NVIDIA teams up with Google Cloud to help companies use Google Gemini AI models on their home turf. They can run these models locally using NVIDIA Blackwell HGX and DGX machines with special data protection tech. Companies need this setup when laws say they must keep sensitive data inside their buildings. The NVIDIA Blackwell system works with Google Distributed Cloud to lock down access to private records like medical charts and money transfers. Their security system keeps both the data and the AI code safe from prying eyes.
This team effort makes sure customer information stays private when talking to AI models. Companies can fine-tune these models without worrying about data leaks. Both sides of the exchange gain protection - the company data and the model code itself. This double shield lets businesses try new ideas with Gemini models but keep everything hush-hush. Sachin Gupta from Google Cloud says this helps customers create new things safely without losing speed or easy use.
Agentic AI changes how companies use technology because it solves problems better than older systems. Regular AI just learns stuff and spits out answers based on what it knows. But agentic AI thinks things through, changes plans, and makes choices when situations change. Take IT support as an example - normal AI just finds help guides for you. Agentic AI actually figures out what broke, fixes itself, and calls for backup if things look really bad.
The same goes for banking—regular AI might flag weird money moves based on patterns it sees. Agentic AI does more by checking out strange stuff and taking action right away. It might block fishy transactions before they happen or update the alarm system rules as things change. Many people already use these smart models that handle text, pictures, code, and other information types all at once to tackle hard problems.
Google Cloud breaks new ground as one of the first providers to offer tough security for agentic AI across all setups. The NVIDIA HGX B200 platform with Blackwell chips powers this whole operation with amazing performance that uses less energy. Google Cloud also announced a new tool called GKE Inference Gateway that makes running AI tasks easier and safer. This gateway talks nicely with other NVIDIA tools for better performance at a lower cost. Google Cloud plans to add even more features soon through something called NVIDIA Dynamo to watch over and manage these complex AI systems.
This team effort makes sure customer information stays private when talking to AI models. Companies can fine-tune these models without worrying about data leaks. Both sides of the exchange gain protection - the company data and the model code itself. This double shield lets businesses try new ideas with Gemini models but keep everything hush-hush. Sachin Gupta from Google Cloud says this helps customers create new things safely without losing speed or easy use.
Agentic AI changes how companies use technology because it solves problems better than older systems. Regular AI just learns stuff and spits out answers based on what it knows. But agentic AI thinks things through, changes plans, and makes choices when situations change. Take IT support as an example - normal AI just finds help guides for you. Agentic AI actually figures out what broke, fixes itself, and calls for backup if things look really bad.
The same goes for banking—regular AI might flag weird money moves based on patterns it sees. Agentic AI does more by checking out strange stuff and taking action right away. It might block fishy transactions before they happen or update the alarm system rules as things change. Many people already use these smart models that handle text, pictures, code, and other information types all at once to tackle hard problems.
Google Cloud breaks new ground as one of the first providers to offer tough security for agentic AI across all setups. The NVIDIA HGX B200 platform with Blackwell chips powers this whole operation with amazing performance that uses less energy. Google Cloud also announced a new tool called GKE Inference Gateway that makes running AI tasks easier and safer. This gateway talks nicely with other NVIDIA tools for better performance at a lower cost. Google Cloud plans to add even more features soon through something called NVIDIA Dynamo to watch over and manage these complex AI systems.
