In SURF Research Cloud, flavours are a predefined set of quantities that describe the dimensions of a workspace.
This is a list of the flavours that Research Cloud offers, depending on the application and the used cloud provider.
HPC Cloud
These flavours are for workspaces that are hosted in SURF's own HPC Cloud.
The creator of an application in the Research Cloud catalog can choose to offer a subset of these flavours for the application.
HPC Cloud flavours
| name | number of CPUs | number of GPUs | RAM [GB] |
|---|---|---|---|
| 1 Core - 8 GB RAM | 1 | 0 | 8 |
| 2 Core - 16 GB RAM | 2 | 0 | 16 |
| 4 Core - 32 GB RAM | 4 | 0 | 32 |
| 8 Core - 32 GB RAM | 8 | 0 | 32 |
| 16 Core - 64 GB RAM | 16 | 0 | 64 |
| 30 Core - 480 GB RAM | 30 | 0 | 480 |
HPC Cloud GPU flavours
| name | number of CPUs | CPU RAM [GB] | number of GPUs | GPU type | GPU RAM [GB] |
|---|---|---|---|---|---|
| RTX2080 - 1 GPU | 5 | 40 | 1 | NVIDIA RTX2080 Ti | 11 |
| RTX2080 - 2 GPU | 10 | 80 | 2 | NVIDIA RTX2080 Ti | 22 |
| RTX2080 - 4 GPU | 20 | 180 | 4 | NVIDIA RTX2080 Ti | 44 |
| A10 - 1 GPU | 11 | 90 | 1 | NVIDIA A10 | 24 |
| A10 - 2 GPU | 22 | 180 | 2 | NVIDIA A10 | 48 |
| A10 - 4 GPU | 44 | 350 | 4 | NVIDIA A10 | 96 |
Amazon Web Services (AWS)
These flavours are available for for workspaces on AWS. We also give the native AWS names for reference. You may also want to refer to the Amazon EC2 Instance Types.
AWS flavours
| name | number of CPUs | number of GPUs | RAM [GB] | AWS-name |
|---|---|---|---|---|
| 2 Core - 4 GB | 2 | 0 | 4 | t3.medium |
| 2 Core - 8 GB | 2 | 0 | 8 | t3.large |
| 2 Core - 16 GB | 2 | 0 | 16 | r5.large |
| 4 Core - 16 GB | 4 | 0 | 16 | t3.xlarge |
| 4 Core - 32 GB | 4 | 0 | 32 | r5.xlarge |
| 8 Core - 32 GB | 8 | 0 | 32 | t3.2xlarge |
| 8 Core - 64 GB | 8 | 0 | 64 | r5.2xlarge |
| 16 Core - 64 GB | 16 | 0 | 64 | m6a.4xlarge |
| 32 Core - 256 GB | 32 | 0 | 256 | r5d.8xlarge |
| 48 Core - 384 GB | 48 | 0 | 384 | r5.12xlarge |
| 96 Core - 768 GB | 96 | 0 | 768 | r5d.24xlarge |
AWS GPU flavours
| name | number of CPUs | number of GPUs | RAM [GB] | AWS-name |
|---|---|---|---|---|
| GPU 4 Core - 61 GB - 1x K80 | 4 | 1 | 193 | p2.xlarge |
| GPU 8 Core - 61 GB - 1x V100 | 8 | 1 | 61 | p3.2xlarge |
| GPU 32 Core - 244 GB - 4x V100 | 32 | 4 | 244 | p3.8xlarge |
| GPU 64 Core - 488 GB - 8x V100 | 64 | 8 | 488 | p3.16xlarge |
| GPU 16 Core - 64 GB - 1x A10 | 16 | 1 | 64 | g5.4xlarge |
| GPU 48 Core - 192 GB - 4x A10 | 48 | 4 | 192 | g5.12xlarge |
Azure
These are the flavours that are available on Microsoft Azure. We also give the native Azure names for reference. You may also want to refer to Azure's Virtual Machine series reference
Azure flavours
| name | number of CPUs | number of GPUs | RAM [GB] | Azure-name |
|---|---|---|---|---|
| 1 Core - 4 GB | 1 | 0 | 4 | Standard_DS1_v2 |
| 2 Core - 8 GB | 2 | 0 | 8 | Standard_D2as_v5 |
| 4 Core - 16 GB | 4 | 0 | 16 | Standard_D4as_v5 |
| 8 Core - 32 GB | 8 | 0 | 32 | Standard_D8as_v5 |
| 16 Core - 64 GB | 16 | 0 | 64 | Standard_D16as_v5 |
| 16 Core - 250GB | 16 | 0 | 250 | Standard_E32-16ds_v5 |
| 32 Core - 512 GB | 32 | 0 | 512 | Standard_E64-32ds_v5 |
Azure GPU flavours
| name | number of CPUs | number of GPUs | RAM [GB] | Azure-name |
|---|---|---|---|---|
| GPU 6 Core- 112 GB - 1x V100 | 6 | 1 | 112 | Standard_NC6s_v3 |
| GPU 12 Core - 110GB - 1/3x A10 | 12 | 1/3 | 110 | Standard_NV12ads_A10_v5 |
| GPU 36 Core - 440 GB - 1x A10 | 36 | 1 | 440 | Standard_NV36ads_A10_v5 |