Bare Metal GPU Servers
Optimized for AI Workloads
Run inference engines, generative AI, and continuously train small language models (SLMs)
LOW
COST
70% lower than
AWS 3-year reserved pricing*
HIGH
PERFORMANCE
Fully-integrated servers with up
to 16 NVlink’d GPUs with 512GB of GPU Memory
LOW
LATENCY
Five US data centers along both
coasts ensure quick response times
DGX-2 (16 v100 32GB)
Starting at $0.59 gpu/hr (~$6,899/mo)
• CPU Dual Intel Xeon Platinum 8168, 2.7 GHz, 24-cores
• System Memory 1.5TB
• Storage OS:
• Boot 2x 960GB NVME SSDs
• Internal Storage: 30TB; (8x 3.84TB) NVME SSDs
• Network TBD
OCP Big Basin (8 v100 16GB)
Starting at $0.29 gpu/hr (~$1,699/mo)
• System Memory 512GB
• Storage
• Boot: 1TB; (1X 1TB) SATA SSD
• Network: 25GB
OCP Big Basin (8 v100 16GB)
Starting at $0.29 gpu/hr (~$1,699/mo)
• System Memory 512GB
• Storage
• Boot: 1TB; (1X 1TB) SATA SSD
• Network: 25GB
OCP BIG BASIN
8 x NVIDIA V100 16GB
128GB GPU Memory
128GB GPU Memory
Starting at $0.29 gpu/hr ($1,699/mo)*
NVIDIA DGX-1
8 x NVIDIA V100 32GB
256GB GPU Memory
256GB GPU Memory
Starting at $0.49 gpu/hr ($2,860/mo)*
NVIDIA DGX-2
16 x NVIDIA V100 32GB
512GB GPU Memory
512GB GPU Memory
Starting at $0.59 gpu/hr ($6,899/mo)*
*Requires a 1-year commitment
OCP Big Basin
8 x NVIDIA V100 16GB
128GB GPU Memory
$0.29 gpu/hr ($1,699/mo)‡
NVIDIA DGX-1
8 x NVIDIA V100 32GB
256GB GPU Memory
$0.49 gpu/hr ($2,859/mo)‡
NVIDIA DGX-2
16 x NVIDIA V100 32GB
512GB GPU Memory
$0.59 gpu/hr ($6,899/mo)‡
‡Requires a 1-year agreement