(All-cores)
(Single-core)
Specifications
Server Metadata
Vendor ID | azure |
Server ID | Standard_ND40rs_v2 |
Name | ND40rs_v2 |
API Reference | Standard_ND40rs_v2 |
Description | NDSv2 family [RDMA capable] [Premium Storage capable] [Intel processor] 40 vCPUs |
Family | NDSv2 |
Hw Virt | - |
Status | active |
Observed At | 2026-11-06T09:11:39.593384 |
Availability
| REGION / ID | SPOT | ONDEMAND |
|---|---|---|
East US (US) / eastus | 4.0715 USD/h | 22.032 USD/h |
West US 2 (US) / westus2 | 4.0715 USD/h | 22.032 USD/h |
South Central US (US) / southcentralus | 4.8857 USD/h | 26.438 USD/h |
West Europe (NL) / westeurope | 5.0868 USD/h | 27.526 USD/h |
Southeast Asia (SG) / southeastasia | 5.6349 USD/h | 30.492 USD/h |
Processor
vCPUs | 40 |
Hypervisor | Microsoft Hyper-V |
CPU Allocation | Dedicated |
CPU Architecture | x86_64 |
System Resources and Accelerators
| MEMORY | |
|---|---|
Memory Amount | 672 GiB |
| ACCELERATORS | |
|---|---|
Accelerator Count | 8 |
Accelerator Memory Min | 30 GiB |
Accelerator Memory Total | 238 GiB |
Accelerator Type | GPU |
Accelerator Manufacturer | NVIDIA |
Accelerator Model | V100 |
Accelerators |
| STORAGE | |
|---|---|
Storage Size | 2970 GB |
Storage Type | ssd |
Storages |
|
| NETWORK | |
|---|---|
Inbound Traffic | 0 GB/month |
Outbound Traffic | 0 GB/month |
IPv4 | 0 |
Server Description
An accelerated virtual machine featuring eight dedicated GPUs, massive memory capacity, and low-latency RDMA networking for high-performance parallel computing.
Microsoft Azure ND40rs_v2 is an NDSv2-family virtual machine optimized for GPU-accelerated workloads and high-performance computing. Running on the Microsoft Hyper-V hypervisor, the server features 40 dedicated x86_64 Intel vCPUs and 672.0 GB of system memory, providing 16.8 GB of RAM per core. The hardware configuration is highlighted by eight NVIDIA V100 GPUs with a total of 238 GB of VRAM, alongside a 2970 GB local SSD. Designed for clustered deployments, the instance is RDMA-capable and supports Premium Storage. This high-density resource profile is qualitatively suited for demanding parallel processing tasks. Appropriate workloads include deep learning model training, large-scale artificial intelligence inference, high-performance computing simulations, and intensive data science operations requiring multi-GPU acceleration.
Economics
Alternatives
Servers of the Same Family
| INSTANCE | vCPUs | MEMORY | Accelerators |
|---|
Similar Servers
| INSTANCE | VENDOR | vCPUs | MEMORY | Accelerators |
|---|