
The world's smallest personal AI supercomputer
ASUS Ascent GX10 is a powerful mini PC designed for local AI work. In a body with a volume of approximately one litre, it offers computing power more commonly found in specialised servers. The device is built on the NVIDIA Grace Blackwell GB10 superchip, which combines a 20-core ARM architecture with 5th generation Tensor cores. With up to 1 petaflop of performance and 128GB of unified memory, it enables local training, fine-tuning, and inference of large language models without the need for cloud services. The ATAGB10-9000 model is equipped with a 2TB M.2 NVMe PCIe 5.0 SSD. Suitable operating systems include NVIDIA DGX OS and Ubuntu Linux. The integrated NVIDIA ConnectX-7 SmartNIC ensures high-speed, low-latency communication. This makes the AI TOP ATOM suitable for data scientists, researchers, and developers who require control over their data and computing environment.
The NemoClaw ASUS Ascent GX10 mini PC with the open-source NVIDIA NemoClaw suite complements OpenClaw with privacy and security controls. NemoClaw installs the NVIDIA OpenShell tool, which enforces privacy rules and security guardrails based on set policies. This gives you control over agent behaviour and data handling. The ASUS Ascent GX10 provides the computing resources to run open models, such as NVIDIA Nemotron, locally, increasing privacy and cost-effectiveness.
With 1 petaFLOP of AI performance and 128GB of memory, the ASUS Ascent GX10 lets you run agentic AI (such as OpenClaw¹ or other frameworks) on demand. By running your preferred AI models locally, you achieve greater cost predictability², enterprise-grade stability, and enhanced protection against unwanted data leaks³.
¹The ASUS Ascent GX10 supports and is compatible with third-party open-source software such as OpenClaw, Hermes Agent...etc. OpenClaw is a trademark/product of its respective owners.
²API dependency refers to the local operation of open-source models (e.g., Llama, Qwen) using the ASUS Ascent GX10's own hardware. Third-party services, premium licensed models, or web plugins may still be charged separately by their providers. Savings on API tokens are indicative and depend on the user's specific workflows and choice of local open-source models. ASUS does not provide API keys or licenses for proprietary third-party models. Users are responsible for complying with the terms of use of any deployed AI models and third-party software integrated into OpenClaw or other frameworks.
³Data leakage can occur if network access is configured or third-party services are used. Applies to standard local setups with external network access disabled.
The ASUS Ascent GX10 proves that you no longer need a server room or a rack full of hardware to run the largest AI models. Measuring 150 × 150 × 50.5mm and weighing 1.2kg, the ASUS Ascent GX10 fits on any desk next to a laptop—while offering performance that, just a few years ago, would have required an entire server full of graphics cards. Four USB-C 4.0 ports with speeds up to 40Gb/s, Wi-Fi 7, 10Gb Ethernet, and an HDMI output ensure all your accessories can be connected without needing an external hub or docking station. The ASUS Ascent GX10 is designed for developers, scientists, and AI professionals who want the full power of a local, personal AI supercomputer without relying on the cloud—right where they work.
The ASUS Ascent GX10 handles even the most demanding AI tasks thanks to its sophisticated cooling solution. A pair of fans with seven-speed control ensures a smooth and precise airflow that adjusts to the current workload. This system offers 1.6× more effective cooling than comparable compact stations. The result is a lower operating temperature and stable maximum performance, even under prolonged load.
A single GX10 can handle models with up to 200 billion parameters. Two interconnected GX10s push the boundary much further. Thanks to the integrated NVIDIA ConnectX-7 network technology, you can combine two ASUS Ascent GX10 systems into a single computing unit to handle models like Llama 3.1 with 405 billion parameters. The ultra-fast data throughput between nodes with minimal latency is ideal for large-scale distributed AI tasks, while hardware acceleration for TLS, IPsec, and MACsec ensures encrypted transmission without burdening the processor.
The power of the Grace Blackwell architecture in a body that fits on any desk – the ASUS Ascent GX10 is the ideal choice for developers, researchers, and data scientists who need full-fledged AI performance without compromise. Whether you are working on large language model inference, fine-tuning pre-trained networks, or developing AI agents, the ASUS Ascent GX10 handles the entire spectrum of AI tasks locally, quickly, and without depending on cloud infrastructure.
The ASUS Ascent GX10 supports the complete NVIDIA AI software stack, which makes it easy to train, manage, and deploy AI tasks in a local environment. The device lets you scale performance across your local infrastructure and integrate it into broader AI solutions without relying on cloud services.
Use the 128 GB of unified memory in the ASUS Ascent GX10 to fine-tune pre-trained models with up to 70 billion parameters directly at your workstation. Training on your own data enables you to specialise AI models for specific needs, industry data, or particular use cases – without sending sensitive data to the cloud or paying for expensive GPU instances. The result is a model precisely tailored to your scenario, created locally, securely, and under your full control.
The combination of 128 GB of unified memory and 1 PFLOP of performance in parallel throughput makes the ASUS Ascent GX10 the ideal workstation for demanding data analytics and machine learning projects. Extensive datasets, complex computational models, and demanding training and analysis processes that previously required a powerful cloud cluster or a dedicated server now run directly on your desk. Fast, local, and without waiting for remote infrastructure.
5th generation Tensor Cores with FP4 format support achieve up to 1 PFLOP of performance and, combined with 128 GB of system memory, let you run inference on state-of-the-art AI models with up to 200 billion parameters directly on your desktop. Test, verify, and deploy models like Llama, Gemma, or Qwen locally and in real time – without cloud latency, shared infrastructure, or the risk of sensitive data leaks.
The ASUS Ascent GX10 is an exceptional platform for developing robotic systems, smart city solutions, and computer vision applications. Pre-installed NVIDIA Isaac frameworks for robotics, Metropolis for video analysis, and Holoscan for real-time data processing, enable developers to fully harness the power of the ASUS Ascent GX10 and quickly prototype and deploy applications – locally, without dependency on cloud infrastructure, and with full support from the NVIDIA ecosystem.
ASUSTek COMPUTER INC. 1F.,No.15, Lide Rd., Beitou Dist., Taipei City 112 Tel: +886-2-2894-3447 Tel: +886-2-2890-7798 https://gr.asus.com/ProductSafety ASUS COMPUTER GmbH Harkortstrasse 21-23 40880 Ratingen Germany, Europe Email: [email protected]
ASUS COMPUTER GmbH, Harkortstrasse 21-23, 40880 Ratingen, Germany, Europe, [email protected]
ASUS is a symbol of the growth of Taiwan's economy. The firm's beginnings date back to April 2, 1989, when as a small manufacturer of motherboards began his work with a few employees. Today, Asus employs more than 12,500 employees worldwide.
ASUS products can be found in almost every category of the information technology marketplace. ASUS is focused on desktops, laptops, phones, optical drives and components. In the manufacture of motherboards, the company is now the dominant player with a staggering 40% market share.
Specifications can be changed without notice. Images are for illustrative purposes only.