Taiwan, and AI infrastructure for Taiwan.

Workshops, training and certification

Full-day workshops will be held on June 2 at the Taipei International Convention Center. NVIDIA says the workshops cover physical AI, robotics, generative AI and large language models.

GTC Taipei Conference and Full-Day Workshop attendees can also take proctored NVIDIA technical certification exams at no extra cost at TICC.

NVIDIA Demo Showcase

The NVIDIA Demo Showcase runs from June 2 to June 4 at the Taipei International Convention Center.

NVIDIA says the showcase will include demonstrations across agentic AI, physical AI and AI infrastructure, with products and solutions for developers, IT leaders, gamers and creators.

Build-a-Claw

Build-a-Claw is an event activity for attendees interested in creating AI agents using OpenClaw and NVIDIA NemoClaw.

NVIDIA says participants can learn how agents run locally, on a cloud virtual machine and at the edge.

COMPUTEX Forum appearances

NVIDIA also lists two appearances at the COMPUTEX Forum. Deepu Talla is scheduled to speak on robotics, autonomous machines and physical AI on June 2. Kevin Deierling, Senior Vice President of Networking at NVIDIA, is scheduled to speak on AI compute, infrastructure and development on June 4.

NVIDIA states that a GTC Taipei Conference pass does not provide access to COMPUTEX keynote or forum sessions.

Why NVIDIA GTC Taipei 2026 matters

NVIDIA GTC Taipei matters because it brings together the software, infrastructure and physical systems required to build and deploy AI.

For developers

The event provides technical sessions, workshops and demonstrations focused on agents, models, robotics and accelerated computing.

For technology leaders

The program gives teams a closer look at NVIDIA's approach to AI infrastructure, networking, deployment and enterprise AI factories.

For manufacturing and robotics teams

Sessions on robotic factories, autonomous machines and physical AI make the event relevant for businesses applying AI in industrial environments.

For product and business teams

GTC Taipei provides useful signals on the tools and infrastructure NVIDIA is promoting across enterprise AI, healthcare, robotics and intelligent products.

For Taiwan's technology ecosystem

Held alongside COMPUTEX, the event connects NVIDIA with Taiwan's technology and manufacturing partners as demand for AI infrastructure grows.

Venue

Keynote venue

Taipei Music Center No. 99, Sec. 8, Civic Blvd., Nangang Dist. Taipei City, 115, Taiwan

Conference, workshops and demo showcase venue

Taipei International Convention Center No. 1, Section 5, Xinyi Rd, Xinyi District Taipei City, 110, Taiwan

NVIDIA confirms that the keynote takes place at Taipei Music Center, while workshops and conference activities run at the Taipei International Convention Center.

Registration and access

NVIDIA states that all in-person GTC Taipei pass types are sold out. Online viewers can still watch Jensen Huang's keynote livestream for free.

Pass types confirmed by NVIDIA:

• Conference Pass: keynote, keynote pregame, sessions, NVIDIA Demo Showcase, one certification attempt, lunch and COMPUTEX exhibit areas (sold out) • Full-Day Workshop + Conference: conference access plus one full-day workshop (sold out) • Keynote Only at Taipei Music Center: keynote, keynote pregame, NVIDIA Demo Showcase and COMPUTEX exhibit areas (sold out) • Keynote Livestream: online keynote viewing (free)

NVIDIA's pricing page lists regular pricing of USD 100 for the Conference Pass and USD 500 for the Full-Day Workshop + Conference package. The Keynote Only pass was listed as free for in-person attendance before passes sold out.