Skip to main content
Compute offers the following platforms for virtual machines (VMs) (private regions are marked with *): Each platform offers different presets for creating VMs. The presets vary by the number of GPUs (for platforms with GPUs), vCPUs and RAM size. The exact amount of these resources is indicated in the preset name. Thus, 1gpu-16vcpu-200gb preset stands for a VM with 1 GPU, 16 vCPUs and 200 RAM. Availability of platforms depends on a region and project. You can get a list of platforms and presets supported in a given project. For VM platforms with GPUs, you can also use the capacity advisor to get information about the availability of computing resources based on your quotas and the current physical capacity.

​
GPU types

Compute provides the following GPU types:
  • NVIDIA B300 (NVIDIA Blackwell Ultra architecture) in the SXM factor. It offers 270 GB of HBM3e RAM with up to 7.7 TB/s of memory bandwidth. Each set of eight physical GPUs uses eight NVIDIA ConnectX-8 adapter cards, each providing 800 Gbps InfiniBand™ interconnectivity, plus an NVIDIA BlueField-3 Ethernet adapter that provides a 400 Gbps connection to the underlying network. NVIDIA B300 SXM is available on the NVIDIA® B300 NVLink with Intel Granite Rapids platform.
  • NVIDIA B200 (NVIDIA Blackwell architecture) in the SXM factor. It offers 180 GB of HBM3e RAM with up to 7.7 TB/s of memory bandwidth. Each set of eight physical GPUs uses eight NVIDIA ConnectX-7 adapter cards, each providing 400 Gbps InfiniBand interconnectivity, plus an NVIDIA BlueField-3 Ethernet adapter that provides a 200 Gbps connection to the underlying network. NVIDIA B200 SXM is available on the NVIDIA® B200 NVLink with Intel Emerald Rapids platform (gpu-b200-sxm and gpu-b200-sxm-a).
  • NVIDIA H200 (NVIDIA Hopper architecture) in the SXM factor. It offers 141 GB of HBM3e RAM with up to 4.8 TB/s of memory bandwidth. Each set of eight physical GPUs uses eight NVIDIA ConnectX-7 adapter cards, each providing 400 Gbps InfiniBand interconnectivity, plus an NVIDIA BlueField-3 Ethernet adapter that provides a 200 Gbps connection to the underlying network. NVIDIA H200 SXM is available on the NVIDIA® H200 NVLink with Intel Sapphire Rapids platform.
  • NVIDIA H100 (NVIDIA Hopper architecture) in the SXM factor. It offers 80 GB of HBM3 RAM with up to 3.35 TB/s of memory bandwidth. Each set of eight physical GPUs uses eight NVIDIA ConnectX-7 adapter cards, each providing 400 Gbps InfiniBand interconnectivity, plus an NVIDIA ConnectX-6 Ethernet adapter that provides a 100 Gbps connection to the underlying network. NVIDIA H100 SXM is available on the NVIDIA® H100 NVLink with Intel Sapphire Rapids platform.
  • NVIDIA RTX PRO 6000 Server Edition (NVIDIA Blackwell architecture) in a PCIe Gen5 form factor. Each GPU offers 96 GB of GDDR7 memory with up to 1.6 TB/s of memory bandwidth. Networking is enabled via a 400 Gbps BlueField-3 adapter. NVIDIA RTX PRO 6000 Server Edition is available on the NVIDIA® RTX PRO™ 6000 with Intel Granite Rapids platform.
  • NVIDIA L40S (NVIDIA Ada Lovelace architecture) in the PCIe factor. It offers 48 GB of GDDR6 RAM with up to 864 GB/s of memory bandwidth. NVIDIA L40S is available on the NVIDIA® L40S PCIe with Intel Ice Lake and NVIDIA® L40S PCIe with AMD EPYC Genoa platforms.

​
Presets for GPU platforms

  • NVIDIA® B300 NVLink with Intel Granite Rapids (gpu-b300-sxm), available in uk-south1 and eu-west2:
  • NVIDIA® B200 NVLink with Intel Emerald Rapids (gpu-b200-sxm and gpu-b200-sxm-a), available in us-central1 and me-west1 respectively:
  • NVIDIA® H200 NVLink with Intel Sapphire Rapids (gpu-h200-sxm), available in eu-north1, eu-north2, eu-west1 and us-central1:
  • NVIDIA® H100 NVLink with Intel Sapphire Rapids (gpu-h100-sxm), available in eu-north1:
  • NVIDIA® RTX PRO™ 6000 with Intel Granite Rapids (gpu-rtx6000), available in us-central1:
  • NVIDIA® L40S PCIe with Intel Ice Lake (gpu-l40s-a), available in eu-north1:
  • NVIDIA® L40S PCIe with AMD EPYC Genoa (gpu-l40s-d), available in eu-north1:

​
Presets compatible with GPU clusters

If you are adding a VM to a GPU cluster, select from the following platforms and presets: Other presets and platforms are not compatible with GPU clusters.

​
Presets for non-GPU platforms

  • Non-GPU AMD EPYC Genoa (cpu-d3):
  • Non-GPU Intel Ice Lake (cpu-e2), available in eu-north1:

​
Compatibility with boot disk images

Nebius AI Cloud provides boot disk images for GPU and non-GPU VMs. The image that you choose for a VM must be compatible with the VM’s platform. For compatibility details, see Boot disk images for Compute virtual machines.

​
Compatibility with VM types

All platforms support regular VMs. All platforms with GPUs also support preemptible VMs. To get an up-to-date list of platforms, run the nebius compute platform list command. The platforms available for preemptible VMs are marked as allowed_for_preemptibles: true.