Apple may be preparing one of its most unusual Siri decisions yet: leaning on Google Cloud infrastructure powered by NVIDIA GPUs to speed up a major overhaul of its voice assistant. The move would pair outside computing power with Apple’s own AI systems, suggesting a more hybrid approach than the company has traditionally favored.
The report comes at a time when generative AI workloads are becoming harder to handle with a purely internal setup. Apple has long been known for keeping tight control over core technologies, but the scale of modern AI appears to be pushing it toward external cloud resources to help bring long-delayed Siri features to market sooner.
According to The Information, the updated Siri is targeted for release in September. The same report says the service could arrive alongside iOS 27, although Apple has not publicly confirmed that plan.
Why Google Cloud and NVIDIA matter
The proposed setup reportedly centers on Google’s Gemini model running on NVIDIA Blackwell GPUs. That combination is considered capable of handling large-scale AI inference and reasoning, which are essential for a more advanced voice assistant.
NVIDIA’s role would go beyond being a simple chip supplier in this scenario. The Blackwell B200 GPU is described as well suited for complex AI tasks, including layered request processing and responses that require deeper reasoning.
That level of computing matters because Siri is expected to do far more than answer basic voice commands. The next version is expected to better understand context, follow longer conversations, and handle multi-step requests with more flexibility.
A shift from Apple’s usual approach
If Apple does use this arrangement, it would mark a notable change in strategy. The company has traditionally emphasized on-device processing and a privacy-first design, but generative AI has changed the technical requirements in a significant way.
Many modern AI features depend on large-scale cloud infrastructure to deliver fast responses and more advanced capabilities. In that context, using NVIDIA-based cloud capacity could help Apple move faster without waiting for every internal system to be fully ready.
Apple is still expected to combine its own privacy technology with Google Cloud, rather than handing Siri over entirely to external servers. Basic tasks would remain processed on the device, while more demanding queries would be routed to Google Cloud running Gemini AI and NVIDIA Blackwell hardware.
What Siri is expected to do differently
The broader goal appears to be turning Siri from a basic voice assistant into a more capable AI companion. That means users may eventually be able to carry on longer exchanges without repeating details at every step.
Context awareness is also expected to improve, making Siri better at understanding what a user means across multiple requests. Multi-step task handling is another key part of the upgrade, allowing the assistant to complete a chain of actions rather than only a single simple instruction.
Those changes require more than software refinements. They depend on fast response times and substantial processing capacity, which is why outside infrastructure may be the most practical way for Apple to keep the project moving.
Broader implications for Apple, Google, and NVIDIA
For NVIDIA, involvement in a project like this would reinforce its growing influence in AI infrastructure. Many major AI models already rely on NVIDIA hardware, and Siri could become another high-profile example.
For Google, the reported arrangement shows that its cloud and AI stack can support services beyond its own products, even when a major rival is involved. For Apple, the main objective appears to be shipping a more modern Siri without introducing another delay.
The setup remains unconfirmed, and Apple has not officially announced the use of Google Cloud, Gemini, or NVIDIA Blackwell for the new Siri. Still, the report makes clear that the technical demands behind the next-generation assistant are substantial, and Apple may be willing to look outward to meet them.