Supermicro H15 SP7 Multi-Node systems are designed to maximize compute density, efficiency, and serviceability across demanding data center environments. Built on AMD EPYC™ 9006 Series processors, these platforms consolidate multiple server nodes within shared chassis designs to optimize performance, power, cooling, and rack space.

From front I/O GrandTwin® systems designed for dense, serviceable compute and GPU acceleration to liquid-cooled FlexTwin™ platforms built for OCP rack environments, the H15 SP7 Multi-Node portfolio supports scale-out workloads across cloud, HPC, AI, virtualization, and enterprise infrastructure.
Back to H15 SP7 Portfolio
Explore Next H15 Systems
Select a Multi-Node Platform
High-Density Multi-Node Platforms for Scalable Compute
H15 SP7 Multi-Node Systems
Multi-node systems help organizations consolidate compute resources while supporting high-performance workloads at scale.
Liquid-cooled, high-density multi-node systems built for OCP rack environments, HPC, and AI workloads.
FlexTwin
Systems
Front I/O multi-node systems designed for dense, serviceable compute with flexible storage, networking, and GPU acceleration options.
GrandTwin
Systems
SP7 MultiNode Systems
Back to H15 SP7 Portfolio
Explore Next H15 Systems
View Details
2U, 2-node front I/O GrandTwin system designed for dense, serviceable compute with flexible storage, networking, and GPU acceleration options.
AS -2117GT7-DNF
Featured Product Lines
Shared Titanium-level power and cooling infrastructure helps improve efficiency across deployments.
Shared Power and Cooling
Expansion options support full-height, half-length, dual-width PCIe cards such as GPU accelerators.
GPU Acceleration Support
Supports multiple drive options and PCIe expansion for workload-specific configurations.
Flexible Storage and Expansion
Two independent single-socket nodes in a 2U chassis maximize compute density.
High-Density Multi-Node Design
Cold-aisle access to storage, networking, and node trays helps simplify installation and maintenance.
Front I/O Serviceability
Key Capabilities:
Shared redundant Titanium-level power with 2x 2200W / 3000W power supply options.
Power Supply
GrandTwin Front I/O module for 10G/25G/200G, 2 USB 3.0, 1 VGA, 1 dedicated IPMI LAN from GrandTwin I/O module, Optional 1 AIOM (OCP 3.0)
Networking & I/O
Up to 4 hot-pluggable 2.5" NVMe drive bays or 8 E3.S 1T / 8 E1.S drives, plus up to 2 M.2 PCIe 5.0 x4 NVMe slots.
Drive Bays
IPMI 2.0, KVM with dedicated LAN, ASPEED AST2700 BMC with RoT, SSM, SDO, TAS, and SAA.
System Management
PCI-Express (PCIe) Configuration: Optional,
 2 PCIe 6.0 x16 (in x16) FHHL slots
Expansion
16 DIMM slots per node supporting up to 4TB ECC DDR5-8000 RDIMM or DDR5-12800 MRDIMM
Memory
Single AMD EPYC™ 9006 Series processor per node, up to 256 cores
Processors
2U rackmount, 2-node front I/O design
Form Factor
Featured System Specs:
Download Datasheet
Supermicro H15 SP7 GrandTwin systems are designed to maximize compute density and simplify serviceability with a 2U, 2-node front I/O architecture. Built on AMD EPYC™ 9006 Series processors, each node delivers high single-socket performance with support for flexible storage, networking, and GPU acceleration options.

With cold-aisle access to storage, networking, and node trays, GrandTwin systems help reduce downtime and simplify maintenance in space-constrained environments. These platforms are well-suited for scale-out workloads including cloud services, HPC, hyperconverged infrastructure, content delivery, big data analytics, and GPU-accelerated applications.
Front I/O Multi-Node Systems for High-Performance Compute
H15 SP7 GrandTwin® Systems
SP7 MultiNode Systems
Back to H15 SP7 Portfolio
Explore Next H15 Systems
View Details
Liquid-cooled FlexTwin system designed for dense HPC and AI workloads in OCP rack environments.
AS -1127FT7-DEO-ALC
Featured Product Lines
Front and rear component access helps support availability and simplified maintenance.
Serviceable Cluster Architecture
Designed for workloads including simulation, financial services, scientific research, AI rendering, and clustered compute.
Optimized for HPC and AI
Built for OCP 3.0 rack environments with centralized power and cooling infrastructure.
OCP Rack-Ready Design
Each node supports dual AMD EPYC™ 9006 Series processors for dense compute performance.
Dual-Socket Nodes
Direct liquid cooling supports high-performance CPUs and helps optimize thermal efficiency.
Liquid-Cooled Compute Density
Key Capabilities:
48–54V DC power busbar with OCP rack power shelf support; 1OU 33kW or 72kW power options
Power Supply
AIOM network connectivity, 1 RJ45 1GbE dedicated BMC LAN port, 1 USB 3.0 Type-A port, Mini-DisplayPort, and TPM onboard / port 80
Networking & I/O
2 E1.S NVMe drive bays by default, plus 2 M.2 PCIe 6.0 x4 NVMe slots.
Drive Bays
Integrated IPMI 2.0, KVM with dedicated LAN, ASPEED AST2700 BMC with RoT, Supermicro Server Manager, Super Diagnostics Offline, and SuperServer Automation Assistant
System Management
PCI-Express (PCle) Configuration: Optional,
1 PCle 6.0 x16 (in x16) FHHL slot or 2 PCle 6.0 x16 (in x16) FHHL slot, 1 PCle 6.0 x16 (in x16) LP slot or 1 AIOM (OCP 3.0)
Expansion
32 DIMM slots per node supporting up to 8TB ECC DDR5-8000 MT/s or up to 8TB DDR5-MRDIMM 12800 MT/s
Memory
Dual AMD EPYC™ 9006 Series processors per node, up to 256 cores per CPU
Processors
1-OU rackmount, OCP rack-ready chassis with two dual-socket nodes
Form Factor
Featured System Specs:
Download Datasheet
Supermicro H15 SP7 FlexTwin systems are designed for high-density, high-performance computing environments where performance, efficiency, and serviceability are critical. Built for OCP rack deployments, FlexTwin combines liquid cooling, centralized rack-level power, and dense multi-node architecture to support demanding HPC and AI workloads.

Each chassis supports two dual-socket nodes powered by AMD EPYC™ 9006 Series processors, enabling up to 4 CPUs and 1024 cores per chassis. With liquid-cooled CPUs, flexible storage and I/O options, and open management tools, FlexTwin helps organizations scale compute-intensive workloads while optimizing power, cooling, and rack space.
Liquid-Cooled Multi-Node Systems for HPC and AI Workloads
H15 SP7 FlexTwin™ Systems