ASUS GX10 Compact AI Compute Platform
Compact, High-Performance GenAI Compute for Secure Development and Edge
Prototyping a Large Language Model (LLM), running multimodal inference, or building an AI-in-a-box solution? AI teams want fast iteration and control. Conventional desktop workstations can't keep pace. Memory capacity, CPU to GPU bandwidth, and performance drops under sustained load can slow iteration and limit how far you can scale models and pipelines.
Cloud resources can help, but they can also introduce cost variability, added latency, and data governance concerns around sensitive datasets and IP.
The ASUS Ascent GX10 is designed to solve these issues and provide the ideal balance of compute performance, system scalability and compact deployment. Powered by the NVIDIA® GB10 Grace Blackwell Superchip and NVIDIA DGX™ Spark, it delivers high-density AI compute in a compact desktop system.
What You Can Do with ASUS Ascent GX10
ASUS Ascent GX10 delivers high-density AI compute in a compact desktop system allowing you to:
- Run LLMs up to 200B parameters locally
- Fine-tune and iterate with 128GB unified system memory
- Scale performance by stacking systems using ConnectX-7 networking
- Reduce latency and control costs by keeping more workloads close
- Keep sensitive data and IP closer to where it is created and used
- Work in familiar developer environments on Ubuntu Linux
ASUS Ascent GX10 Highlights
The GX10 combines a developer-optimized platform with scalable architecture in an ultra-small form factor.
- Extreme AI performance: Built on the NVIDIA® GB10 Grace Blackwell Superchip delivering up to 1 petaFLOP of AI performance and 128GB unified memory for 200B model fine-tuning
- Developer-optimized platform: Built for AI researchers and developers with full-stack compatibility and streamlined workflows
- Scalable architecture: Integrated NVIDIA® NVLink™-C2C for ultra-fast CPU to GPU memory communication and NVIDIA® ConnectX-7 networking that supports dual GX10 system stacking
- Advanced thermal design: Engineered cooling for sustained performance and reliability
- Minimal footprint, maximum efficiency: Compact system designed for labs, offices, and edge environments
- Full stack AI solution: GB10 plus the NVIDIA AI software stack for AI development and deployment
Key Specifications
At the core of the GX10 is an architecture built for high-throughput AI processing in a small footprint.
| Category | Specifications |
|---|---|
| Performance and Architecture |
|
| Memory | 128GB LPDDR5x unified system memory |
| Storage Options |
|
| Networking |
|
| Operating System | Ubuntu Linux |
| Form Factor |
|
What We Deliver with ASUS Ascent GX10
Build complete GX10-based solutions or source ASUS Ascent GX10 systems, with reliable supply and expert support from Arrow System Builder. We help you select the right configuration and deliver a ready-to-deploy platform that fits your performance, connectivity, and rollout requirements.
With ASUS Ascent GX10, Arrow System Builder supports OEMs, integrators, and solution developers with:
- Configuration support for repeatable builds and program scale
- Integration guidance for networking, storage, and deployment readiness
- Lifecycle planning and supply chain support for production rollouts
- System design support for lab, office, and edge environments
Edge AI Teams, OEM Solution Builders, and Infrastructure Teams
ASUS Ascent GX10 is built for teams that need serious desk-side AI capability and a clear path to repeatable deployments. It supports fast iteration and fits solution builds where footprint, reliability, and scaling matter. Common use cases include:
- LLM prototyping and pipeline iteration
- Vision and language inference testing
- Fine-tuning on local datasets
- Secure edge inference deployments
- OEM-ready AI-in-a-box systems for repeatable customer rollouts