AceleMax DGS-428WS
Summary:
The AceleMax DGS-428WS is a 4U Dual Processor (DP) server designed for high-performance tasks, supporting 3rd Generation Intel® Xeon® processors. It is optimized for artificial intelligence (AI), machine learning (MI), and high-performance computing (HPC) applications. Key features include:
- Support for Dual 3rd Gen Intel® Xeon® Scalable Processors with up to 270W TDP.
- 32 DIMM slots with a capacity of up to 8TB of ECC DDR4 3200MHz SDRAM.
- Equipped with 8 NVIDIA A100 SXM4 GPUs.
- High power efficiency with 6x 3000W Redundant Titanium Level PSU.
LiquidMax™ LA-2
Summary:
The LiquidMax™ LA-2 Workstation is a liquid-cooled system designed for high-performance computing, particularly in AI and machine learning applications. It features support for 7 NVIDIA Data Center GPUs and a single socket 3rd Gen AMD EPYC processor. Key aspects of this workstation include:
- Full liquid cooling for both CPU and GPU, maximizing peak performance.
- Supports up to 7 full-height, full-length (FHFL) GPUs such as the NVIDIA H100 and RTX 6000 Ada Generation.
- Incorporates a single socket 3rd Gen AMD EPYC processor with up to 64 cores.
- Flexible and high scalability with 6 hot-swap bays and multiple PCIe Gen4 GPU support.
AceleMax DL-E440
Summary:
The AceleMax DL-E440W is a versatile and powerful workstation, designed as a Tower/4U Dual Processor (DP) unit. It is capable of supporting up to four double-width GPUs and is powered by Dual 3rd Generation Intel® Xeon® Scalable Processors. The key features of the AceleMax DL-E440W include:
- Support for up to 4 double-width GPUs.
- Equipped with Dual 3rd Gen Intel® Xeon® Scalable Processors.
- Designed for applications in research, scientific computing, simulation, design, machine learning, and computer-aided engineering.
- Suitable for AI compute, model training, deep learning, real-time high-quality multi-GPU ray tracing, and high-performance 3D simulation.
AceleMax DGS-428AS
Summary:
The AceleMax DGS-428AS is a high-density 4U GPU server designed for advanced computing tasks, featuring support for dual AMD EPYC processors and eight NVIDIA A100 SXM4 GPUs. Key highlights of this server include:
- High GPU peer-to-peer communication enabled by NVIDIA NVLink.
- Low latency with direct attached GPUs and next-gen PCIe 4.0 support.
- Support for two AMD EPYC™ 7002 or 7003 series processors family.
- Eight NICs for GPU Direct Attach RDMA and 4 NVMe GPU Direct Storage.
AceleMax DGS-410W
Summary:
The AceleMax DGS-410W is a powerful 4U Dual Processor (DP) server designed for data center applications, particularly those requiring intensive GPU computing. This server is optimized for a variety of advanced computing tasks, including AI, machine learning, and high-performance computing. Key features of the AceleMax DGS-410W include:
- Dual 3rd Gen Intel® Xeon® Scalable Processors support, with up to 270W TDP.
- Capacity for 10 NVIDIA GPUs, compatible with various models like A100, A40, A10, and others.
- Up to 8TB of DDR4 memory across 32 DIMM slots.
- High-efficiency power supply with four 2000W Titanium level units.
AceleMax DGS-260W
Summary:
The AceleMax DGS-260W is a 2U Dual Processor server designed for high-performance data center applications. It supports the 3rd Generation Intel® Xeon® processors and is equipped with six PCIe GPUs. This server is particularly suitable for tasks involving artificial intelligence, machine learning, deep learning, and high-performance computing. Key features include:
- Dual 3rd Gen Intel® Xeon® Scalable Processors support, with up to 270W TDP.
- Accommodates 6 double-width, full-height, full-length NVIDIA GPUs.
- 16 DIMM slots supporting up to 4TB ECC DDR4-3200MHz memory.
- Two 2600W High-efficiency (Titanium level) power supplies.
AceleMax DGS-224AS
Summary:
The AceleMax DGS-224AS is a high-density 2U server designed for advanced computing tasks. It features four NVIDIA HGX A100 GPUs and supports dual AMD EPYC 7002 or 7003 series processors. Key aspects of this server include:
- Hardware-level root of trust support.
- Supports NVIDIA HGX A100 with 4 x SXM4 GPU.
