Compare GCP Instances
| Instance Type | a4x-highgpu-4g | t2a-standard-4 |
|---|---|---|
| Details | Optimized for accelerated high performance computing workloads. a – Accelerator optimized 4 – Generation highgpu – High amound of GPU memory - 40x RAM/GPU ratio 4g – Number of GPUs | Tau T-instances where each vCPU is equivalent to an entire core that gives cost-effective performance. Scale-out workloads, Web servers, Containerized microservices. t – General purpose with entire core 2 – Generation standard – 4x RAM/CPU ratio 4 – Number of CPUs |
| Instance Family | Accelerator optimized | General-purpose |
| Description | Accelerator Optimized: 4 NVIDIA GB200 GPU, 140 vCPUs, 884GB RAM | 4 vCPUs, 16 GB RAM |
| vCPUs | 140 | 4 |
| Memory | 884 GiB | 16 GiB |
| CPU Architecture | arm64 | arm64 |
| Processor | NVIDIA Grace | Ampere Altra |
| Coremark Score | 94096 | |
| Shared CPU | no | no |
| GPU | yes | no |
| GPUs | 4 | |
| GPU Name | nvidia-gb200 | |
| Sustained Discounts (SUDs) | no | no |
| 25%-50% Usage Sustained Discount | 0.8 | |
| 50%-75% Usage Sustained Discount | 0.6 | |
| 75%-100% Usage Sustained Discount | 0.4 | |
| Sole Tenancy | no | no |
| Nested Virtualization | no | no |
| Custom Shapes | no | no |
| Confidential Compute | no | no |
| Maximum Persistent Disks | 128 | 128 |
| Maximum Persistent Disks Size | 524288 GB | 263168 GB |
| Disk Interface Type | NVMe | NVMe |
| Local SSD | yes | no |
| Max Local SSD | 12000 GB | |
| Extreme Persistent Disks | no | no |
| Hyperdisk Extreme | yes | no |
| Hyperdisk Throughput | no | no |
| Network Interfaces | gVNIC, MRDMA | gVNIC |
| Avg Price Diff | Infinityx |
Regional Prices
Geography | Region | a4x-highgpu-4g | t2a-standard-4 |
|---|---|---|---|
| North America | Iowa (us-central1) | n/a | 0.1540 |
| Europe | Netherlands (europe-west4) | n/a | 0.1618 |
| APAC | Singapore (asia-southeast1) | n/a | 0.1899 |