ASUS Desktop AI Supercomputers
Bring data‑centre‑class AI compute to the desktop with ASUS platforms powered by NVIDIA® Grace Blackwell Superchips. Built for modern AI development, these systems are designed for local fine‑tuning, inference, and data science workflows — from compact deployments to deskside AI supercomputing.
Local AI, faster iteration
Run demanding AI workflows close to your data and teams — without waiting on shared cloud capacity.
Unified CPU–GPU memory
Architected to reduce traditional CPU/GPU bottlenecks for large model workflows.
Enterprise software foundation
Includes NVIDIA DGX OS and the NVIDIA AI Software Stack on supported platforms for an AI‑ready environment.
Platforms in this Category
ASUS Desktop AI Supercomputers span compact and deskside form factors — including the ultra‑small ASUS Ascent GX10 and the ASUS ExpertCenter Pro ET900N G3 deskside AI supercomputer.
ASUS Ascent GX10 (Ultra‑Small AI Supercomputer)
A compact platform designed to bring petaflop‑scale AI computing capabilities directly to the desk, powered by the NVIDIA® GB10 Grace Blackwell Superchip and NVIDIA AI software.
ASUS ExpertCenter Pro ET900N G3 (Deskside AI Supercomputer)
A deskside AI supercomputer powered by the NVIDIA® GB300 Grace Blackwell Ultra Desktop Superchip, built for large‑scale AI training and inference workloads with a massive coherent memory pool and data‑centre‑class connectivity options.
Key Features & Value Adds
Grace Blackwell architecture for AI workloads
Supported systems pair an NVIDIA Grace CPU with a Blackwell‑class GPU and NVIDIA NVLink‑C2C chip‑to‑chip interconnect for high‑bandwidth coherent transfers between CPU and GPU — aimed at accelerating AI development and execution.
FP4 & next‑generation Tensor Core acceleration
The ET900N G3 platform highlights Blackwell‑generation CUDA® cores and 5th‑generation Tensor Cores, enabling FP4 AI compute to support next‑generation model efficiency.
Large coherent memory pool (ET900N G3)
Training and running models within a large coherent memory pool helps reduce CPU/GPU bottlenecks typical of traditional architectures — supporting larger models in a unified memory space.
Data‑centre‑class connectivity & scaling
The ET900N G3 references NVIDIA ConnectX®‑8 SuperNIC support with up to 800 Gb/s networking for fast, efficient connectivity, supporting scale‑out and “AI factory” style workflows.
DGX OS & NVIDIA AI Software Stack (Delivered with ASUS Platforms)
ASUS Desktop AI Supercomputers aren’t just hardware — supported platforms include NVIDIA DGX OS and the NVIDIA AI Software Stack to provide an AI‑ready environment for fine‑tuning, inference, and data science workflows.
NVIDIA DGX™ OS
DGX OS is a stable, fully qualified operating system for running AI, machine learning, and analytics applications on NVIDIA DGX platforms. It includes system‑specific configurations, drivers, and diagnostic and monitoring tools.
- Stable OS foundation for AI/ML and accelerated analytics workloads
- Includes system‑specific configuration, drivers, diagnostics, and monitoring tools
- Built to support development environments that can scale beyond a single workstation
NVIDIA AI Software Stack
ASUS platforms reference the NVIDIA AI Software Stack as a full‑stack environment for AI workloads — including fine‑tuning, inference, and data science — enabling teams to move faster from experimentation to execution.
- Full‑stack software environment for AI workloads (fine‑tuning, inference, data science)
- Designed to scale across NVIDIA DGX Station systems, NVIDIA DGX Cloud, and other accelerated data‑centre/cloud infrastructure
- Helps keep workflows consistent as projects grow from desktop development to larger environments
From desk to scaled infrastructure
DGX OS is designed to make it easier to scale across multiple NVIDIA DGX Station systems, NVIDIA DGX Cloud, or other accelerated data‑centre and cloud infrastructure — enabling a path from local development to larger deployments without rebuilding your environment from scratch.
Ideal For
LLM work
Fine‑tuning and inference workflows for large language models and generative AI use cases.
Research & simulation
Deep learning research, multimodal AI experimentation, and simulation workloads.
Enterprise AI teams
Teams needing powerful local compute with an enterprise software foundation and scale‑out path.
Talk to DiGiCOR about ASUS Desktop AI Supercomputers
Tell us your workloads and constraints (data, model size, inference vs training, power/space). We’ll help size the right ASUS platform and deployment approach for Australia & New Zealand.
No spam — just availability updates and practical options.