- Capabilities for up to 600GB/s GPU to GPU interconnection via NVIDIA NVLink and NVSwitch technology.
- Dual AMD EPYC 7003 Series processors and 16 DIMM slots for DDR4 memory.
AceleMax DGS-140W
Summary:
The AceleMax DGS-140W is a 1U dual processor server designed for a range of high-performance computing tasks. It supports the 3rd Generation Intel® Xeon® processors and is equipped with four PCIe GPUs. The server is tailored for machine learning, artificial intelligence, deep learning training, and high-performance computing applications. Key features of the AceleMax DGS-140W include:
- Dual 3rd Gen Intel® Xeon® Scalable Processors support.
- Compatibility with 4 NVIDIA GPUs including models like A100, A40, and A10.
- 16 DIMM slots for up to 4TB of DDR4-3200 MHz memory.
- Two 2000W High-efficiency (Titanium level) power supplies.
ServMax® X-52-AG
Summary:
The ServMax X-52-AG is a high-performance 5U server system optimized for AI and GPU-intensive applications. It supports dual 4th Gen. Intel® Xeon® processors and features the NVIDIA HGX™ H100 8-GPU design. Key features include:
- Dual 4th Gen. Intel® Xeon® processor support with a maximum TDP of 350W and 8GB DDR5 4800 memory.
- Integration with the latest NVIDIA HGX™ H100 8-GPU design, facilitating streamlined AI development and deployment.
- Energy-efficient design with 4+2 3000W (240V) 80 PLUS Titanium redundant power supply.
- 12 low-profile Gen5 x16 and 1 Gen4 x16 expansion slots, with 8 2.5' Gen5 NVMe/SATA/SAS hot-swappable HDD/SSD bays.
AceleMax DGS-828GS
Summary:
The ServMax X-52-AG is a 5U server system designed for AI and GPU-intensive workloads. It incorporates dual 4th Gen. Intel® Xeon® processors and features NVIDIA HGX™ H100 8-GPU design. Key features include:
- Dual 4th Gen. Intel® Xeon® processor support with a maximum TDP of 350W and 8GB DDR5 4800 memory.
- Integration with NVIDIA HGX™ H100 8-GPU design for AI development and deployment.
- Energy-efficient with 4+2 3000W (240V) 80 PLUS Titanium redundant power supply.
- Extensive expansion capabilities with 12 Gen5 x16 and 1 Gen4 x16 expansion slots, and 8 hot-swappable HDD/SSD bays.
AceleMax DGS-528AS
Summary:
The AceleMax DGS-528AS is a 5U HPC server that offers robust computing capabilities, primarily designed for GPU-intensive tasks. It supports NVIDIA HGX™ H100 with 8 x SXM5 GPUs, making it suitable for high-performance computing applications. Key features include:
- NVIDIA HGX™ H100 support with 8 x SXM5 GPUs for advanced computing performance.
- AMD EPYC™ 9004 series processor family support, leveraging the latest in processor technology.
- High-speed interconnects with up to 900GB/s GPU to GPU communication.
- Extensive expansion options including 2 x 10Gb/s BASE-T LAN ports, 2 M.2 slots, and multiple PCIe slots.
AceleMax DGS-728ES
Summary:
The AceleMax DGS-728ES is a 7U server designed for large-scale AI and HPC workloads, featuring the NVIDIA HGX H100 8-GPU SXM5 design. Key highlights of this server include:
- Dual 4th Gen Intel Xeon Scalable processors support, optimized for AI and HPC tasks.
- Incorporates NVIDIA HGX H100 eight-GPU server technology for advanced AI capabilities.
- Features up to 12+1 PCIe slots, 32 DIMM slots, and supports 10 NVMe drives.
- Includes dual 10Gb LAN and OCP 3.0 (optional), with direct GPU-to-GPU interconnect via NVLink providing 900GB/s bandwidth.
AceleMax DGS-528ES
Summary:
The AceleMax DGS-528ES is a 5U high-performance computing server designed for intensive AI and HPC workloads. It features the NVIDIA HGX H100 with 8 x SXM5 GPUs. Key aspects of this server include:
- NVIDIA-Certified system ensuring scalability, functionality, security, and performance.
- Support for NVIDIA HGX™ H100 with 8 x SXM5 GPUs.
- NVIDIA® NVLink® and NVSwitch™ technology for high-speed GPU interconnections.
