■ 4U 8-GPU AMD Instinct™ MI300X AI Supercomputer
ENTERPRISE RACK BIS CERTIFIED ISO 9001

AARIV 4U MI300X Supercomputer Beyond Limits

Datacenter 4U AI supercomputer node powered by dual AMD EPYC™ 9004 CPUs and 8x AMD Instinct™ MI300X 192GB HBM3 GPUs.

■ TELEMETRY CONTROL

IPMI 2.0 Remote KVM

Dedicated HTML5 web console for 24/7 out-of-band power monitoring, fan diagnostics, and remote ISO deployment.

Active BMC Connectivity: 100G Redundant
+
+
AARIV Server Stage
EXPLORE SERVER SPECIFICATIONS
■ ARCHITECTURE

8x AMD Instinct™ MI300X 192GB

Ultra-high 1.5TB total HBM3 memory per 4U node, accelerating massive local AI inference and generative AI workloads.

99.999% SLA RELIABILITY

Enterprise AI & Research Datacenters

Equipped with 4x 3000W Titanium 3+1 redundant PSUs and 8x 400G InfiniBand NIC slots for AI supercomputer scaling.

Unleash Enterprise Hardware power

8x AMD Instinct™ MI300X


Ultra-high 192GB HBM3 memory per GPU with 8x AMD Instinct™ MI300X accelerators for trillion-parameter AI models.

Dual AMD EPYC™ 9004 Series


Dual 128-core AMD EPYC™ 9754 processors providing max PCIe 5.0 lanes to all GPU nodes.

2TB DDR5 ECC & 8x NVMe


2TB DDR5 ECC Registered memory and 8x E1.S/E3.S Gen5 NVMe drives for fast AI checkpointing.

Quad 3000W Titanium PSUs


4x 3000W 80 Plus Titanium 3+1 redundant power supplies engineered for 24/7 AI datacenter cluster loads.

CRAFTED TO ELEVATE

Where innovation meets customization.

Elevate your enterprise compute with AARIV 4U GPU Supercomputer (AMD) — designed in Pune and assembled with precision high-density engineering.

Form Factor & Chassis
4U High-Density Datacenter GPU Server
Material Build
Heavy-Duty Electro-Galvanized Industrial Steel
Power Supply Unit
4x 3000W 80+ Titanium Redundant PSU
Thermal Cooling System
High-CFM Hot-Swappable Redundant Blower Fans
Processor Silicon
Dual AMD EPYC™ + 8x AMD Instinct™ MI300X
Memory Infrastructure
Multi-Channel DDR5 Registered ECC Memory
Outer Dimensions (mm)
W 448 D 850 H 175 (4U)
(45.0 KG net system weight).
PCIe & Drive Expansion
8x OAM/SXM GPU Slots, 8x 400G InfiniBand NICs
*BIS Safety Standard & ISO 9001 certified.
KEY INNOVATIONS

Enterprise Engineering Features

AMD PRO & Memory Guard

AMD PRO & Memory Guard

Hardware-level AMD Memory Guard encryption and dedicated Secure Processor.

Explore Security ›
PCIe Gen5 High Speed Lanes

PCIe Gen5 High-Speed Bus

Up to 128 PCIe 5.0 lanes delivering max bandwidth to GPUs, RAID controllers, and 100G NICs.

Explore Expansion ›
IPMI 2.0 Out-of-Band Management

IPMI 2.0 Remote BMC

Dedicated LAN port with HTML5 web console for 24/7 remote power monitoring and KVM management.

Explore Remote BMC ›
3-Year Commercial SLA

3-Year Commercial SLA

Backed by OMBASE 3-Year commercial warranty with 4-hour engineer dispatch across 28 states.

Explore Warranty ›
AMD EPYC™ POWERED GPU SUPERCOMPUTER PORTFOLIO

Enterprise Configurations & Silicon Options

AMD EPYC™ PLATFORM SERVERS ONLY
RACK SERVER 1U/2U • AMD EPYC 9004

OMBASE 1U/2U Rack Server (AMD)

128-core AMD EPYC™ 9004 high-density rack server node for enterprise cloud virtualisation.

OMBASE 1U/2U Rack Server (AMD)
Dual AMD EPYC™ 9754 (256 Cores) • 1TB ECC DDR5 • 12x NVMe U.2 Bays
GPU SUPERCOMPUTER • 4U EPYC + 8x MI300X

AARIV 4U GPU Supercomputer (AMD)

Datacenter 4U AI node with 8x AMD Instinct™ MI300X / NVIDIA H100 GPUs for massive AI LLM workloads.

AARIV 4U GPU Supercomputer (AMD)
Dual AMD EPYC™ 9654 + 8x AMD Instinct™ MI300X 192GB • 2TB ECC DDR5

Find the Right Platform for Your Enterprise

From ultra-light AI laptops to high-density datacenter GPU supercomputers, explore hardware tailored for your workloads.

The AI Laptop for everything you do

Find the ultra-light business laptop that fits your daily productivity with 45+ TOPS integrated NPU acceleration.

What an OMBASE AI PC does for you

See how local LLMs, stable diffusion image generation, and enterprise security work seamlessly behind the scenes.

A look into OMBASE Workstation T-Series

Take a detailed walkthrough into dual-socket Intel® Xeon® and quad NVIDIA RTX 6000 Ada workstation power.

Explore the AARIV 4U GPU Supercomputer

Find the right high-density rack server node engineered for enterprise AI datacenter deployment and cluster scaling.

4U datacenter AI supercomputer node for large language model training.

AARIV 4U GPU Supercomputer (AMD) combines dual AMD EPYC™ 9004 CPUs, 8x AMD Instinct™ MI300X 192GB GPUs, and 3000W Titanium redundant PSUs for AI clusters.

SHOP NOW

Contact OMBASE Sales & Service

Our enterprise experts and dedicated technical support teams are here to assist you.

Contact for Sales

Get official enterprise hardware quotes, bulk government procurement specs, RFP pricing, and tailored custom configurations for your organization.


Contact for Support & Service

Reach our 24/7 technical customer desk for official driver downloads, BIOS updates, warranty SLA registration, and 4-hour on-site dispatch servicing.


Contact Support →