confidential-computing scope per CVM — Hopper passes all eight HGX GPUs with unencrypted NVLink; Blackwell supports up to eight GPUs with encrypted peer-to-peer NVLink

Name: confidential-computing scope per CVM — Hopper passes all eight HGX GPUs with unencrypted NVLink; Blackwell supports up to eight GPUs with encrypted peer-to-peer NVLink
License: https://aidatacenterguide.com/terms#intellectual-property

Hopper: 8 GPUs/CVM; Blackwell: up to 8 GPUs/CVMobserved

Value kind	observed — Reported measurements, counts, and specifications keep the precision and scope stated by their source; an exact specification is not treated as a range.
As of	2025-04
Source	NVIDIA Technical Blog, Announcing NVIDIA Secure AI General Availability (2025-04-23; Hopper PPCIe: eight GPUs and four switches per CVM, NVLink encryption removed); NVIDIA WP-12554 (Blackwell multi-GPU with encrypted NVLink)
Review	checking…review by 2027-03-31 · standard cadence
Recorded changes	last 2026-10-16 · 3 revisions tracked
Claim id	hopper-class-confidential-computing-scope-multi

Where the guide uses it

11.5 GPU Confidential Computing & Trusted Execution

← Full numbers register — every date-stamped figure in the guide, with revision history.