- Up to 900GB/s GPU to GPU interconnection speed.
AceleMax DGS-428A
Summary:
The AceleMax DGS-428A is an 8U air-cooled server featuring 8x AMD Instinct MI300X Accelerators, designed for high-performance computing applications. Key features of this server include:
- Dual AMD EPYC 9004 Series Processors with up to 128 cores/256 threads per socket.
- A total of 1.5TB HBM3 memory from the 8x AMD Instinct MI300X Accelerators.
- Up to 896GB/s of peer-to-peer inter-GPU connectivity with AMD Infinity Fabric™ Links.
- 6TB of 4800MT/s DDR5 Memory via 24 DIMMs, along with advanced networking options.
AceleMax DGS-214A
Summary:
The AceleMax DGS-214A is a 2U server featuring liquid cooling and equipped with 4 AMD Instinct MI300A Accelerators. It's designed for high-performance computing tasks. Key features include:
- 4x AMD Instinct MI300A with a total of 912 AMD CDNA 3 GPU compute units.
- Peer-to-peer inter-GPU connectivity up to 384GB/s via AMD Infinity Fabric Links.
- 96 AMD “Zen4” 9004 cores across 4x AMD Instinct MI300A APUs.
- Unified 512 GB HBM3 memory with up to 5.3 TB/s bandwidth, and up to 400G networking fabrics.
AceleMax DGS-214GC
Summary:
The AceleMax DGS-214GC is a 2U server optimized for AI and HPC workloads, supporting 4x PCIe 5.0 GPUs. Its key features include:
- AMD EPYC™ 9004 series processors with AMD 3D V-Cache™ technology.
- PCIe Gen5.0 ready, enhancing data throughput and processing speed.
- Tailored for AI and HPC workloads, providing robust computing power.
- Comprehensive cooling solutions, ensuring optimal performance and reliability.
StorMax J-4390
Summary:
The StorMax J-4390 is a 4U dual-path storage enclosure designed for high availability applications. It is optimized for performance and reliability in data-intensive environments. Key features of the StorMax J-4390 include:
- Dual-path storage enclosure for high availability applications, ensuring continuous data access.
- 1+1 redundancy with dual canister controllers for enhanced data protection.
- Supports 90x 3.5”/2.5” top loading SAS3 12Gb/s hot-swappable HDDs, providing extensive storage capacity.
- Equipped with 6x SAS3.0 IO ports (MiniSAS HD/SFF-8644) per node, offering high-speed data transfer capabilities.
ServMax X-22424-L
Summary:
The ServMax X-22424-L is a 2U liquid-cooled system designed for high scalability. This server is tailored for demanding applications requiring efficient cooling and high processing power. Key features include:
- Support for Dual 5th/4th Gen Intel® Xeon® Scalable processors.
- 32 DIMM slots supporting up to 8TB of DDR5-5600 memory.
- Up to 8 PCIe 5.0 x8 or 4 PCIe 5.0 x16 slots, supporting 4x double-width GPUs.
- Up to 24x 2.5” hot-swap NVMe/SATA/SAS drive bays for enhanced storage options.
AceleMax X-48-CH200
Summary:
The AceleMax X-48-CH200 is a 4U liquid-cooled GPU system designed for high-performance computing tasks. It supports NVIDIA HGX H100 SXM5 Board and is optimized for AI and deep learning applications. Key features include:
- Support for Dual 5th/4th Gen Intel® Xeon® Scalable processors.
- 32 DIMM slots supporting up to 8TB of DDR5-5600 memory.
- Up to 7x PCIe 5.0 x8 or 4 PCIe 5.0 x16 slots, compatible with 4 double-width GPUs.
- Up to 6x 2.5” hot-swap NVMe/SATA/SAS drive bays for versatile storage options.
ServMax X-224-L
Summary:
The ServMax X-224-L is a 2U liquid-cooled system, ideal for scalable high-performance setups. It's designed for demanding applications that require efficient cooling and high processing power. Key features of the ServMax X-224-L include:
- Support for Dual 5th/4th Gen Intel® Xeon® Scalable processors.
- 32 DIMM slots, supporting up to 8TB of DDR5-5600 memory.
- Up to 8 PCIe 5.0 x8 or 4 PCIe 5.0 x16 slots, compatible with 4x double-width GPUs.
- Up to 16x 2.5” hot-swap NVMe/SATA/SAS drive bays for extensive storage capacity.
