Reserved NVIDIA H100 SXM SuperPods
Utilize the best GPUs in the industry today, starting at $1.80/hr
Reserve Your NVIDIA H100 Cloud Cluster
Enable large-scale model training with top-of-the-line NVIDIA H100 SXM GPUs. Arc Compute's cloud clusters are available for a minimum 2-year commitment.
2-Year
Starting at $2.00/hr per GPU
3-Year
Starting at $1.80/hr per GPU
Instance Type
8 x NVIDIA H100 SXM5 GPUs with 80 GB of GPU memory
vCPUs
224 vCPUs featuring Intel or AMD processors
Storage
Minimum of 6 TB NVMe SSD local storage
Network Bandwidth
3,200 Gbps unblocked InfiniBand
NVIDIA A100 80GB Cloud Instances
6-Month
Starting at $1.20/hr per GPU
1-Year
Starting at $1.15/hr per GPU
3-Year
Starting at $1.05/hr per GPU
Instance Type
8 x NVIDIA A100 SXM4 GPUs with 80 GB of GPU memory
vCPUs
224 vCPUs featuring Intel or AMD processors
Storage
Minimum of 6 TB NVMe SSD local storage
Network Bandwidth
1,600 Gbps unblocked InfiniBand
NVIDIA A100 40GB Cloud Instances
6-Month
Starting at $1.15/hr per GPU
1-Year
Starting at $1.10/hr per GPU
3-Year
Starting at $1.00/hr per GPU
Instance Type
8 x NVIDIA A100 SXM4 GPUs with 40 GB of GPU memory
vCPUs
224 vCPUs featuring Intel or AMD processors
Storage
Minimum of 6 TB NVMe SSD local storage
Network Bandwidth
800 Gbps unblocked InfiniBand
Transformational AI Training
H100 features fourth-generation Tensor Cores and the Transformer Engine with FP8 precision that provides up to 9X faster training over the prior generation for mixture-of-experts (MoE) models. The combination of fourth-generation NVlink, which offers 900 gigabytes per second (GB/s) of GPU-to-GPU interconnect; NVLINK Switch System, which accelerates communication by every GPU across nodes; PCIe Gen5; and NVIDIA Magnum IO™ software delivers efficient scalability from small enterprises to massive, unified GPU clusters.
Deploying H100 GPUs at data center scale delivers outstanding performance and brings the next generation of exascale high-performance computing (HPC) and trillion-parameter AI within the reach of all researchers.
Learn MoreUp to 9X Higher AI Training on Largest Models
NVSwitch + InfiniBand
NVLink is a direct GPU-to-GPU interconnect that scales multi-GPU input/output (IO) within the server and is available in both form factors. Exclusive to SXM5, NVSwitch connects multiple NVLinks to provide all-to-all GPU communication at full NVLink speed within a single node. InfiniBand enables NVSwitch connections to be extended across nodes to create a seamless, high-bandwidth, multi-node GPU cluster - effectively forming a data center-sized GPU, making it possible to solve even the most extensive AI jobs rapidly.
