GPU Fleet & Specs
We provide a mix of latest‑gen GPUs tuned for training and inference. Below are sample instance types — final availability depends on region and capacity.
h100.large
GPU: 1x NVIDIA H100
Memory: 80 GB HBM3
CPU: 16 vCPU
Storage: 2 TB NVMe
h100.xlarge
GPU: 8x NVIDIA H100
Memory: 640 GB HBM3
CPU: 128 vCPU
Storage: 16 TB NVMe
a100.large
GPU: 1x NVIDIA A100
Memory: 40/80 GB HBM2e
CPU: 12 vCPU
Storage: 1 TB NVMe