Skip to main content
Unveiling H13 System Portfolio

Featuring Linus Tech Tips logo

Thumbnail for Episode 0: Intro/Overview H13 server powered by AMD EPYC


Linus Sebastian joins Supermicro to get a first look on their latest H13 server lineup powered by AMD EPYC™ 9004 series processors, and see firsthand what it means for these systems to be Better, Faster, Greener.

Thumbnail for Episode 1: CloudDC

Episode 1: CloudDC

Follow Linus and Clinton as they showcase Supermicro’s most versatile single-socket solution. The H13 CloudDC systems offer a cost-optimized solution with ultimate expandability for a wide range of web and cloud applications.

Thumbnail for Episode 2: Hyper

Episode 2: Hyper

What do you get when you take the CloudDC server and double it? With up to 24 PCIe Gen 5 Memory, CXL 1.1+ support, expansion slots for GPU and accelerators, Linus takes you inside the box of the H13 Hyper systems and its unparalleled versatility and performance.

Thumbnail for Episode 3: GrandTwin

Episode 3: GrandTwin

Journey with Linus to discover incredible performance, density, and efficiency packed into every H13 GrandTwin™ server, with up to 4x 96 core single processor nodes, shared power, shared cooling, and that’s not all…

Thumbnail for Episode 4: GPU Systems

Episode 4: GPU Systems

Linus goes face to face with the 8U uGPU system featuring 8x NVIDIA H100 GPUs, and the flexible 4U GPU systems offering maximum acceleration to tackle the most intensive HPC, AI, ML applications out there.

Thumbnail for Episode 5: AMD Rack Server

Episode 5: Rack Scale Integration and Deployment

Linus heads into the testing facilities to show you Supermicro’s rack scale integration from design to deployment, and what their L11 and L12 testing and validation offers.

H13 Generation with AMD EPYC™ 9004 Series Processors

H13 GPU Optimized Systems

H13 4U GPU System

The Supermicro 4U GPU servers are 4U dual processor systems, supporting up to 10 FHFL double-width PCIe GPU cards, including the latest AMD Instinct MI200 Series and NVIDIA H100 GPUs. The 4U GPU-optimized systems provide maximum acceleration, flexibility, and balance for AI, deep learning, and HPC applications.

H13 Multi-Node GrandTwin Systems

H13 GrandTwin™ System

The Supermicro GrandTwin™ are the newest multi-node architecture solution with Front and Rear I/O, designed for maximum density and purpose-built for single-processor performance per node. The flexible modular design is optimized for a range of applications, with a front I/O option simplifying installation for space-constrained environments.

H13 Hyper Systems

H13 Hyper System

The Supermicro Hyper solutions are enterprise-focused servers built with versatility and performance. Uncompromised performance design with dual processors and 12 channel 24 DIMMs optimized for supporting the highest TDPs and offering a flexible range of computing, networking, storage, and I/O expansion capabilities.

H13 CloudDC Systems

H13 CloudDC System

The Supermicro CloudDC are single-socket servers optimized for I/O flexibility with many cloud-focused applications. Offering convenient serviceability with tool-less brackets, hot-swap drive trays, and redundant power supplies ensure rapid deployment and efficient maintenance in data centers.

H13 8U Universal GPU System

H13 8U Universal GPU System

Propel AI/ML workloads with dual 4th Gen AMD EPYC processors and the latest NVIDIA HGX H100 8-GPU. The 8U Universal GPU System provides unparalleled I/O and thermal capacity. It supports 8 of 700W TDP GPUs with NVLink, GPUDirect Storage and RDMA with 8 of 400G networking to enable 1:1 GPU ratio to keep feeding deep learning models at massive scale.

H12 Systems, Powered by AMD EPYC™ 7003/7002 Series Processors
Choose from one of the most comprehensive lineups in the industry


  • Dual sockets and 32 DIMMs optimized for supporting the highest processor TDPs
  • Best-in-class server features including all NVMe, hybrid storage and low latency optimization
  • Vast networking and expansion possibilities with Ultra Riser card


  • World record performance with up to 17% uplift with AMD 3D V-Cache Technology on SPECjbb
  • Up to 20 hot-pluggable blades, integrated 200G HDR InfiniBand switch, and 4x 25GbE switches or pass-through modules
  • GPU blades supporting up to 20 AMD Instinct™ MI200/MI100 Series Accelerators

BigTwin® and TwinPro®

  • Highly configurable 2U 4-node systems
  • 2-socket with 16 DIMMs per node (BigTwin) or 1-socket with 8 DIMMs per node (TwinPro)
  • Flexible storage and I/O options including NVMe/SATA3 and SIOM networking


  • Highly configurable 4U, 4- and 8-node systems
  • 1-socket with 16 DIMMs per node
  • Flexible NVMe and SATA3 drive options with front or rear I/O


  • 1U or 2U 1-socket, 16 DIMMs with tool-less design for rapid deployment and easy maintenance
  • 12 (in 2U) or 10 (in 1U) SATA drive bays with optional NVMe support
  • 4 (in 2U), or 2 (in 1U) PCIe 4.0 x16 slots, as well as dual AIOM slots for flexible networking


  • Highly simplified 1- and 2-socket designs, 2U or mid-tower form factor
  • Up to 16 DIMMs, and flexible NVMe or SATA3 drive options
  • Built-in Gigabit Ethernet connectivity and Platinum-Level efficiency power supplies