Advanced Micro Devices · Capability

AMD AI GPU Computing

Unified workflow capability for AI and HPC workloads on AMD Instinct GPUs — provision instances, deploy LLMs, monitor performance, and manage cloud credits. Designed for AI researchers, ML engineers, and HPC developers.

Run with Naftiko AMDAICloud ComputingGPUHPCMachine Learning

What You Can Do

GET
List instances — List GPU instances.
/v1/instances
POST
Create instance — Create a GPU instance.
/v1/instances
GET
List models — List deployed AI models.
/v1/models
POST
Deploy model — Deploy an AI model.
/v1/models

MCP Tools

list-gpu-instances

List all AMD Instinct GPU compute instances in the developer cloud.

read-only
create-gpu-instance

Launch an AMD Instinct GPU instance (MI300X, MI250, MI210) for AI training or HPC.

get-gpu-instance-status

Check the current status and configuration of an AMD GPU compute instance.

read-only
monitor-gpu-performance

Get real-time GPU utilization, memory usage, temperature, and power draw for an AMD instance.

read-only
deploy-llm

Deploy a large language model on AMD Instinct GPUs using vLLM for high-throughput inference.

list-deployed-models

List AI models deployed and serving on AMD Instinct GPUs.

read-only
check-credit-balance

Check available AMD Developer Cloud GPU credit balance and usage history.

read-only
terminate-instance

Terminate an AMD GPU compute instance to stop billing and release resources.

idempotent

APIs Used

amd-cloud