• Silver supermicro as-4124gs-tnr 4u dual amd epyc gpu server with black drive bays and cooling fans in sleek industrial design
1 1

Supermicro

Supermicro AS-4124GS-TNR | 4U Dual AMD EPYC GPU Server, Up to 8x GPU, 24x 2.5" Bays

AS-4124GS-TNR

POA
POA
POA
Sale Sold out
Taxes included. Shipping calculated at checkout.

Custom Built Servers

Need a different spec, quantity or configuration?

We build configured-to-order refurbished servers from Cisco, Dell, HPE, NVIDIA Lenovo and Supermicro. Tell us what you need and we'll come back with options within 24 hours.

Configure your server now →

Description

The Supermicro AS-4124GS-TNR is a purpose-built 4U rackmount GPU server designed for AI, deep learning, HPC, and cloud gaming workloads. It supports dual AMD EPYC 7003/7002 series processors with up to 128 cores, up to 8TB DDR4 3200MHz memory across 32 DIMM slots, and accommodates up to 8 double-width GPUs including NVIDIA A100, V100, A40, T4, and AMD Instinct series. The system features 24 hot-swap 2.5-inch hybrid drive bays, dual-root PCIe 4.0 architecture for optimized GPU performance, redundant 2000W Titanium-level power supplies, and onboard dual 1GbE networking with expansion options for high-speed NICs.

Features

Dual-root PCIe 4.0 architecture distributes 4 GPUs per processor for maximized direct-connect GPU performance and reduced latency
- Supports up to 8 double-width GPUs with optional NVIDIA NVLink bridge or AMD Infinity Fabric for GPU-to-GPU communication
- Up to 10x PCIe 4.0 x16 expansion slots (9x with NVMe storage installed) for flexible network, storage, and accelerator card configuration
- Hot-swap storage and power components with redundant 2+2 power supply configuration for high availability
- 80 PLUS Titanium Level (96% efficiency) power supplies reduce operational costs and thermal output
- 8 hot-swap high-speed cooling fans (11.5K RPM) with optimized airflow for GPU thermal management
- Hybrid 2.5-inch drive bay architecture supports mix of NVMe, SAS, and SATA drives with optional RAID configurations
- IPMI 2.0 remote management with KVM-over-IP and virtual media for lights-out administration
- Native onboard RAID 1 support via Marvell 9230 controller for boot drive redundancy
- Supports AMD EPYC 7003 series (Milan), 7002 series (Rome), and Milan-X processors with expanded L3 cache
- Compatible with VMware ESXi, Citrix Hypervisor, Microsoft Windows Server, Red Hat Enterprise Linux, Oracle Linux, and Ubuntu Server
- NVIDIA-Certified system for validated GPU workload compatibility and performance
- Accommodates alternative chassis lids (1U and 0.5U) for improved cable clearance with dense GPU configurations

Warranty

All products sold by XS Network Tech include a 12-month warranty on both new and used items. Our in-house technical team thoroughly tests used hardware prior to sale to ensure enterprise-grade reliability.

All technical data should be verified on the manufacturer data sheets.

View full details

specs-tabs

Technical Specifications

FAQs

Technical Specifications

Form Factor: 4U rackmount server, 437mm x 178mm x 737mm (17.2" x 7" x 29")
- Processors: Dual socket SP3 supporting AMD EPYC 7003/7002 series processors, up to 128 cores total (64 per CPU), up to 280W TDP per CPU
- Memory: 32x DDR4 DIMM slots (16 per CPU), supports up to 8TB DDR4-3200MHz ECC RDIMM/LRDIMM including 3DS modules
- GPU Support: Up to 8x double-width or single-width PCIe 4.0 x16 GPUs with direct-connect dual-root architecture; compatible with NVIDIA A100, V100, A40, T4, Quadro RTX, and AMD Instinct series
- PCIe Expansion: 9x PCIe 4.0 x16 slots (10x slots available without NVMe drives installed)
- Storage: 24x hot-swap 2.5" hybrid drive bays; natively supports 2x SATA (RAID 1 via onboard Marvell 9230) and 4x NVMe PCIe 4.0; optional SAS controller supports up to 18x additional SAS/SATA drives
- Networking: Dual onboard 1GbE Intel Ethernet controllers (2x RJ45); dedicated IPMI 2.0 management port (1GbE RJ45)
- Management: IPMI 2.0 with KVM and virtual media support; compatible with Supermicro Server Manager (SSM), SuperDoctor 5 (SD5), SuperCloud Composer, and Supermicro Update Manager (SUM)
- Power Supply: 4x 2000W redundant (2+2) 80 PLUS Titanium Level (96% efficiency) hot-swap power supplies
- Cooling: 8x hot-swap 11.5K RPM 92mm cooling fans
- BIOS: AMI 32MB SPI Flash ROM
- Front I/O: 2x USB 3.0 ports, 1x VGA port

FAQs

Q: What GPU models are compatible with the AS-4124GS-TNR?
A: The AS-4124GS-TNR supports up to 8 double-width or single-width PCIe 4.0/3.0 GPUs, including NVIDIA A100, V100, A40, T4, Quadro RTX series, and AMD Instinct series. Both actively and passively cooled GPUs are supported. The dual-root PCIe architecture assigns four GPUs to each processor for optimized performance in certain workloads.

Q: Can I configure this server with NVMe storage?
A: Yes, the system natively supports 4x 2.5-inch PCIe 4.0 NVMe drive bays and 2x 2.5-inch SATA drives (RAID 1 via onboard controller). Note that installing NVMe drives reduces available PCIe expansion slots from 10 to 9. With an optional SAS controller, you can configure up to 18 additional SAS/SATA drives for a total of 24 hot-swap bays.

Q: What networking options are available for high-speed data transfer?
A: The AS-4124GS-TNR includes dual onboard 1GbE RJ45 ports as standard. For high-speed networking, PCIe-mounted NICs can be installed in the available PCIe 4.0 x16 slots to support 10GbE, 25GbE, 100GbE, or other high-throughput network interfaces required for GPU-accelerated workloads and data center deployments.

Q: What is the maximum memory capacity and configuration?
A: The server supports up to 8TB of DDR4-3200MHz ECC memory using 32x DIMM slots (16 per processor). Both Registered (RDIMM) and Load-Reduced (LRDIMM) DIMMs are supported, including 3DS variants. A minimum of 4 DIMMs per CPU is required when using AMD EPYC 7003 series processors for optimal memory channel population.

Q: What workloads is this server optimized for?
A: The AS-4124GS-TNR is specifically designed for GPU-intensive applications including AI and deep learning model training, high-performance computing (HPC) simulations, cloud gaming infrastructure, molecular dynamics simulation, and data analytics. The dual-root architecture and support for up to 8 high-end GPUs with NVLink or AMD Infinity Fabric interconnects make it ideal for large-scale parallel processing tasks.

  • Request a Quote

    Looking for competitive pricing? Submit a request, and our team will provide a tailored quote that fits your needs.

  • Contact Us Directly

    Have a question or need immediate assistance? Call us for expert advice and real-time support.

    Call us Now  

  • Contact Us Directly

    Have a question or need immediate assistance? Call us for expert advice and real-time support.

    Contact us