According to @FuSheng_0306, Nvidia's quarterly revenue reached $68.1 billion, with CEO Jensen Huang emphasizing the critical role of computational power in driving token usage and revenue. Fu Sheng highlighted his personal experience with AI agents running continuously, consuming significant token resources but yielding high productivity in tasks like writing and research. This underscores the growing demand for AI-driven applications and Nvidia's pivotal role in supporting this ecosystem.

According to Sundar Pichai, Gemma 3 introduces groundbreaking efficiency in AI models, with its largest 27B model operating on just one H100 GPU. This represents a significant leap, requiring at least 10 times less computational power for similar performance compared to other models, potentially reducing operational costs and energy consumption for traders and developers.