# Machine types

You can deploy the following machine types inside your lab that fit your scientific explorations.

Pending activation

Machine types marked with a star (*) in the tables below are available in location 2.

# Compute optimized machine types

These machines are optimized for computation with 2 GB, 4 GB, 8 GB or 16 GB of memory per vCPU respectively.

# B-series

2 GB of memory per vCPU.

Type vCPU Memory (GB)
default.b1 2 4
default.b2 4 8
default.b3 8 16
default.b4 16 32
default.b5 32 64
default.b6 48 96
default.b7* 64 128
default.b8* 80 160

# C-series

4 GB of memory per vCPU.

Type vCPU Memory (GB)
default.c1 2 8
default.c2 4 16
default.c3 8 32
default.c4 16 64
default.c5 32 128
default.c6 48 192
default.c7* 64 256
default.c8* 80 320

# D-series

8 GB of memory per vCPU.

Type vCPU Memory (GB)
default.d1 2 16
default.d2 4 32
default.d3 8 64
default.d4 16 128
default.d5 24 192
default.d6* 48 384
default.d7* 64 512
default.d8* 80 640

# E-series

16 GB of memory per vCPU.

Type vCPU Memory (GB)
default.e1 2 32
default.e2 4 64
default.e3 8 128
default.e4 16 256
default.e5* 24 384
default.e6* 32 512
default.e7* 48 768

# GPU accelerator types

All the above machine types may be deployed with one or two GPU accelerators:

Type Model Architecture Memory Interface
nvidia.p100 NVIDIA P100 Pascal 16GB HBM2 PCIe Gen3
nvidia.v100 NVIDIA V100 Volta 32GB HBM2 PCIe Gen3
nvidia.v100s NVIDIA V100S Volta 32GB HBM2 PCIe Gen3
nvidia.a30* NVIDIA A30 Ampere 24GB HBM2 PCIe Gen3
nvidia.a40* NVIDIA A40 Ampere 48GB HBM2 PCIe Gen3
nvidia.a100* NVIDIA A100 Ampere 40GB HBM2 PCIe Gen3
nvidia.a100* NVIDIA A100 Ampere 80GB HBM2 PCIe Gen3

Note that availability vary between locations. See the GPU on HUNT Cloud documentation for more information.

# Custom machine types

Not finding the type of your choice? Please contact us so we can tailor machine types for your need (within our limitations of available memory and vCPUs).

Last Updated: 12/9/2024