Snowpark Container Services: Google Cloud Instance Families¶
Instance families are grouped by type: general compute, high memory, and GPU accelerated. GCP instance families are available in three types:
- General Compute (GEN): Best price-performance for general-purpose containerized workloads.
- High Memory (MEM): High memory-to-vCPU ratio for applications that require large amounts of RAM, such as CPU-based model serving, large-scale in-memory data processing, and vector index serving.
- GPU Accelerated (GPU): For machine learning training, inference, and AI workloads requiring GPU acceleration.
For pricing information, see the Snowflake Service Consumption Table.
Note
Region availability is subject to change. To retrieve current availability and instance family specifications programmatically, use SHOW COMPUTE POOL INSTANCE FAMILIES.
Current Generation Instance Families¶
General Compute Instance Families (Current Generation)¶
Current generation x86 instances offering the best price-performance for general-purpose workloads.
| Instance Family | vCPU | Memory (GiB) | Storage (GB) | Bandwidth limit (Gbps) | Node limit | Region Availability |
|---|---|---|---|---|---|---|
| CPU_X64_XS | 1 | 6 | 100 | 10.0 | 500 | Available everywhere |
| CPU_X64_S | 3 | 13 | 100 | 10.0 | 500 | Available everywhere |
| CPU_X64_M | 6 | 28 | 100 | 16.0 | 500 | Available everywhere |
| CPU_X64_SL | 14 | 58 | 100 | 32.0 | 500 | Available everywhere |
| CPU_X64_L | 28 | 116 | 100 | 32.0 | 500 | Available everywhere |
High Memory Instance Families (Current Generation)¶
Current generation x86 instances optimized for memory-intensive workloads.
| Instance Family | vCPU | Memory (GiB) | Storage (GB) | Bandwidth limit (Gbps) | Node limit | Region Availability |
|---|---|---|---|---|---|---|
| HIGHMEM_X64_S | 6 | 58 | 100 | 16.0 | 500 | Available everywhere |
| HIGHMEM_X64_M | 28 | 240 | 100 | 32.0 | 500 | Available everywhere |
| HIGHMEM_X64_SL | 92 | 654 | 100 | 67.0 | 500 | Not available in me-central2 |
GPU Accelerated Instance Families (Current Generation)¶
GCP GPU instance families feature two NVIDIA GPU architectures, each suited to different AI and ML workloads.
NVIDIA L4¶
Ada Lovelace GPU for efficient AI inference and media workloads.
| Instance Family | vCPU | Memory (GiB) | Storage (GB) | Bandwidth limit (Gbps) | GPU | GPU Memory per GPU (GB) | Node limit | Region Availability |
|---|---|---|---|---|---|---|---|---|
| GPU_GCP_NV_L4_1_24G | 6 | 28 | 100 | 16.0 | 1 NVIDIA L4 | 24 | 10 | Available everywhere |
| GPU_GCP_NV_L4_4_24G | 44 | 178 | 100 | 50.0 | 4 NVIDIA L4 | 24 | 10 | Available everywhere |
NVIDIA A100¶
High-throughput Ampere GPU for large-scale model training and large dataset processing.
| Instance Family | vCPU | Memory (GiB) | Storage (GB) | Bandwidth limit (Gbps) | GPU | GPU Memory per GPU (GB) | Node limit | Region Availability |
|---|---|---|---|---|---|---|---|---|
| GPU_A100_G1_12 | 10 | 77 | 100 | 10.0 | 1 NVIDIA A100 | 40 | On Request | Only available in us-central1 and europe-west4 |
| GPU_A100_G1_48 | 44 | 324 | 100 | 50.0 | 4 NVIDIA A100 | 160 | On Request | Only available in us-central1 and europe-west5 |
| GPU_GCP_NV_A100_8_40G | 92 | 654 | 100 | 100.0 | 8 NVIDIA A100 | 40 | On Request | Available only in us-central1 and europe-west4 |
Previous Generation Instance Families¶
There are no previous generation instance families on Google Cloud. All listed instance families are current generation.