[[["容易理解","easyToUnderstand","thumb-up"],["確實解決了我的問題","solvedMyProblem","thumb-up"],["其他","otherUp","thumb-up"]],[["難以理解","hardToUnderstand","thumb-down"],["資訊或程式碼範例有誤","incorrectInformationOrSampleCode","thumb-down"],["缺少我需要的資訊/範例","missingTheInformationSamplesINeed","thumb-down"],["翻譯問題","translationIssue","thumb-down"],["其他","otherDown","thumb-down"]],["上次更新時間:2025-09-04 (世界標準時間)。"],[],[],null,["# Provision GPUs and enable Vertex AI pre-trained APIs\n\nEach Vertex AI service requires different computing resources,\nincluding graphics processing units (GPUs), as described in\n[Capacity planning and computing requirements](/distributed-cloud/hosted/docs/latest/appliance/application/ao-user/vertex-ai-overview#capacity).\nThis page describes how to provision GPU resources to run artificial\nintelligence (AI) and machine learning (ML) workloads in a GPU environment. This\npage also walks you through enabling the\nVertex AI pre-trained APIs on Google Distributed Cloud (GDC) air-gapped appliance so\nyou can start implementing Vertex AI capabilities.\n\nMost tasks to configure GPU resources and enable or deactivate\nVertex AI pre-trained APIs require administrator access. If you\nlack the necessary permissions, ask your administrator to enable GPUs and the\nVertex AI pre-trained APIs on your behalf.\n\nVertex AI on GDC air-gapped appliance includes three APIs, one for\neach of its pre-trained models. To learn more about these pre-trained models,\nsee the following documentation:\n\n- **Optical Character Recognition (OCR)** : [Learn about character recognition features](/distributed-cloud/hosted/docs/latest/appliance/application/ao-user/vertex-ai-ocr).\n- **Speech-to-Text** : [Learn about speech recognition features](/distributed-cloud/hosted/docs/latest/appliance/application/ao-user/vertex-ai-stt).\n- **Vertex AI Translation** : [Learn about translation features](/distributed-cloud/hosted/docs/latest/appliance/application/ao-user/vertex-ai-translation).\n\nUse the GDC console to enable, deactivate, and view the endpoints of\nthe Vertex AI pre-trained APIs.\n\nBefore you begin\n----------------\n\nTo get the permissions that you need to enable pre-trained APIs, ask your\nOrganization IAM Admin or Project IAM Admin to grant you the AI Platform Admin\n(`ai-platform-admin`) role in the project namespace.\n\nFor information about this role, see [Prepare IAM permissions](/distributed-cloud/hosted/docs/latest/appliance/application/ao-user/vertex-ai-ao-permissions).\nTo learn how to grant permissions to a subject, see [Grant and revoke access](/distributed-cloud/hosted/docs/latest/appliance/platform/pa-user/iam/set-up-role-bindings).\n\nFollow these steps to provision GPUs before enabling the pre-trained APIs:\n\n1. [Set up a project to use Vertex AI](/distributed-cloud/hosted/docs/latest/appliance/application/ao-user/vertex-ai-set-up-project).\n2. Ensure that your project has the adequate ingress communication configured.\n3. [Manage GPU workloads](/distributed-cloud/hosted/docs/latest/appliance/application/ao-user/containers/deploy-gpu-container-workloads).\n4. [Sign in to the GDC console](/distributed-cloud/hosted/docs/latest/appliance/platform/pa-user/iam/sign-in). If you can't sign in, see [Recover device if identity provider is unavailable](/distributed-cloud/hosted/docs/latest/appliance/admin/recover-idp).\n\nEnable pre-trained APIs\n-----------------------\n\nYou can enable the OCR, Speech-to-Text, and\nVertex AI Translation pre-trained APIs using the GDC console.\n| **Caution:** GDC air-gapped appliance includes a single NVIDIA A100 80GB GPU. This limits the device to running only *one* GPU-dependent Vertex AI pre-trained API at a time. If you attempt to enable both Vertex AI Translation and OCR, the second API will fail to enable with an error message indicating insufficient GPU resources. You can, however, run Speech-to-Text alongside either Vertex AI Translation or OCR, as Speech-to-Text only requires CPU resources.\n\nAfter [meeting the prerequisites](#before-you-enable-apis), follow these steps\nto enable the pre-trained APIs:\n\n1. [Sign in to the GDC console](/distributed-cloud/hosted/docs/latest/appliance/platform/pa-user/iam/sign-in).\n2. In the navigation menu, click **Vertex AI** \\\u003e **Pre-trained APIs**.\n3. On the **Pre-trained APIs** page, click **Enable** on a specific service to\n enable that API.\n\n | **Note:** The GDC console doesn't display the buttons to enable the pre-trained APIs if you don't have the correct permissions.\n4. In the confirmation dialog, click **Enable**. A progress message displays.\n\nThe enablement duration varies. It might take between 15 and 45 minutes to\nfinish.\n\nIf you want to view the status of the pre-trained APIs,\n[view the service status and endpoints](/distributed-cloud/hosted/docs/latest/appliance/application/ao-user/vertex-ai-api-status).\n| **Note:** You can't enable an API while you wait for another API to finish deactivating.\n\nDeactivate pre-trained APIs\n---------------------------\n\nYou can deactivate the OCR, Speech-to-Text, and\nVertex AI Translation pre-trained APIs using the GDC console.\n\nAfter [meeting the prerequisites](#before-you-enable-apis), follow these steps\nto deactivate the pre-trained APIs:\n\n1. [Sign in to the GDC console](/distributed-cloud/hosted/docs/latest/appliance/platform/pa-user/iam/sign-in).\n2. In the navigation menu, click **Vertex AI** \\\u003e **Pre-trained APIs**.\n3. On the **Pre-trained APIs** page, click **Disable** on a specific service to\n deactivate that API.\n\n | **Note:** The GDC console doesn't display the buttons to deactivate the pre-trained APIs if you don't have the correct permissions.\n4. In the confirmation dialog, enter `disable` in the text field to confirm\n that you want to take that action. Then, click **Disable**. A progress\n message displays.\n\nIf you want to view the status of the pre-trained APIs,\n[view the service status and endpoints](/distributed-cloud/hosted/docs/latest/appliance/application/ao-user/vertex-ai-api-status).\n| **Note:** You can't deactivate an API while you wait for another API to finish enabling."]]