NVIDIA H200
The NVIDIA H200 is an ideal choice for large-scale AI applications. It uses the NVIDIA Hopper architecture that combines advanced features and capabilities, accelerating AI training and inference on larger models.
NVIDIA H200
Infrastructure and technology partners
Perfect for a range of workloads
Deploying AI based workloads on CUDO Compute is easy and cost-effective. Follow our AI related tutorials.
Deploying rendering based workloads on CUDO Compute is easy and cost-effective.
From video editing to image generation, virtualization is ideal for your content creation needs.
Purpose-built H200 clusters, designed and managed by CUDO
Deployed across 16 ISO-certified data centres
From 8 to 1,000+ GPUs in a single deployment
NVIDIA Quantum-X800 InfiniBand or Spectrum-X Ethernet networking
Expert rack-level design, installation, and benchmarking before handoff
24/7 monitoring, management, and engineering support
Compatible with Slurm, Kubernetes, and NVIDIA Base Command
Available at the most cost-effective pricing
Launch your AI products faster with on-demand GPUs and a global network of data center partners
Bare metal
Powered by renewable energy
SpectrumX local networking
8x GPUs per host
No noisy neighbors
300Gbps external connectivity
NVMe SSD storage
Enterprise
Powerful GPU clusters
Scalable data center colocation
Large quantities of GPUs and hardware
Optimize to your requirements
Expert installation
Scale as your demand grows
Specifications
Browse specifications for the NVIDIA H200 GPU
Starting from
Contact us for pricing
Architecture
NVIDIA Hopper
GPU
8x NVIDIA H200 Tensor Core GPUs
GPU memory
1.1 TB total HBM3e (141 GB per GPU), 38.4 TB/s aggregate bandwidth (4.8 TB/s per GPU)
FP64 tensor core performance
34 TFLOPS
FP32
67 TFLOPS
NVIDIA NVSwitch
4x Third-generation NVIDIA NVSwitch
NVIDIA NVLink bandwidth
7.2 TB/s aggregate bidirectional bandwidth (900 GB/s per GPU)
System power usage
~10.2 kW max
CPU
Dual 5th Gen Intel Xeon Scalable Processors (e.g., Platinum 8558P) or AMD EPYC 9004 series equivalents
System memory
2 TB, configurable to 4 TB
Networking
8x OSFP ports serving 8x single-port NVIDIA ConnectX-7 VPI (Up to 400 Gb/s NVIDIA InfiniBand/Ethernet) 1x to 2x NVIDIA BlueField-3 DPUs or additional ConnectX-7 NICs for North/South traffic
Management network
1GbE or 10GbE onboard network interface card (NIC) with RJ45. Host baseboard management controller (BMC) with RJ45
Storage
OS: 2x 1.92 TB (or similar) NVMe M.2, internal storage: 8x 3.84 TB NVMe U.2 (Partner configurable depending on OEM like Supermicro, Lenovo, or Dell)
Software
NVIDIA AI Enterprise (optimized AI software suite included), NVIDIA Base Command, NVIDIA Magnum IO, Ubuntu / Red Hat Enterprise Linux / Rocky
Rack units (RU)
8
Operating temperature
5-30°C (41-86°F)
Ideal uses cases for the NVIDIA H200 GPU
Explore uses cases for the NVIDIA H200 including AI inference, Deep learning, High-performance computing.
AI inference
AI developers can utilize the NVIDIA H200 to accelerate AI inference workloads, such as image and speech recognition, at lightning speed. The H200 GPU’s powerful Tensor Cores enable it to quickly process large amounts of data, making it perfect for real-time inference applications.
Deep learning
The NVIDIA H200 empowers data scientists and researchers to achieve groundbreaking milestones in deep learning. Its massive memory and processing power guarantee significantly reduced training and deployment times for complex, large-scale models and enables model training on significantly larger datasets.
High-performance computing
From complex scientific simulations to weather forecasting and intricate financial modeling, the H200 empowers diverse organizations to accelerate high-performance computing tasks. Its unmatched memory bandwidth and processing capabilities ensure smooth operation for workloads of any scale, allowing you to achieve unmatched results faster than ever.
Sovereign and regulated AI
Deploy B200 clusters in ISO-certified data centres globally. Meet data residency and regulatory requirements with full infrastructure control, your hardware, your jurisdiction, managed by CUDO.
Blog
Resources
- Emmanuel Ohiri
Resources
- Emmanuel Ohiri
Resources
- Emmanuel Ohiri
Resources
- Emmanuel Ohiri
Resources
- Emmanuel Ohiri
Resources
- Emmanuel Ohiri
Browse alternative GPU solutions for your workloads
Access a wide range of performant NVIDIA and AMD GPUs to accelerate your AI, ML & HPC workloads
NVIDIA H100 PCIe
Price on request
Scale with high performance H100 GPUs on our reserved cloud.
