The NDv2-series virtual machine is a new addition to the GPU family designed for the needs of the most demanding GPU-accelerated AI, machine learning, simulation, and HPC workloads.
NDv2 is powered by 8 NVIDIA Tesla V100 NVLINK-connected GPUs, each with 32 GB of GPU memory. Each NDv2 VM also has 40 non-HyperThreaded Intel Xeon Platinum 8168 (Skylake) cores and 672 GiB of system memory.
NDv2 instances provide excellent performance for HPC and AI workloads utilizing CUDA GPU-optimized computation kernels, and the many AI, ML, and analytics tools that support GPU acceleration 'out-of-box,' such as TensorFlow, Pytorch, Caffe, RAPIDS, and other frameworks.
Critically, the NDv2 is built for both computationally intense scale-up (harnessing 8 GPUs per VM) and scale-out (harnessing multiple VMs working together) workloads. The NDv2 series now supports 100-Gigabit InfiniBand EDR backend networking, similar to that available on the HB series of HPC VM, to allow high-performance clustering for parallel scenarios including distributed training for AI and ML. This backend network supports all major InfiniBand protocols, including those employed by NVIDIA’s NCCL2 libraries, allowing for seamless clustering of GPUs.
Host specifications
| Part |
Quantity
Count Units |
Specs
SKU ID, Performance Units, etc. |
| Processor |
40 vCPUs |
Intel Xeon Platinum 8168 (Skylake) [x86-64] |
| Memory |
672 GiB |
|
| Local Storage |
1 Disk |
2,948 GiB |
| Remote Storage |
32 Disks |
80,000 IOPS
800 MBps |
| Network |
8 NICs |
24,000 Mbps |
| Accelerators |
8 |
|
For features supported by this series, see the Feature support section.
Sizes in series
vCPUs (Qty.) and Memory for each size
| Size Name |
vCPUs (Qty.) |
Memory (GiB) |
| Standard_ND40rs_v2 |
40 |
672 |
VM Basics resources
Local (temp) storage info for each size
| Size Name |
Temp Storage Disks (Qty.) |
Temp Disk Size (GiB) |
| Standard_ND40rs_v2 |
1 |
2,948 |
Storage resources
Table definitions
- Temp disk performance depends on many factors including block size, workload patterns of read/writes, queue depth (QD), and others. Temp disk performance specifications should be viewed as best case performance numbers, assuming 4k block sizes and QD=256 for IOPS, and 256k block sizes with QD=64 for throughput. Additionally, temp disk performance often differs between read and write operations. During steady state operations, write performance is expected to be lower than read performance.
- Storage capacity is shown in units of GiB or 1024^3 bytes. When you compare disks measured in GB (1000^3 bytes) to disks measured in GiB (1024^3) remember that capacity numbers given in GiB may appear smaller. For example, 1023 GiB = 1098.4 GB.
- Disk throughput is measured in input/output operations per second (IOPS) and MBps where MBps = 10^6 bytes/sec.
- To learn how to get the best storage performance for your VMs, see Virtual machine and disk performance.
Remote (uncached) storage info for each size
| Size Name |
Max Remote Storage Disks (Qty.) |
Uncached Disk IOPS |
Uncached Disk Speed (MBps) |
| Standard_ND40rs_v2 |
32 |
80,000 |
800 |
Storage resources
Table definitions
- Some sizes support bursting to temporarily increase disk performance. Burst speeds can be maintained for up to 30 minutes at a time.
- Special Storage refers to either Ultra Disk or Premium SSD v2 storage.
- Storage capacity is shown in units of GiB or 1024^3 bytes. When you compare disks measured in GB (1000^3 bytes) to disks measured in GiB (1024^3) remember that capacity numbers given in GiB may appear smaller. For example, 1023 GiB = 1098.4 GB.
- Disk throughput is measured in input/output operations per second (IOPS) and MBps where MBps = 10^6 bytes/sec.
- Data disks can operate in cached or uncached modes. For cached data disk operation, the host cache mode is set to ReadOnly or ReadWrite. For uncached data disk operation, the host cache mode is set to None.
- To learn how to get the best storage performance for your VMs, see Virtual machine and disk performance.
Network interface info for each size
| Size Name |
Max NICs (Qty.) |
Max Bandwidth (Mbps) |
| Standard_ND40rs_v2 |
8 |
24,000 |
Networking resources
Table definitions
- Expected network bandwidth is the maximum aggregated bandwidth allocated per VM type across all NICs, for all destinations. For more information, see Virtual machine network bandwidth
- Upper limits aren't guaranteed. Limits offer guidance for selecting the right VM type for the intended application. Actual network performance will depend on several factors including network congestion, application loads, and network settings. For information on optimizing network throughput, see Optimize network throughput for Azure virtual machines.
- To achieve the expected network performance on Linux or Windows, you may need to select a specific version or optimize your VM. For more information, see Bandwidth/Throughput testing (NTTTCP).
Accelerator (GPUs, FPGAs, etc.) info for each size
| Size Name |
Accelerators (Qty.) |
Accelerator-Memory (GB) |
| Standard_ND40rs_v2 |
|
256 |
Feature support
List of all available sizes: Sizes
Pricing Calculator: Pricing Calculator
Information on Disk Types: Disk Types
Next steps
Take advantage of the latest performance and features available for your workloads by changing the size of a virtual machine.
Utilize Microsoft's in-house designed ARM processors with Azure Cobalt VMs.
Learn how to Monitor Azure virtual machines.