ServMax X-1224
Summary:
The ServMax X-1224 is a 1U application server, designed for efficient and high-performance tasks. It supports dual Intel 4th/5th Gen Xeon® Scalable processors, making it suitable for demanding workloads. Key features include:
- Dual 5th/4th Gen Intel® Xeon® Scalable processors support.
- 16 DIMM slots, allowing up to 4TB of DDR5-5600 memory.
- 2x PCIe 5.0 x16 full-height, full-length expansion slots.
- Up to 4x 3.5”/2.5” hot-swap NVMe/SATA/SAS drive bays for versatile storage solutions.
Ultrastar® EDGE
Summary:
The Ultrastar Edge Server is a high-performance edge computing solution designed for remote data capture and analytics. It is well-suited for low-latency applications and can function in environments with intermittent or non-existent network connections. Key features include:
- Extreme performance with 40 CPU cores and 512GiB memory, suitable for intensive data processing.
- Equipped with a GPU for AI/ML tasks, enhancing its capabilities in advanced computing.
- Over 60TB of Ultrastar NVMe™ flash storage, providing substantial data storage capacity.
- Features high-performance 100GbE networking and Trusted Platform Module 2.0 for enhanced security.
ServMax® A-22-IG
Summary:
The ServMax A-22-IG is a 2U rackmount GPU server optimized for high-density computing with support for dual AMD EPYC™ 7003 Series processors and advanced cooling solutions. It is designed to accommodate multiple GPUs, making it ideal for AI, ML, and high-performance computing tasks.
Key Features:
- Processor Support: Dual AMD EPYC™ 7002 processors, supporting up to 64 cores per processor and max TDP of 280W
- Memory Capacity: Features 16x DDR4 RDIMM slots, expandable up to 3200MT/s
- GPU Configuration Options: Supports up to 16x single-slot GPUs or 8x double-slot GPUs
- Storage and Expansion: Offers a range of storage options including 2.5" hot-swappable SAS/SATA drives and Gen4 U.2 ports, along with multiple PCI-E slots for expansion, ensuring ample storage and adaptability for expansion.
ServMax® X-22-IG
Summary:
The ServMax® X-22-IG is a 2U GPU server optimized for high-density computing environments, supporting dual 3rd Gen. Intel® Xeon® Scalable processors and up to 8 double-slot GPUs. It's designed for efficiency and high availability, with features geared towards minimizing downtime and protecting data.
- Immersion Cooling Ready: Ensures full warranty coverage, including system and components, for immersion cooling environments.
- High Density and Performance: Supports dual 3rd Gen Intel® Xeon® Scalable processors, catering to demanding compute tasks.
- Enhanced Data Protection: Features Smart Crises Management and Protection (SCMP) and Smart Ride Through (SmaRT) to prevent server downtime and data loss during power failures.
- Flexible Storage Options: Offers a variety of hot-swappable HDD/SSD bays and expansion slots, including support for SATA, SAS, and Gen4 NVMe devices.
ServMax® X-21-IN
Summary:
The "ServMax X-21-IN" is a high-density server designed for intensive computing tasks. It features high expandability and is equipped to support advanced AI workloads with its capability to house multiple NVIDIA GPUs and NVMe drives.
- Immersion Cooling Ready: Comes with a full warranty covering both system and components.
- GPU-Optimized: Offers an optional GPU-optimized design allowing for dual-slot GPUs suitable for AI workloads, including support for NVIDIA A100 GPUs.
- High Storage Capacity: Equipped to handle up to eight all-flash NVMe and four SATA/SAS drives on the front, plus an additional two NVMe in the back.
- Extensive Connectivity: Includes 6 PCIe Gen4 x16 FH/FL slots, 2 PCIe Gen4 x8 FH/HL slots, and 1 OCP NIC 3.0 PCIe Gen4 x16 for diverse and advanced management and expansion capabilities.
ServMax® A-28-IN
The ServMax® A-28-IN is a high-density server designed for demanding data center environments, optimizing performance, scalability, and cooling efficiency. This 2U 4-node server is specifically engineered to support up to 8 AMD EPYC™ 7003 series processors, offering a powerful computing solution for applications requiring high throughput and processing power. It is immersion cooling ready, ensuring reliability and performance even under extensive workloads.
- Immersion Cooling Ready: Full warranty coverage that encompasses both the system and its components, ensuring reliability and performance under extensive workloads.
