• NVIDIA HGX AI supercomputer

    NVIDIA HGX AI supercomputer The most powerful end-to-end platform for AI supercomputing

sysGen/SUPERMICRO10U GPU-Server inkl. NVIDIA HGX B200 8-GPU SXM
sysGen/SUPERMICRO
10U GPU-Server inkl. NVIDIA HGX B200 8-GPU SXM
SYS-A22GA-NBR
CPU
2x Intel Xeon 6960P (72-Cores, 2.70 GHz)
RAM
2TB (16x 128GB) DDR5-6400MHz ECC RDIMM
SSD OS
2x 1.92TB Gen4 NVMe
SSD Storage
8x 3.8TB Gen4 NVMe
GPU
NVIDIA HGX B200 8-GPU SXM
Total GPU-Memory size of 1,440GB HBM3e
Network
8x 400GbE / Infiniband NDR OSFP
2x 100GbE QSFP56
2x 10GbE RJ-45
1x Dedicated IPMI Management
Support
incl. 3 Years Parts Warranty
On Request
Request Now
sysGen/SUPERMICRO10U GPU-Server inkl. NVIDIA HGX B200 8-GPU SXM
sysGen/SUPERMICRO
10U GPU-Server inkl. NVIDIA HGX B200 8-GPU SXM
SYS-A126GS-TNB
CPU
2x AMD EPYC TURIN 9965 (192-Cores, 2.25 GHz)
RAM
3TB (24x 128GB) DDR5-6000MHz ECC RDIMM
SSD
2x 1.92TB Gen4 NVMe
SSD Storage
8x 3.8TB Gen4 NVMe
GPU
NVIDIA HGX B200 8-GPU SXM
Total GPU-Memory size of 1,440GB HBM3
Network
8x 400GbE / Infiniband NDR OSFP
2x 100GbE QSFP56
2x 10GbE RJ-45
1x Dedicated IPMI Management
Support
incl. 3 Years Parts Warranty
On Request
Request Now
sysGen/SUPERMICRO8U GPU-Server inkl. NVIDIA HGX H200 8-GPU SXM5
sysGen/SUPERMICRO
8U GPU-Server inkl. NVIDIA HGX H200 8-GPU SXM5
SYS-821GE-TNHR
CPU
2x Intel Xeon Platinum 8468 (48 Cores, 2.10GHz)
RAM
2TB (32x 64GB) DDR5 4800MHz ECC RDIMM
SSD
1.92TB Gen4 NVMe
GPU
NVIDIA HGX H200 8-GPU 141GB SXM5
Network
2x 400GbE / Infiniband OSFP
1x Dedicated IPMI Management
Support
incl. 3 Years Parts Warranty
267.162,- €NETTO
Request Now
sysGen/SUPERMICRO8U GPU-Server inkl. NVIDIA HGX H100 8-GPU SXM5
sysGen/SUPERMICRO
8U GPU-Server inkl. NVIDIA HGX H100 8-GPU SXM5
SYS-8125GS-TNHR
CPU
2x AMD EPYC GENOA 9654 (96 Cores, 2.40GHz)
RAM
1.5TB (24x 64GB) DDR5 4800MHz ECC RDIMM
SSD
2x 3.84TB Gen4 NVMe
GPU
NVIDIA HGX H100 8-GPU 80GB SXM5
Network
2x 10GbE RJ-45
1x Dedicated IPMI Management
Support
incl. 3 Years Parts Warranty
256.370,- €NETTO
Request Now
sysGen/SUPERMICRO4U GPU-Server inkl. NVIDIA HGX H100 4-GPU SXM5
sysGen/SUPERMICRO
4U GPU-Server inkl. NVIDIA HGX H100 4-GPU SXM5
SYS-421GU-TNXR
CPU
2x Intel Xeon Platinum 8558 (48-Cores, 2.10 GHz)
RAM
1TB (16x 64GB) DDR5-4800MHz ECC RDIMM
SSD
2x 3.84TB Gen4 NVMe
GPU
NVIDIA HGX H100 4-GPU SXM5
Network
2x 25GbE SFP28
2x 10GbE RJ-45
1x Dedicated IPMI Management
Support
incl. 3 Years Parts Warranty
136.447,- €NETTO
Request Now
Specially developed for the convergence of simulations, data analyses and AI

Massive data sets, huge deep learning models and complex simulations require multiple GPUs with extremely fast connections and a fully accelerated software stack. The NVIDIA HGX™ platform combines powerful GPUs with fast NVLink and InfiniBand interconnects and an optimised software stack from the NVIDIA NGC catalogue. This enables maximum performance for AI training, simulations and data-intensive analyses. Thanks to its end-to-end performance and flexibility, NVIDIA HGX enables researchers and scientists to combine simulations, data analyses and AI to drive scientific progress.

