NVIDIA GPUs with Confidential Computing are now used for confidential inference in Apple’s Private Cloud Compute (PCC), as it expands beyond Apple’s data centers to Google Cloud. The move was unveiled during Apple’s annual WWDC gathering for developers.
The Announcement
At WWDC, Apple said NVIDIA GPUs will support server-side inference for Apple Foundation Models — custom-built by Apple and Google, leveraging the technologies behind the Gemini family of models. NVIDIA is collaborating with Apple and Google to power next-generation Apple Intelligence features using Blackwell GPUs with Confidential Computing, integrated into PCC’s hardware security architecture running on Google Cloud.
The expansion also extends PCC support to Apple Silicon, giving iOS, iPadOS and macOS a single hardware security architecture across platforms.
Confidential Computing at Scale
NVIDIA Confidential Computing provides a hardware-based security layer for accelerated AI workloads. The technology protects data while it is being processed by isolating workloads in trusted execution environments, and enables systems to cryptographically verify that the infrastructure has not been tampered with before any sensitive data is sent to the server.
For end users, the practical effect is blunt: no one — not even the system’s builders — can look at their data, chats or conversations.
How Privacy Is Enforced
Adoption at this scale reflects a broader shift in AI infrastructure — as AI experiences combine on-device and cloud processing, there is a growing need for high-performance, server-side inference with strong privacy guarantees. The key capabilities:
- Hardware-rooted trust, helping establish that systems run on genuine, untampered NVIDIA GPUs
- Encrypted communication paths, protecting data as it moves between components
- Remote attestation, letting software verify the security state of the platform before releasing sensitive data
- Accelerated AI inference and training, helping organizations run privacy-sensitive workloads without sacrificing GPU performance
What This Means
For NVIDIA, being named Apple’s confidential-inference partner validates confidential computing as a mainstream architecture rather than a niche enterprise feature. For users of Apple Intelligence, it means the heaviest requests get more capable, secure compute — processed in a way where the infrastructure providers themselves cannot inspect the workload.