ASUS Ascent GX10 Compact AI Supercomputer - 2TB

ASUS Ascent GX10 Compact Desktop AI Supercomputer - 2TB (GX10-GG0030BN) /  (90MS0371-M000Y0) - 1 Year Onsite Warranty + Limited-Time Offer: Free Upgrade to 3 Years Warranty.

ASUS Ascent GX10 – Compact AI Supercomputer for Advanced Workflows

The ASUS Ascent GX10 Compact Desktop AI Supercomputer is built for creators, researchers, and engineers who require extreme performance in a space-efficient form factor. It brings supercomputer-class AI processing to the desktop, enabling high-speed data analysis, model training, rendering, and advanced simulation without needing a full-scale rack system.

AI-Optimized Performance for Intensive Computing

Powered by ASUS’s next-generation AI-accelerated design, the GX10 delivers exceptional compute power for machine learning, deep learning, and generative AI workloads. From LLM experimentation to professional 3D rendering, this system ensures smooth, optimized performance across the most demanding applications.

2TB High-Speed Storage for Large Datasets

Equipped with 2TB of ultra-fast NVMe storage, the Ascent GX10 offers massive room for training datasets, project files, media assets, and AI model checkpoints. The high read/write throughput ensures minimal latency, faster loading, and seamless workflow transitions across data-intensive tasks.

Compact Design with Advanced Thermal Engineering

Despite its compact footprint, the GX10 features intelligent cooling architecture to maintain stable performance under continuous heavy workloads. ASUS engineered the system for quiet operation and thermal efficiency, making it suitable for research labs, studios, and office environments where performance and reliability must coexist.

Technical Specification: 

Processor / Compute Architecture

• Built on the NVIDIA GB10 Grace-Blackwell Superchip, integrating CPU + GPU into a unified AI-optimized platform

• Delivers next-generation AI performance with up to 1 petaFLOP FP4 compute (depending on configuration)

• Unified memory architecture reduces bottlenecks for LLMs, generative AI, multimodal workloads, and large dataset processing

Graphics / AI Acceleration

• NVIDIA Blackwell-based GPU architecture with advanced tensor cores

• Fully optimized for CUDA, TensorRT, NVIDIA AI Enterprise, and end-to-end GPU acceleration frameworks

• Supports large-scale inference, model fine-tuning, vector embeddings, and enterprise-level AI deployment

Memory

• High-bandwidth LPDDR5X Unified Memory

• Enables fast CPU-GPU data sharing for real-time AI workloads

• Suitable for running transformer models, vision pipelines, and multi-instance inference environments

Storage

• 2TB PCIe NVMe SSD

• High-speed sequential read/write performance for model checkpoints, datasets, and local AI storage

• Ideal for edge-AI, embedded AI systems, and enterprise research workloads

Connectivity & I/O

• Network:

  • Built-in 10GbE LAN

  • NVIDIA ConnectX-7 SmartNIC for advanced clustering and high-bandwidth network acceleration

• Ports:

  • 3 × USB-C (USB 3.2 Gen 2×2, 20Gbps + DisplayPort Alt-Mode)

  • 1 × USB-C power input

  • 1 × HDMI 2.1 display output

• Wireless:

  • Wi-Fi 6E + Bluetooth support (varies by configuration)

Cooling System

• Advanced Quiet-Cooling Thermal Design engineered for sustained full-load AI compute

• Ensures stable performance during long training and inference cycles

• Low acoustic profile suitable for office, lab, and desktop environments

Operating System Support

• Compatible with NVIDIA DGX OS (Ubuntu-based AI-tuned environment)

• Supports major Linux distributions for machine learning frameworks

• Windows 11 Pro supported for general use, developer workflows, and administrative tasks

Form Factor

• Ultra-compact chassis: 150 × 150 × 51 mm

• Lightweight, portable design (~1.48 kg)

• Ideal for desks, research labs, edge deployments, or multi-node clusters

Warranty

• Standard 1-Year Onsite Warranty

• Additional service extensions available