General purpose instances featuring full passthrough on a affordable fraction of a NVIDIA A16 GPU.
The NVIDIA A16 is a data center GPU designed for high-density virtual desktop infrastructure (VDI), cloud gaming, and virtualized workstation workloads. Based on NVIDIA Ampere architecture, it combines four GPUs on a single PCIe card, delivering efficient multi-user performance and hardware-accelerated graphics. The A16 supports NVIDIA Virtual GPU (vGPU) software, enabling secure resource sharing across multiple virtual machines. With low power consumption and optimized scalability, it helps organizations deploy cost-effective, graphics-rich virtual environments while maximizing server utilization and operational efficiency.
| Name | NVMe Storage | Memory | vCPUs | Transfer | Price | Regions |
|---|---|---|---|---|---|---|
G1-gpu-a16-8m-2c-2g
| 50 GiB | 8 GB | 2 vCPUs | 1 TB | $0.0753 /hr $55 /mo | 9 |
G1-gpu-a16-16m-2c-4g
| 80 GiB | 16 GB | 2 vCPUs | 2 TB | $0.1507 /hr $110 /mo | 9 |
G1-gpu-a16-32m-3c-8g
| 170 GiB | 32 GB | 3 vCPUs | 3 TB | $0.3014 /hr $220 /mo | 9 |
G1-gpu-a16-64m-6c-16g
| 350 GiB | 64 GB | 6 vCPUs | 6 TB | $0.6027 /hr $440 /mo | 9 |
G1-gpu-a16-128m-12c-32g
| 700 GiB | 128 GB | 12 vCPUs | 10 TB | $1.2055 /hr $880 /mo | 8 |
G1-gpu-a16-256m-24c-64g
| 1200 GiB | 256 GB | 24 vCPUs | 12 TB | $2.3288 /hr $1 700 /mo | 5 |
G1-gpu-a16-496m-48c-128g
| 1500 GiB | 496 GB | 48 vCPUs | 15 TB | $4.6575 /hr $3 400 /mo | 2 |
G1-gpu-a16-960m-96c-256g
| 1700 GiB | 960 GB | 96 vCPUs | 25 TB | $9.3151 /hr $6 800 /mo | Sold out |