Now it can be seen that Lao Huang's strong support for open source at the end of July is also paving the way for Nvidia's own open source model. According to a report in The Wall Street Journal, Nvidia paid $6 billion to non exclusively license Poolside's "Model Factory" technology - the core software system used by Poolside to build its Laguna series of open-source weight encoded AI models. Nvidia has issued job offers to over 100 Poolside engineers who will join Nvidia. Focusing on its Nemotron open-source weight model project (launched by Dazi in 2023, currently developing larger and more advanced versions with rumored parameter scales reaching trillions). 1. It seems that Dazi plans to use this transaction to recreate one of the world's most powerful open-source weighted AI models, as a challenger to China's open-source models in the United States. In fact, this also directly challenges American closed source cutting-edge companies such as OpenAI and Anthropic, as open-source weight models have lower operating costs and are easier to customize. 2. This transaction is also a landmark move by Nvidia to further delve into "gold mining" (software/model layer) from "selling shovels" (GPU hardware) to "selling AI factories". Lao Huang wants to bind the entire upstream and downstream to NVIDIA's ecosystem. 3. NVIDIA's ecological moat is further deepening, but it is also intensifying competition with customers: Nvidia hardware has taken the lead in training/inference. Now strengthening the ability of open source models can better drive their GPU requirements (more models → more computing power). But Nemotron directly competes with closed source models from Nvidia's major clients such as OpenAI and Anthropic, which may cause tension. Large model manufacturers develop their own chips, while chip manufacturers develop their own models. The competition has become more intense 4. The wave of open-source weight models is accelerating, and the core is still to promote the open-source AI ecosystem in the United States and counter China's open-source advantages. Helps establish strong and customizable open-source alternatives in the United States. This may lower the threshold for enterprises to adopt open source models, promote wider localization and customized deployment, and reduce dependence on closed source APIs. See how the new version of Nemotron's open-source weight model performs in the future. If the performance can indeed catch up, it will actually further benefit cloud vendors, FDE models, and software and Saas that master customer workflows and data.