Confidential VMs with H100 GPUs Interest Form on Google Cloud

Request edit access

JavaScript isn't enabled in your browser, so this file can't be opened. Enable and reload.

This is the form to fill out if you are interested in Confidential GPUs, specifically Confidential VMs on accelerator optimized A3 machine series with NVIDIA H100 GPUs on Google Cloud. Note that the Confidential VMs with H100 GPUs Private Preview will be a limited release and interest requests will be evaluated on a case by case basis.

Google’s Privacy Policy: https://policies.google.com/privacy

Data provided in the form will ONLY be used by Google Cloud teams to opt users into learning more about Confidential VMs with H100 GPUs and the upcoming Private Preview of Confidential VMs with H100 GPUs. It will not be shared outside of Google.

If you have any questions, please email confidential-compute@google.com.

What's your email? *

What company or organization are you associated with? *

What GPUs do you use today on prem or in the public clouds? Check all that apply. https://cloud.google.com/compute/docs/gpus
*

A100

H100

I don't use any GPUs today anywhere.

Other:

Required

What are your use cases for your GPUs today? *

What are you planning to use Confidential GPU for (e.g. use cases)?
*

What are the concerns you are trying to mitigate or address with Confidential Computing and/or Confidential GPUs? Check all that apply. *

Increase data or model or workload protection, privacy, and isolation (increase security)

Decrease trust in cloud service provider (don't trust the cloud service provider)

Comply with data privacy laws, regulations, and auditors (compliance)

Perform untrusted multi-party computation

Other:

Required

What are your Confidential GPU requirements? This can include capacity requirements (# of GPUs), # of GPUs per VM (single vs multi-GPU), # of nodes (single vs multi-node), RAM, networking speed and bandwidth, workload size, AI/ML model size (# of parameters), remote attestation, performance, and latency.
*

What locations do you need Confidential GPUs? Please select all that apply. If you want specific regions and/zones, please also select "Other" and specify. *

APAC (e.g. asia-southeast1)

North America (e.g. us-central1, us-east4, us-east5)

South America

Europe (e.g europe-west4)

Middle East

Other:

Required

How do you want to consume Confidential GPUs on Google Cloud? Check all that apply. (e.g. GCE, GKE, Vertex, DWS)
*

Google Compute Engine (GCE): Just attached to VMs (e.g. A2 or A3 VM)

Google Kubernetes Engine: Standard (GKE Standard): Containers

Google Kubernetes Engine: Autopilot (GKE Autopilot): Managed Containers

Vertex AI: Google Managed AI/ML Service

Dynamic Workload Scheduling (DWS): Reservation system for GPUs

Other:

Required

OPTIONAL: Do you have any specific benchmarks or example workloads you'd like to see Confidential GPUs be tested with? (e.g. workload size, model size, # of parameters, serving vs training)

Submit

Clear form

Never submit passwords through Google Forms.

This form was created inside of Google.com. Privacy & Terms

Forms