- Processor and Memory Support: Each node can accommodate two 3rd Gen AMD EPYC™ 7003 series processors and includes 16 DDR4 DIMM slots, providing substantial computing power and memory capacity.
- Flexible Storage Options: Features include 2x 2.5” Gen4 U.2 hot-swappable SSD bays and 1x M.2 slot with PCIe Gen4 x4 interface per node, offering versatile storage configurations.
- Connectivity and Expansion: Offers 2 low profile half-length PCIe x16 slots, 1 OCP 3.0 mezzanine slot with PCIe Gen4 x16 bandwidth, and multiple USB and RJ45 ports for comprehensive connectivity and expansion capabilities.
ServMax® A-11-IG
The ServMax® A-11-IG is a 1U GPU server engineered for high-performance computing environments, especially those that benefit from or require immersion cooling technology. This server is optimized for computational intensity and AI workloads, supporting a single 3rd Gen AMD EPYC™ 7003 processor with up to 64 cores. It can accommodate up to 6 NVIDIA A100 GPUs, offering exceptional parallel computing power. Additionally, it provides storage flexibility with up to four internal 2.5” NVMe SSDs, making it suitable for a wide range of data-intensive applications.
- Immersion Cooling Ready: Comes with a full warranty covering both the system and components, designed for environments utilizing immersion cooling technology for enhanced efficiency and cooling performance.
- High-Core Processor Support: Utilizes a single AMD EPYC™ 7003 series processor with AMD 3D V-Cache™ Technology, supporting up to 64 cores for intensive computing tasks.
- GPU Acceleration: Capable of supporting up to 6x NVIDIA A100 GPUs, ideal for AI, machine learning, and high-performance computing applications.
- Robust Storage and Power: Features up to 4 internal 2.5" NVMe SSDs for fast storage options and is equipped with a 1+1 redundant 2000W Platinum power supply for reliable operation.
StorMax X-21-IG
The StorMax X-21-IG is a 2U 26-bay storage system designed for high-density data storage, leveraging Intel Xeon Scalable processors for enhanced performance. It's suitable for a wide range of applications, from data centers to cloud storage solutions. Key features include:
- Single-phase immersion-ready server design for efficient cooling.
- Support for 2nd Gen Intel Xeon Scalable and Intel Xeon W-3200 processor families.
- Offers up to 24x 3.5” SATA/SAS hot-swappable bays plus 2x 2.5” SATA hot-swappable bays.
- Equipped with dual 1Gb/s LAN ports (Intel I210-AT) and a management LAN port for robust connectivity options.
ServMax® A-22-IG-16
The ServMax A-22-IG-16 is a 2U DP server designed for immersion cooling, supporting up to 16 single-slot GPU cards. It leverages AMD EPYC™ 7003 series processors for powerful computing capabilities. Here are four key features:
- Single-phase immersion-ready server configuration.
- Capable of supporting 16 single-slot GPU cards for high-density computing tasks.
- Utilizes AMD EPYC™ 7003 series processors with dual processor 7nm technology.
- Features 8-channel RDIMM/LRDIMM DDR4 per processor, totaling 16 DIMM slots.
AceleMax DGS-324A
The AceleMax DGS-324A is a 3U server equipped with 4 AMD Instinct MI300A Accelerators and additional PCIe slots for GPUs. This server is designed for high-performance computing applications. Key features include:
- 4 AMD Instinct MI300A APUs with a total of 912 AMD CDNA 3 GPU compute units.
- Peer-to-peer inter-GPU connectivity up to 384GB/s via AMD Infinity Fabric Links.
- Each APU includes 24 Zen4 9004 cores and provides 128GB of onboard HBM3 memory, totaling 512GB.
- Dual 10G NIC and support for 8 hot-swap 2.5” U.2 NVMe drives.
NVIDIA A-800
The NVIDIA A800 40GB Active GPU datasheet outlines a powerful workstation development platform designed for AI, data science, and high-performance computing (HPC). It leverages the NVIDIA Ampere architecture to deliver breakthrough performance, handling complex data science workflows, AI training and inference, and large-scale simulations with speed and efficiency. Key features include NVIDIA Ampere Architecture, Third-Generation Tensor Cores, and Ultra-Fast HBM2 Memory, enabling accelerated processing and enhanced scalability for demanding computing tasks.
