newsroom
AI Infrastructure · Vera Rubin's 35x performance leap and Feynman roadmap extend Nvidia's compute lead
now playing · AI Infrastructure
AI Infrastructuretailwindscore 9/10ejaaz
Vera Rubin's 35x performance leap and Feynman roadmap extend Nvidia's compute lead
Nvidia's Vera Rubin platform delivers 35x performance over prior generation and 10x performance-per-watt versus Blackwell, with Feynman already in pipeline giving 18-month roadmap visibilit…
Autonomous Vehiclestailwindscore 8/10ejaaz
Nvidia's modular FSD stack targets non-Tesla OEMs with BYD and Uber partnerships
Nvidia is offering a complete sensor-to-software autonomous driving stack to automakers like BYD, Nissan, Hyundai, and Geely, plus Uber's ride-hail network. This modular approach contrasts…
Space Economytailwindscore 7/10josh kale
Orbital AI data centers with Vera Rubin chips could slash compute cost per watt
Deploying Nvidia's Vera Rubin architecture in space via Starship launches could exploit vacuum cooling and continuous solar power to dramatically reduce energy costs for AI training and inf…
DGX Spark brings 20 petaflops and 750GB memory to desktop for $4,700
Nvidia's DGX Spark compresses Grace Blackwell architecture into a desk-side form factor capable of running trillion-parameter models locally, democratizing high-end AI compute for developer…
DLSS 5 rendering tech transfers to robotics and autonomous driving simulation
The same neural rendering technology behind DLSS 5's game graphics enhancement powers Nvidia's robotics simulation (Groot) and autonomous driving training, creating a cross-domain flywheel…
Enterprise AI Adoptiontailwindscore 6/10ejaaz
Nemo Claw secures open-source coding agents for enterprise deployment
Nvidia's Nemo Claw wraps Open Claw's viral coding agent with enterprise-grade security, enabling companies to deploy autonomous software agents without data leakage risks. This complementar…