Showing 730–740 of 893 results

The Supermicro SYS-820GH-TNR2 is a specialized 8U AI Training SuperServer designed to accelerate deep learning and large-scale model training. Built on the X12 generation architecture, this system integrates eight Intel Gaudi 2 accelerators via high-speed OAM interconnects, offering a powerful and cost-efficient alternative to traditional GPU clusters. It features massive on-board networking with six 400GbE ports directly attached to the AI processors, enabling seamless scale-out for massive clusters without complex external switching layers. Ideal for research institutions and enterprise AI labs, this server delivers exceptional compute density and throughput for Generative AI, Computer Vision, and Natural Language Processing workloads.

The Supermicro AS-2114GT-DPNR is a high-density 2U GPU server designed for AI, HPC, and deep learning workloads. Featuring two hot-pluggable nodes, each with a direct-attached backplane supporting two NVMe hot-swap drives, it ensures ultra-fast local storage performance. Each node can host up to three GPUs, including NVIDIA A40 or AMD Instinct MI210, enabling powerful parallel processing for compute-intensive applications. Networking flexibility is provided via AIOM card options, while a dedicated RJ45 port with IPMI 2.0 allows secure remote management, KVM access, and virtual media support. With 80 PLUS Titanium power supplies and a rackmount rail kit included, this server delivers scalable, energy-efficient performance for demanding enterprise and research environments.

Supermicro ARS-111GL-NHR-LCC sets a new benchmark for AI computing, powered by the NVIDIA GH200 Grace Hopper Superchip with integrated H100 GPU. This high-density 1U platform delivers exceptional acceleration for large language models, deep learning training, HPC workloads, and generative AI. It features 480GB ECC LPDDR5X delivering up to 500GB/s throughput and 96GB of HBM3 memory with up to 4TB/s bandwidth for unparalleled data performance. Designed for liquid-cooled efficiency, it includes eight hot-swap E1.S NVMe bays and two M.2 NVMe slots for ultra-fast storage scalability.

The Supermicro ARS-111GL-DNHR-LCC is a high-density 1U GPU barebone system engineered for advanced AI workloads. Featuring two independent nodes, each equipped with an NVIDIA GH200 Grace Hopper Superchip, this platform delivers unmatched performance for LLM training, generative AI, and high-performance computing. Each node combines a Grace CPU with an integrated liquid-cooled NVIDIA H100 Tensor Core GPU, supported by up to 480GB ECC LPDDR5X and 96GB HBM3 memory for extreme bandwidth and accelerated AI processing. With hot-swap E1.S NVMe storage, dual M.2 support, and 2700W power delivery, it provides optimal scalability, efficiency, and mission-critical reliability for next-generation data centers.

Supermicro ARS-111GL-NHR is a high-performance GPU server designed for AI, deep learning, and HPC workloads. Equipped with the NVIDIA GH200 Grace Hopper Superchip and H100 GPU, it delivers unparalleled computational power for large-scale training and inference tasks. The server supports up to 480 GB memory and features 8 hot-plug NVMe drives for high-speed storage and data access. Integrated IPMI 2.0 enables remote management, monitoring, and virtual media support for streamlined administration. Redundant 2000 W 80 PLUS Titanium power supplies ensure energy efficiency and reliability in demanding environments. Delivered with a rackmount kit, this server is ideal for AI research, data analytics, and enterprise GPU-accelerated computing deployments.

The Supermicro ARS-121L-DNR is a high-density 1U two-node platform powered by NVIDIA Grace™ CPU Superchips for next-generation AI and HPC workloads. Each node integrates dual 72-core Arm CPUs connected via a 900GB/s NVLink® C2C interface, delivering exceptional compute throughput and ultra-low latency. Supporting up to 480GB LPDDR5X memory, the system is optimized for accelerated applications with compatibility for single- or double-width NVIDIA GPUs. With PCIe 5.0 x16 slots for BlueField-3 or ConnectX-7 networking and flexible NVMe storage via E1.S and M.2 drives, this liquid-efficient design offers outstanding scalability, energy efficiency, and high-performance data processing for modern data centers.

The Supermicro SYS-2028TP-HC0R is a high-density 2U TwinPro™ system engineered for mission-critical enterprise workloads and scale-out data center deployments. This architecture integrates four independent, hot-pluggable compute nodes into a single 2U chassis, significantly reducing rack space requirements while maintaining high performance. Each node is powered by dual Intel Xeon E5-2600 v3/v4 processors and supports up to 2TB of DDR4 ECC memory across 16 DIMM slots, making it an ideal candidate for virtualization, hyper-converged infrastructure (HCI), and high-performance computing (HPC) clusters.

The Supermicro SYS-621BT-HNC8R is a high-density 2U 4-node BigTwin system engineered for enterprise-grade hyper-converged infrastructure (HCI) and high-performance computing (HPC) workloads. Each of the four hot-pluggable nodes supports dual 4th Gen Intel Xeon Scalable processors, delivering massive core density and compute power within a compact rack footprint. This multi-node architecture is specifically designed to solve space and power efficiency challenges in modern data centers, allowing for granular scaling of compute resources without the overhead of individual chassis management.

The Supermicro SYS-821GE-TNMR2 is an 8U high-density GPU SuperServer engineered for AI, HPC, and advanced analytics workloads. It features dual Intel 5th/4th Gen Xeon Scalable processors supporting up to 350W and 8 AMD Instinct MI300X GPUs for ultra-fast compute acceleration. Memory scales up to 8TB DDR5-4400 in 2DPC or 4TB DDR5-5600 in 1DPC across 32 DIMMs. Storage includes 8 hot-swap SATA bays, 2 M.2 NVMe boot drives, and up to 16 hot-swap NVMe bays (12 default + 4 optional). Flexible networking, PCIe Gen 5.0 expansion slots, and 6 redundant 3000W Titanium power supplies ensure reliable, scalable, and energy-efficient performance. The 8U rackmount chassis provides a compact solution for enterprise and research-grade deployments.

The Supermicro AS-2114GT-DNR is a high-performance 2U GPU server featuring two independent nodes, each powered by AMD EPYC processors and supporting up to 2TB of memory and two NVMe drives. Each node can accommodate up to three GPUs, including NVIDIA A40 or AMD Instinct MI210, making it ideal for AI training, deep learning, and HPC workloads. Flexible networking is enabled through AIOM expansion cards, while IPMI 2.0 provides secure remote management with KVM and virtual media support. Engineered for energy efficiency with 80 PLUS Titanium power supplies, the system includes a rackmount rail kit for easy deployment, delivering a scalable, reliable platform for demanding enterprise and research applications.