NVIDIA's Leadership and Near-Monopoly in Artificial Intelligence Technology is 'Staggering'


According to data from the National Data Administration, by 2025, the total volume of AI training and inference data in China is expected to reach 199.48 EB, representing a year-on-year growth of 42.86%. Among this, inference data has exceeded training data for the first time, reaching 101.34 EB, indicating the wide application of AI across industries. At the same time, the data volume generated by system software and AI is expected to reach 26.92 ZB.
Nvidia unveils the open multimodal model Nemotron 3 Nano Omni, integrating video, audio, image, and text reasoning. It uses a 30B-A3B mixture-of-experts architecture with built-in vision and audio encoders, eliminating extra perception models. This enhances large-scale inference efficiency and excels in complex text processing.....
Nvidia unveils Nemotron3Nano Omni, an open multimodal model integrating video, audio, image, and text reasoning with a 30B-A3B MoE architecture and built-in vision/audio encoders for faster, smarter developer interactions.....
NVIDIA VP Wu Xinzhu recently revealed the company's full-stack 'five-layer cake' ecosystem for autonomous driving, covering chips, systems, algorithms, data, and cloud platforms. L4 autonomy is expected to be realized by 2025. NVIDIA shifts from chipmaker to a 'three-computer' model, enabling in-vehicle inference, training, and simulation synergy, accelerating physical AI commercialization in mobility.....
Meituan has recently launched a trillion-parameter AI large model test. The model is trained entirely on domestic computing power clusters, marking a significant breakthrough in the application of domestic technology. It is currently only available to invited users and has not been widely released, demonstrating Meituan's leading position in the AI field.