Skip to main content
Shop with CelesteCeleste

Celestehere!

I’m your shopping assistant for PC builds, upgrades, and everything tech.

Got questions? Ask away!

Start a conversation

Get to know me

Hey!

How can we help you today?

Track a packageStart a returnCancel an orderResolve a payment issueBilling, declined cards, and payment method updateMy Account & SecurityUpdate passwords and security settings

Need more help?

Visit Help Center
All Categories
Build Tools & Resources New & Trending Offers for You
For your BusinessNew
Your Browsing History
  1. Home
  2. AI Workstation
  3. Servers & Workstations
  4. Server Barebones
  5. ASUS

  6. Item#: N82E16816110257

Out of stock
  • $10,430.00
  • Free 91-day Returns
Sold by Newegg

Shipped by Newegg

Price alert
Add to Wish List
Shop it at NeweggBusiness.
Shop All ASUS Products

ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server | Dual Intel Xeon 6 Scalable | 32x DDR5 ECC | 8x Hot-Swap NVMe | PCIe 5.0 | 3200W Titanium Redundant PSU

Be the first to review this product...
  • "4U flagship GPU chassis, supports up to 8 dual-slot full-size AI accelerators
  • Dual Intel Xeon 6 Scalable processors, powerful next-gen multi-core parallel computing
  • 32x DDR5 ECC registered memory, large capacity & high bandwidth for massive AI models
  • 8 hot-swappable Tri-Mode NVMe high-speed all-flash storage bays
  • Latest PCIe Gen 5.0 high-speed expansion architecture, NVLink GPU interconnection support
  • ASMB12-iKVM KVM-over-IP remote BMC out-of-band management, TPM-SPI security
  • 3+1 redundant 3200W Titanium high-efficiency ultra-stable power supply"
+
+
Overview
Reviews
See all Q & A

Any questions? Ask CelesteSM will help you find out quickly.

Try me now
ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server
ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server
ESC8000-E12P
Intel® Xeon® 6 dual-processor 4U NVIDIA® MGX server that supports eight dual-slot NVIDIA®H200 GPUs, NVIDIA RTX PRO™ 6000 Blackwell Server Edition, NVIDIA RTX PRO™ 4500 Blackwell Server Edition, or Intel® Gaud®3 PCIe AI Accelerator, up to 32 DIMM, five PCIe® 5.0 slots and eight 2.5-inch NVMe.
ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server

Form Factor


4U

ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server

CPU Number


2

ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server

Memory Number


32

KEY FEATURES
  • Equipped with Intel Xeon 6 processors featuring 86 cores, eight-channel DDR5 memory up to 6400MHz and a maximum TDP of 350 watts per socket
  • Fully compatible with NVIDIA MGX architecture, enabling fast and large-scale deployment
  • High-density 4U server for eight dual-slot high-end NVIDIA H200 GPUs, NVIDIA RTX PRO™ 6000 Blackwell Server Edition, NVIDIA RTX PRO™ 4500 Blackwell Server Edition, or Intel® Gaudi® 3 PCIe AI Accelerator, supporting up to 600 watts each
  • Optimized server configuration, five PCI 5.0 slots for high-bandwidth PCIe NICs and DPU to enable performance scaling
  • ASUS-exclusive toolless design empowers easy maintenance for maximum efficiency
  • ASUS Control Center IT management software and hardware-level Root-of-Trust solution
Eight dual-slot GPUs in a high-density 4U compute server

Powered by Intel Xeon 6 dual-processors, ESC8000-E12P delivers exceptional performance, featuring top-tier GPUs, faster GPU interconnects, and higher-bandwidth fabric. This powerhouse server supports up to eight dual-slot GPUs, either active or passive, and offers flexible scalability with options for NVIDIA® NVLink® Bridge. This configuration allows seamless scaling and performance enhancement, making ESC8000-E12P an ideal choice for tackling the most Enterprise AI and HPC tasks.

ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server
Scale with ease and flexibility using MGX architecture

ESC8000-E12P takes AI infrastructure to the next level by leveraging NVIDIA® MGX™, an open and flexible reference design that ensures future-proof compatibility. With its innovative modular architecture offering up to 160 customizable configurations, clients can easily tailor their solutions to meet specific AI and high-performance computing needs. MGX sets a new industry benchmark for modular server design, enhancing ROI and slashing time to market — empowering businesses to stay ahead in the fast-evolving AI landscape.

ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server
ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server
Intel Xeon 6 processors

Intel Xeon 6 processors represent a significant leap forward in performance and efficiency, specifically designed to meet the demands of advanced AI, networking and data center workloads. Featuring an all-P-core architecture, these processors deliver exceptional computational power, enabling enterprises to tackle the most demanding applications with ease. With support for high-speed DDR5 memory and PCIe 5.0 I/O, the Intel Xeon 6 processors ensure seamless scalability and enhanced data throughput, making them ideal for modern high-performance environments. Additionally, their optimized energy efficiency and robust design cater to the evolving needs of next-generation data centers, solidifying Intel's leadership in AI and networking solutions.

ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server
Supercharge your system with PCIe 5.0 connectivity switch solutions

This server features the latest Broadcom PEX89000 series of switches, supporting up to 1,024Gbps (128GBps) of raw bandwidth per port (x16) to create high-availability hyper-scale systems. It uses the PCI Express 5.0 bus to empower communication between in-rack hosts and endpoints, simplifying connectivity while providing the highest PCIe-switching performance available for data center servers, storage and networks. Take advantage of these industry-first features today for the most demanding hyper-converged AI/ML/DL applications.

Effortless maintenance with a toolless design for speed and simplicity

Designed for easy maintenance with a refined mechanical structure, ESC8000-E12P allows the entire machine to be disassembled completely without tools.

ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server

Toolless removal

ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server

Quick installation

ASUS ESC8000-E12P-32W 4U 8-GPU AI Training Server

Toolless removal and quick installation for PCIe card

Server-management Solution

Warranty & Returns

Warranty, Returns, And Additional Information

Warranty

  • Limited Warranty period (parts): 3 years
  • Limited Warranty period (labor): 3 years
  • Read full details

Return Policies

    We want to ensure your holiday shopping this year is hassle-free. That’s why we’re automatically extending our return policies on almost every product! For purchases placed between November 1st, 2026 and December 31, 2026 you will have until January 31, 2027 to return your product.

  • Return for refund until: January 31, 2027
  • Return for replacement until: January 31, 2027
  • This item is covered by Newegg.com's Extended Holiday Return Policy.
  • Read full return policy for details.

Manufacturer Contact Info

LOADING...

The ASUS ESC8000-E12P-32W is a 4U barebones server designed for organizations building AI training or high-performance computing systems, with support for dual Intel Xeon 6 processors and up to eight dual-slot full-size accelerators.

Recommended For

  • AI teams training or fine-tuning models across multiple compatible GPUs in a rack environment; the 4U platform is designed for up to eight dual-slot accelerators.
  • Organizations deploying GPU-backed inference or AI services that need a shared, centrally managed server rather than a desk-side workstation.
  • Data-center and infrastructure teams building high-density accelerator nodes, with rack space, suitable power, and server-management operations already in place.
  • Teams combining accelerator compute with fast local datasets or checkpoints that benefit from multiple front hot-swap NVMe bays.

Real-World Use Scenarios

  • An AI team training or fine-tuning models across multiple GPUs can configure the server with supported accelerators and run the workload in one 4U system; the eight-GPU design is intended for dense AI compute, though actual results depend on the selected GPUs and software.
  • A data-center team serving generative AI or LLM applications can use the server as a multi-GPU inference host; its dense accelerator layout is suited to consolidating GPU compute, with capacity and throughput depending on the installed accelerators and deployment.
  • An HPC group running GPU-accelerated workloads such as scientific computing can use the system as a shared compute node; its GPU and high-bandwidth expansion options support that role when matched to the application and network.
  • An organization building or expanding an enterprise AI cluster can deploy the server as a configurable GPU node; its MGX-based design and networking expansion slots provide options for fitting it into a larger infrastructure, subject to the chosen configuration.
  • A data-center administrator managing a remote GPU server can use out-of-band management to monitor or troubleshoot it without relying on the host operating system; this is useful for routine maintenance and recovery in a rack-based deployment.

Spec Summary

AttributeValue
Best ForAI training workloads that benefit from a multi-GPU server platform