gpu
Cloud GPUs
Compute & WebGCP
General
General information about the service.
Description
Google Cloud GPUs provide high-performance GPU accelerators for machine learning training and inference workloads.
Service ID
compute.googleapis.comResource Pattern
projects/{project}/zones/{zone}/acceleratorTypes/{type}References
Chat
Powered by AITalk to this service to learn more about it.
Suggested questions:
Naming
The conventions, rules, and restrictions for naming this service.
Naming Convention
gpu-{workload}-{environment}Length
1-63
Valid Characters
Lowercase letters, numbers, hyphens
Scope
project
Code
Deploy your infrastructure as code using your preferred tooling.
Official Documentation
terraform
1resource "google_compute_instance" "gpu" {2 name = "gpu-${var.environment}"3 machine_type = "n1-standard-4"4 zone = var.zone56 guest_accelerator {7 type = "nvidia-tesla-t4"8 count = 19 }1011 scheduling {12 on_host_maintenance = "TERMINATE"13 }1415 boot_disk {16 initialize_params {17 image = "deeplearning-platform-release/tf-latest-gpu"18 }19 }2021 network_interface {22 network = "default"23 }24}
Utilities
Utilities to support app deployment or configuration.