Question · Q4 2026
Atif Malik questioned the increasing importance of CUDA, particularly as a growing share of AI investment dollars are directed towards inference workloads.
Answer
Jensen Huang, President and CEO of NVIDIA, emphasized CUDA's critical role in inference, citing TensorRT-LLM and NVLink Switch's ability to deliver significant performance per watt and cost reductions. He highlighted that inference performance directly translates to customer revenues due to the exponential growth of token generation by agentic AI.
Ask follow-up questions
Fintool can predict
NVDA's earnings beat/miss a week before the call


