The way forward for AI is touchdown in Taiwan. At NVIDIA GTC Taipei at COMPUTEX, the world’s builders, researchers and trade leaders are converging to dive into the most recent breakthroughs shaping each trade, masking matters spanning AI factories and scaling infrastructure to agentic and bodily AI and extra.
Hear from NVIDIA founder and CEO Jensen Huang reside on stage at Taipei Music Heart on Monday, June 1, 11 a.m. Taiwan time. Tune in early to catch the GTC Reside at Taipei 2026 pregame present, that includes vigorous conversations with trade leaders concerning the newest improvements in AI and accelerated computing.
That is the place to search out all the most recent — keep tuned to the weblog for reside updates.
Thursday, Might 21, 9 a.m. PT 🔗
NVIDIA Wins COMPUTEX 2026 Greatest Selection Awards for Improvements Spanning AI Factories, Robotics and Autonomous Automobiles
NVIDIA Vera Rubin NVL72, NVIDIA Jetson Thor and NVIDIA Alpamayo have been honored throughout 4 classes at Asia’s premier expertise and laptop commerce exhibition.
At this yr’s COMPUTEX Greatest Selection Awards (BCA), NVIDIA at present obtained honors recognizing its innovation in AI computing, built-in circuits and autonomous car (AV) improvement.
The NVIDIA Vera Rubin NVL72 rack-scale AI supercomputer gained a Golden Award and the Sustainable Tech Particular Award; the NVIDIA Jetson Thor platform for edge AI and robotics gained a Golden Award; and the NVIDIA Alpamayo open platform for AV improvement gained the Automobile Know-how and Good Cockpit Class Award.
Entries have been evaluated on their performance, innovation and market potential, showcased on the premier laptop and expertise commerce exhibition.
Jensen Huang, founder and CEO of NVIDIA, will ship a keynote at COMPUTEX on Monday, June 1, at 11 a.m. Taiwan time.
NVIDIA Vera Rubin NVL72 Takes Residence COMPUTEX Awards
Securing a Golden Award and the Sustainable Tech Particular Award, Vera Rubin NVL72 connects 36 NVIDIA Vera CPUs and 72 NVIDIA Rubin GPUs — unified by the sixth-generation NVIDIA NVLink Change for scale-up — with ConnectX-9 SuperNICs and Spectrum-X Ethernet Photonics co-packaged optics switches for scale-out and scale-across, in addition to BlueField-4 DPUs to speed up knowledge processing throughout storage and safety.
Vera Rubin NVL72 delivers as much as 10x greater inference efficiency per watt and 10x decrease value per token. When paired with NVIDIA Groq 3 LPX, Vera Rubin NVL72 delivers as much as 35x greater throughput per watt for trillion-parameter fashions.
Designed for agentic AI, reasoning and long-context workloads, it allows AI factories to scale intelligence contained in the rack and throughout the information heart with safe, repeatedly accessible deployment.
The Vera Rubin NVL72 units the bar for scalability, resiliency and sustainable AI infrastructure. Its cable-free, hose-free, fanless modular tray design reduces meeting time from two hours to 5 minutes per compute tray.
The system’s energy cabinets ship 6x extra onboard power storage for clever energy smoothing, defending each the rack and the broader energy grid from steep load swings. As well as, its 100% liquid-cooled structure operates at 45 levels Celsius, that means it drops seamlessly into present liquid-cooled knowledge facilities and allows ambient-air, dry-cooler designs that redirect energy from cooling overhead into token technology.
Extra BCA Wins for NVIDIA Applied sciences
NVIDIA Jetson Thor gained a Golden Award as essentially the most highly effective edge AI compute platform constructed for bodily AI and autonomous robots. Powered by the NVIDIA Blackwell GPU structure, it delivers as much as 2,070 FP4 teraflops of AI efficiency — 7.5x the compute and three.5x the power effectivity of the earlier NVIDIA Jetson Orin technology — in a compact module configurable between 40 and 130 watts.
Already in manufacturing throughout a whole bunch of functions, Jetson Thor is constructed to carry generative AI to sensible robots, industrial techniques, medical units and autonomous machines whereas maximizing run-time efficiency and reminiscence optimization.
Plus, NVIDIA Alpamayo gained the Automobile Know-how and Good Cockpit Class Award for pioneering open, reasoning-based autonomous car improvement. Alpamayo is designed to assist builders sort out uncommon, advanced long-tail driving situations — similar to deciphering an ambiguous hand sign from a pedestrian, figuring out the right-of-way when visitors lights and street markings contradict one another, and safely passing an emergency car parked partially within the lane forward — which fall exterior typical coaching expertise
The Alpamayo open platform contains Alpamayo 1.5 and Alpamayo 1, 10-billion-parameter chain-of-thought reasoning imaginative and prescient language motion fashions for AV analysis; AlpaSim, an open supply, end-to-end simulation framework for high-fidelity AV improvement; and NVIDIA Bodily AI Open Datasets, which embrace greater than 1,700 hours of driving knowledge throughout geographies and situations.
Study extra about NVIDIA’s newest improvements at NVIDIA GTC Taipei, operating June 1-4 at COMPUTEX.