NVIDIA
8 articles found in this topic.
NVIDIA Unveils Nemotron 3 Open-Source Model Family, Featuring Hybrid Mamba-Transformer Architecture
NVIDIA introduces its Nemotron 3 family of open-source models, featuring a hybrid Mamba-Transformer architecture. The Nano version is available now, with Super and Ultra variants planned for 2026. This strategic move expands NVIDIA's role beyond hardware into large language model development, offering efficient solutions for various applications.
Jensen Huang Named FT Person of the Year 2025 as Nvidia Market Cap Exceeds $5 Trillion
Jensen Huang, Nvidia's CEO, was named FT Person of the Year 2025 as Nvidia's market cap exceeded $5 trillion. His leadership, rooted in a 'masochistic' culture and resilience through hardship, has driven the company's success and shaped the global AI economy. Huang's strategic gambles, like CUDA, proved pivotal for Nvidia's growth.
Nvidia CEO Jensen Huang Manages 36 Direct Reports in Flat Structure
Nvidia CEO Jensen Huang directly manages 36 executives, a flat management structure designed for efficient information flow and rapid decision-making. This approach, which deviates from typical Silicon Valley norms, emphasizes direct access and innovation within the $5 trillion company.
NVIDIA's 8B Orchestrator Model Surpasses GPT-5 in Efficiency and Cost
NVIDIA's new 8B Orchestrator model outperforms GPT-5 in efficiency and cost-effectiveness, achieving higher accuracy on benchmarks like Humanity's Last Exam while significantly reducing operational expenses. This model optimizes AI tool usage through a lightweight scheduling hub, coordinating specialized tools for enhanced performance.
Nvidia Lobbying Effort Halts US AI Chip Export Restriction Bill
Nvidia successfully lobbied against the GAIN AI Act, a proposed defense bill provision that would have restricted advanced AI chip sales to prioritize American customers. CEO Jensen Huang's intervention led the White House to side with Nvidia, preventing the bill's inclusion in defense legislation.
NVIDIA Unveils Major CUDA 13.1 Update and Secures AGI Competition Win
NVIDIA has released CUDA 13.1, its most significant update in two decades, introducing the CUDA Tile programming model for next-gen GPUs. Concurrently, NVIDIA's KGMoN team won the Kaggle ARC Prize 2025, showcasing high-performance AGI with a 4-billion-parameter model.
Argos Translate: An Open-Source, Offline Translation Engine
Argos Translate introduces an open-source, offline translation engine, offering comprehensive language translation without internet reliance. It supports a wide range of languages and provides flexible deployment options including Python library, CLI, and GUI. The system also features GPU acceleration for enhanced performance.
Google's TPU Strategy Challenges NVIDIA's Dominance in AI Hardware
Google's Tensor Processing Units (TPUs) are emerging as a strong competitor to NVIDIA in the AI hardware market, especially after the performance of its Gemini 3 Pro model. TPUs offer significant energy efficiency and specialized architecture for AI workloads. Google's shift to selling its TPUs, with Meta reportedly interested, signals a potential disruption to NVIDIA's long-standing dominance.