Specially developed for the unique requirements of AI
The NVIDIA DGX SuperPOD is a state-of-the-art, turnkey AI data centre solution designed specifically for demanding AI workloads. With a fully integrated infrastructure that combines powerful hardware and optimised software, the DGX SuperPOD delivers maximum performance and scalability for complex AI applications. The liquid-cooled DGX SuperPOD variant with DGX GB200 systems can scale to thousands of NVIDIA GB200-Grace Blackwell superchips and is specifically designed for training and inferencing generative AI models with trillions of parameters. Thanks to full-stack resilience and an intelligent control plane, the DGX SuperPOD ensures constant availability, making it an ideal solution for organisations that want to run their AI projects efficiently and flexibly.
An enterprise solution for the entire life cycle
The NVIDIA DGX SuperPOD with DGX GB200 systems provides a comprehensive enterprise solution that covers the entire lifecycle of AI projects. From development and training to inference, ongoing maintenance and scaling, this platform enables organisations to run their AI applications efficiently and without interruption.
Enterprise infrastructure for mission-critical AI
The NVIDIA DGX SuperPOD™ with DGX GB200 systems is a highly scalable, liquid-cooled AI rack infrastructure that can be expanded to thousands of NVIDIA GB200-Grace Blackwell superchips. Designed for training and inferencing generative AI models with trillions of parameters, it provides full-stack resilience and an intelligent control plane that ensures constant availability for even the most demanding AI deployments. This platform covers the entire lifecycle of AI projects - from development and training to inference, maintenance and scaling - enabling organisations to run their AI applications uninterrupted and efficiently.
What makes the NVIDIA DGX SuperPOD so special?
The NVIDIA DGX SuperPOD is the ultimate solution for organisations looking to take their AI applications to the next level. This turnkey AI data centre solution offers unparalleled computing power, optimised software and seamless operation - all combined in a turnkey infrastructure.
Extreme performance for your AI workloads
The DGX SuperPOD is designed to handle even the most complex and computationally intensive tasks, from large-scale language modelling to deep learning recommendation systems.
Simplified administration
NVIDIA Base Command software makes managing your AI infrastructure a breeze. This software automates deployments, monitoring and health checks to keep your systems running optimally at all times.
Expert support
Benefit not only from a powerful infrastructure, but also from our direct access to NVIDIA experts. Support every step of the way - from installation to optimising your AI workloads.
Technical specifications
The NVIDIA DGX SuperPOD is available in two specialised variants designed to meet different AI requirements. The version with DGX B200 systems is ideal for organisations of all sizes that require scalable infrastructures to support complex and diverse AI workloads. It is particularly suitable for training large language models, optimising supply chains or analysing large volumes of data. The DGX H200 systems, on the other hand, are specifically designed for the most demanding AI applications and offer the highest computing power for transformer-based workloads such as large language models with the NVIDIA NeMo framework and deep recommendation systems. Both variants enable companies to realise their AI projects efficiently, flexibly and on a large scale.
Configuration | Grace-Blackwell GB200 Superchip | GB200 NVL72 |
|---|---|---|
Configuration | 1 Grace CPU: 2 Blackwell | 36 Grace CPU: 72 Blackwell |
FP4 tensor computing unit | 40 PFLOPS | 40 PFLOPS |
FP8/FP6 tensor computing unit | 20 PFLOPS | 720 PFLOPS |
INT8 tensor computing unit | 20 POPS | 720 POPS |
FP16/BF16 tensor computing unit | 10 PFLOPS | 360 PFLOPS |
FP16/BF16 tensor computing unit | 5 PFLOPS | 180 PFLOPS |
FP64 Tensor computing unit | 90 TFLOPS | 3.240 TFLOPS |
Graphics processor memory | bandwidth | Up to 13.5 TB HBM3e | 576 TB/s | Up to 13.5 TB HBM3e | 576 TB/s |
NVLink bandwidth | 3.6 TB/s | 130 TB/s |
Number of CPU computing units | 72 Arm® Neoverse V2 computing units | 2.592 Arm® Neoverse V2 computing units |
CPU memory | bandwidth | Up to 480 LPDDR5X | Up to 512 GB/s | Up to 17 TB LPDDR5X | Up to 18.4 TB/s |
Key features of the NVIDIA DGX SuperPOD
Ready-to-use AI infrastructure
DGX SuperPOD is the world's first ready-to-use AI data centre tailored to the specific requirements of artificial intelligence. It offers a turnkey solution that can be integrated directly into companies.
Scalable and powerful architecture
The platform enables outstanding computing power for any type of AI workload, including large language models and complex data analyses. Companies can use it to develop innovative applications without having to deal with the complexity of the platform.
Proven and advanced software
DGX SuperPOD uses the NVIDIA Base Command software, which facilitates the management of AI workflows and optimises the computing, storage and network infrastructure with specialised libraries. This software is constantly being further developed to ensure maximum efficiency.
Simple administration
The NVIDIA Base Command Manager automates deployments, software updates, monitoring and health checks, greatly simplifying the management of the SuperPOD.
Dedicated expertise and services
Companies benefit from direct access to NVIDIA experts who provide support with the installation, scaling and optimisation of the AI infrastructure.
Why sysGen?
As an Elite Partner of NVIDIA, we not only offer you access to the latest technologies, but also comprehensive support and expertise. We understand the specific challenges of your industry and help you find the right solution to realise your AI projects quickly and efficiently.
Get in touch now!
Rely on the best AI infrastructure on the market
With the NVIDIA DGX SuperPOD, you get the technology that the world's leading companies and research institutions are already using successfully. Let's work together to realise your AI vision and prepare your company for the future.
Contact sysGen today and find out how the NVIDIA DGX SuperPOD can revolutionise your AI strategy. Together we can put your AI projects in the fast lane!