- Powerful Compute Capabilities: The A800 40GB Active GPU is built to accelerate end-to-end data science workflows, AI development, and HPC workloads with its high-speed memory and scalability.
- Advanced Features for Efficiency: It includes Third-Generation NVIDIA NVLink and Multi-Instance GPU (MIG) for improved performance and resource utilization.
- Support for AI Development: Each GPU comes with a three-year subscription to NVIDIA AI Enterprise, offering access to AI frameworks, libraries, and tools for rapid development and deployment of AI and data science applications.
AceleMax X-88-CH200
The AceleMax X-88-CH200 is an 8U air-cooled system designed for high-density GPU computing. It supports 8 NVIDIA H200 SXM5 GPUs, making it ideal for demanding AI and HPC workloads. Here are four key features:
- Dual Socket AMD EPYC™ 9004 Series Processors for high-performance computing.
- Supports HGX H200 GPUs, enhancing LLM inference and fine-tuning performance.
- Offers up to 6TB of 4800MHz DDR5 RAM across 24 DIMMs for extensive memory capacity.
- Features 18 x 2.5” hot-swap NVMe/SATA drives and flexible networking options for superior storage and connectivity.
The AceleMax DGS-428A
The AceleMax DGS-428A is a 4U server equipped with quad AMD Instinct™ MI300A APUs, designed for high-performance computing with a focus on advanced processing and memory capabilities. Key features include:
- Supports up to 512GB HBM3 memory for high-speed data processing.
- Offers flexible PCIe 5.0 expansion slots for customization and scalability.
- Provides versatile networking and storage options, including AIOM compatibility and hot-swap drive bays.
- Features 2700W redundant power supplies and advanced cooling for reliability in intensive workloads.
AceleMax A-528X
The AceleMax A-528X is a 5U dual processor server, leveraging AMD EPYC™ 9004 Series processors and AMD Instinct™ MI300X OAM GPU modules. It is specifically designed for demanding compute tasks in fields like artificial intelligence (AI), machine learning (ML), and high-performance computing (HPC). The server's architecture supports groundbreaking speeds and capacities in processing and memory, making it a robust solution for intensive computational workloads. Key features include:
- Support for 8 x AMD Instinct™ MI300X OAM GPU modules, providing high-speed GPU to GPU interconnections at 128GB/s.
- Utilization of AMD EPYC™ 9004 Series processors, including options with AMD 3D V-Cache™ Technology, emphasizing its focus on cutting-edge processing power.
- Capacity for dual processors using 5nm technology, with up to 128-core 256 threads per processor, illustrating the server's immense computational potential.
- Memory support for 24 x DDR5 DIMMs, enabling high-speed memory configurations up to 4800 MHz, catering to applications that require extensive memory bandwidth and capacity.
AceleMax DGS-828ES
The AceleMax DGS-828ES is an 8U Dual Processor (DP) system designed for demanding computing environments, supporting 5th Gen and 4th Gen Intel® Xeon® Scalable processors. It integrates NVIDIA HGX H100 8-GPU for advanced applications in AI/Deep Learning, Industrial Automation, Healthcare, and more. This system is engineered for high performance, reliability, and scalability in various applications, including business intelligence, drug discovery, and climate modeling. Key features include:
- Support for Dual Socket E (LGA-4677) with 5th Gen and 4th Gen Intel® Xeon® Scalable processors.
- Equipped with NVIDIA SXM: HGX H100 8-GPU (80GB) for unparalleled computational power.
- Features 32 DIMM slots for extensive memory capacity and 15 hot-swap drive bays for flexible storage options.
- Includes advanced networking options and 6x 3000W Redundant Titanium Level power supplies for high power efficiency.
- Compatible with a wide range of operating systems and designed with 10 heavy duty fans for optimal thermal management.
AceleMax® AXG-728IS
NVIDIA HGX H200 8-GPU SXM5 H200 AI Server for Generative AI Workloads
• 7U NVIDIA HGX H200 eight-GPU server with dual 4th Gen Intel Xeon Scalable processors, designed for large-scale AI and HPC and with up to 12+1 PCIe slots, 32 DIMM, 10 NVMe, dual 10Gb LAN and OCP 3.0 (optional)
• Direct GPU-to-GPU interconnect via NVLink delivers 900GB/s bandwidth for efficient scaling
• Energy-efficient design with independent CPU- and GPU-airflow tunnels for power savings