NVIDIA's Fourth Quarter Financial Report: Record-Breaking for Three Consecutive Quarters Amid Surge in Global AI Demand


The CEO of Europe's largest AI startup, Mistral AI, stated that developing custom AI chips is a trend in the long term. With the surge in computing power demand for large models, tech giants and unicorns like OpenAI are increasingly developing their own chips or collaborating with semiconductor companies to reduce costs.
At NVIDIA's employee meeting, some employees were concerned that using AI intensively might only be superficial. CEO Huang Renxun responded by encouraging employees to embrace new technology, pointing out that when facing new tools, perfection is not necessary, and the key is to take the first step. He emphasized that it is acceptable to slightly waste money when exploring AI, but time must never be wasted, as AI is far from being superficial.
NVIDIA open-sources Polar, a reinforcement learning training framework. Its core innovation allows mainstream code agents like Codex and Claude Code to integrate GRPO training without modifying native code. It addresses industry pain points in evolving agents from single-step tasks to complex long-flow tasks (e.g., repository-level modifications, OS interactions), breaking down barriers in agent reinforcement learning.....
OpenRouter raised $113M in Series B funding, led by Alphabet's CapitalG, with participation from NVIDIA, ServiceNow, MongoDB, Snowflake, Databricks, a16z, and Menlo Ventures. The platform transformed from an experimental field to a productivity cornerstone amid explosive business growth in six months.....
NVIDIA announced the establishment of an AI research lab in Singapore, focusing on embodied intelligence to advance automation and robotics. This technology integrates AI with physical systems, offering significant potential in manufacturing and robotics. The lab aims to improve AI model training efficiency and reduce costs.....