For Large Scale AI TrainingLarge Language Models, Generative AI Training, Autonomous Driving, Robotics
For Large Scale AI TrainingLarge Language Models, Generative AI Training, Autonomous Driving, Robotics

Large AI models in areas such as language processing, generative AI, autonomous driving and robotics require a scalable computing infrastructure. Platforms such as NVIDIA HGX are designed to efficiently bundle parallel GPU performance - for training complex models with large amounts of data in a reduced time frame. High-performance connections between GPUs and the direct networking of several nodes enable high-performance processing of even very extensive parameter structures.

For HPC/AIEngineering Simulation, Scientific Research, Genomic Sequencing, Drug Discovery

For HPC/AI

Engineering Simulation, Scientific Research, Genomic Sequencing, Drug Discovery

Research institutions and engineering teams are increasingly relying on GPU-accelerated computing power to carry out complex simulations and data-intensive analyses more efficiently. Applications range from computational fluid dynamics and molecular simulation to genome research and drug discovery.

The systems from sysGen in co-operation with Supermicro are based on the NVIDIA HGX platform, support flexible configurations and can be scaled for rack operation.

Relevant solution on the topic of AIEngineering Simulation, Scientific Research, Genomic Sequencing, Drug Discovery

Relevant solution on the topic of AI

Engineering Simulation, Scientific Research, Genomic Sequencing, Drug Discovery

To shorten the time to discovery for scientists, researchers and engineers, more and more HPC workloads are being augmented with machine learning algorithms and GPU-accelerated parallel computing. Many of the world's fastest supercomputing clusters now utilise GPUs and the power of AI.

HPC workloads typically require data-intensive simulations and analyses with massive data sets and precision requirements. GPUs like NVIDIA's H100/H200 offer unprecedented performance with double the precision, delivering 60 teraflops per GPU.

Our Supermicro's highly flexible HPC platforms enable high GPU and CPU counts in various dense form factors with rack-scale integration and liquid cooling.

First step

Contact sysGen now

Are you interested in our solutions or do you have further questions? Contact us now and find out more about the most powerful end-to-end platform for AI supercomputing. We will be happy to advise you and find the perfect solution for your requirements.

Allgemeine Daten
Weitere Informationen

FAQ - NVIDIA HGX platform

  • What characterises the NVIDIA HGX platform?

    The NVIDIA HGX platform combines scalable GPU performance with optimised system architecture for data-intensive workloads. Thanks to direct GPU connections (NVLink, NVSwitch) and a customised software stack, complex AI models and simulations can be processed efficiently.

  • For which areas of application is the HGX platform suitable?

    HGX-based systems are used in areas such as deep learning, engineering simulation, scientific research, genomics and drug development. They are particularly suitable for computationally intensive, parallelised applications with large amounts of data.

  • Which technologies are used in HGX systems?

    The platform integrates advanced GPU architectures from NVIDIA with direct GPU-GPU communication via NVLink and NVSwitch as well as fast network connections such as InfiniBand. This is complemented by management tools, optimised drivers and support for frameworks from the NVIDIA NGC catalogue. These include NVIDIA HGX H100, NVIDIA HGX H200, NVIDIA HGX B200 and other current models.

  • How can an HGX system be integrated into existing data centres?

    HGX systems have a modular design and can be integrated into standard racks. sysGen offers scalable solutions - air or liquid cooled - with optional support for GPU clusters, storage connection and management software. Integration is customised to existing IT environments.

Ihre optimale Website-Nutzung

Diese Website verwendet Cookies und bindet externe Medien ein. Mit dem Klick auf „✓ Alles akzeptieren“ entscheiden Sie sich für eine optimale Web-Erfahrung und willigen ein, dass Ihnen externe Inhalte angezeigt werden können. Auf „Einstellungen“ erfahren Sie mehr darüber und können persönliche Präferenzen festlegen. Mehr Informationen finden Sie in unserer Datenschutzerklärung.

Datenschutzerklärung Impressum
Detailinformationen zu Cookies & externer Mediennutzung

Externe Medien sind z.B. Videos oder iFrames von anderen Plattformen, die auf dieser Website eingebunden werden. Bei den Cookies handelt es sich um anonymisierte Informationen über Ihren Besuch dieser Website, die die Nutzung für Sie angenehmer machen.

Damit die Website optimal funktioniert, müssen Sie Ihre aktive Zustimmung für die Verwendung dieser Cookies geben. Sie können hier Ihre persönlichen Einstellungen selbst festlegen.

Noch Fragen? Erfahren Sie mehr über Ihre Rechte als Nutzer in der Datenschutzerklärung und Impressum!

Datenschutzerklärung Impressum
Ihre Cookie Einstellungen wurden gespeichert.