Trusted Partners
Please Note: Product images are for illustrative purposes only and may differ from the actual product.
Hand-pick every part with live stock, compatibility, and instant pricing.
Try the new builderHand-pick every part with live stock, compatibility, and instant pricing.
Try the new builderASUS Ascent GX10 Mini PC
Powered by the NVIDIA GB10 Grace Blackwell Superchip alongside the comprehensive NVIDIA AI software stack, the ASUS Ascent GX10 delivers a complete ecosystem for developing and deploying artificial intelligence. Its space-saving footprint ensures straightforward integration, making high-performance AI accessible for professionals seeking peak capability. Equipped with cutting-edge AI utilities and NVIDIA® ConnectX®-7 technology, this miniature server elevates your computational power to support custom technological solutions.
Powered by the NVIDIA GB10 Grace Blackwell Superchip alongside the comprehensive NVIDIA AI software stack, the ASUS Ascent GX10 delivers a complete ecosystem for developing and deploying artificial intelligence. Its space-saving footprint ensures straightforward integration, making high-performance AI accessible for professionals seeking peak capability. Equipped with cutting-edge AI utilities and NVIDIA® ConnectX®-7 technology, this miniature server elevates your computational power to support custom technological solutions.
Explore the compact engineering and advanced capabilities of the ASUS Ascent GX10 AI Supercomputer in this featured video presentation highlighting its architecture and full-stack development potential.
Explore the compact engineering and advanced capabilities of the ASUS Ascent GX10 AI Supercomputer in this featured video presentation highlighting its architecture and full-stack development potential.
Measuring a space-saving 150 x 150 x 51mm, this device yields up to 1 petaFLOP of AI processing power utilizing FP4 and incorporates 64GB or 128GB of LPDDR5x Coherent Unified System Memory to effortlessly drive intensive model creation, testing, and inferencing tasks.
Measuring a space-saving 150 x 150 x 51mm, this device yields up to 1 petaFLOP of AI processing power utilizing FP4 and incorporates 64GB or 128GB of LPDDR5x Coherent Unified System Memory to effortlessly drive intensive model creation, testing, and inferencing tasks.
Built around the NVIDIA GB10 Grace Blackwell Superchip featuring fifth-generation Tensor Cores, a high-performance 20-core Arm CPU for streamlined data handling, and NVLink-C2C technology delivering five times the bandwidth of PCIe 5.0 for a unified CPU and GPU memory framework.
Built around the NVIDIA GB10 Grace Blackwell Superchip featuring fifth-generation Tensor Cores, a high-performance 20-core Arm CPU for streamlined data handling, and NVLink-C2C technology delivering five times the bandwidth of PCIe 5.0 for a unified CPU and GPU memory framework.
Engineered to prototype and fine-tune reasoning models up to 100 billion parameters on the 64GB configuration or scale up to 200 billion parameters on the 128GB build, with built-in NVIDIA ConnectX-7 networking allowing clusters of up to four systems to handle frontier-scale models.
Engineered to prototype and fine-tune reasoning models up to 100 billion parameters on the 64GB configuration or scale up to 200 billion parameters on the 128GB build, with built-in NVIDIA ConnectX-7 networking allowing clusters of up to four systems to handle frontier-scale models.
Equipped with NVIDIA ConnectX-7 technology to offer ultra-fast data transmission speeds, built-in hardware acceleration for IPsec, TLS, and MACsec data encryption without CPU overhead, and IEEE 1588v2 PTP support for microsecond-level synchronization across distributed clusters.
Equipped with NVIDIA ConnectX-7 technology to offer ultra-fast data transmission speeds, built-in hardware acceleration for IPsec, TLS, and MACsec data encryption without CPU overhead, and IEEE 1588v2 PTP support for microsecond-level synchronization across distributed clusters.
Empowers continuous 24/7 execution of agentic frameworks like OpenClaw locally on onboard hardware, utilizing NVIDIA NemoClaw and NVIDIA OpenShell to enforce policy-based security guardrails, lower API dependence, and maintain strict on-premises data privacy.
Empowers continuous 24/7 execution of agentic frameworks like OpenClaw locally on onboard hardware, utilizing NVIDIA NemoClaw and NVIDIA OpenShell to enforce policy-based security guardrails, lower API dependence, and maintain strict on-premises data privacy.
Features an advanced dual-fan cooling layout with 7-level speed regulation, engineered to deliver 1.6 times more efficient thermal coverage than typical compact workstations to maintain peak hardware performance under heavy workloads.
Features an advanced dual-fan cooling layout with 7-level speed regulation, engineered to deliver 1.6 times more efficient thermal coverage than typical compact workstations to maintain peak hardware performance under heavy workloads.
Allows users to stack units together, combining two 64GB systems into 128GB of memory and up to 2 petaflops of compute power, or expanding up to four units for 512GB of pooled memory and 4 petaFLOPS of performance using NVIDIA Sync.
Allows users to stack units together, combining two 64GB systems into 128GB of memory and up to 2 petaflops of compute power, or expanding up to four units for 512GB of pooled memory and 4 petaFLOPS of performance using NVIDIA Sync.
Features a comprehensive selection of peripheral interfaces including a Kensington Lock Slot, one USB 3.2 Gen 2x2 Type-C port supporting 20Gbps transfer speeds with DisplayPort 2.1 alternate mode and 180W EPR PD3.1 power delivery input, three additional USB 3.2 Gen 2x2 Type-C ports supporting 20Gbps with DisplayPort 2.1 alternate mode, an HDMI 2.1b port, a 10 GbE LAN port, and an integrated NVIDIA ConnectX-7 NIC.
Features a comprehensive selection of peripheral interfaces including a Kensington Lock Slot, one USB 3.2 Gen 2x2 Type-C port supporting 20Gbps transfer speeds with DisplayPort 2.1 alternate mode and 180W EPR PD3.1 power delivery input, three additional USB 3.2 Gen 2x2 Type-C ports supporting 20Gbps with DisplayPort 2.1 alternate mode, an HDMI 2.1b port, a 10 GbE LAN port, and an integrated NVIDIA ConnectX-7 NIC.
Utilize the NVIDIA AI platform software architecture to seamlessly transition models from local desktop environments to DGX Cloud or alternative accelerated cloud and data center infrastructures with minimal code modifications. This platform serves as a cost-effective experimentation workspace that preserves valuable cluster compute resources for training and deploying production-scale models.
Utilize the NVIDIA AI platform software architecture to seamlessly transition models from local desktop environments to DGX Cloud or alternative accelerated cloud and data center infrastructures with minimal code modifications. This platform serves as a cost-effective experimentation workspace that preserves valuable cluster compute resources for training and deploying production-scale models.
Engineered to support early-stage prototyping workflows for developing advanced computational models and applications.
Engineered to support early-stage prototyping workflows for developing advanced computational models and applications.
Provides robust processing capabilities tailored for model fine-tuning and executing real-time inference tasks efficiently.
Provides robust processing capabilities tailored for model fine-tuning and executing real-time inference tasks efficiently.
Optimized to handle demanding data science tasks, enabling researchers and developers to analyze complex datasets locally.
Optimized to handle demanding data science tasks, enabling researchers and developers to analyze complex datasets locally.
Everything you need to know about this product
The ASUS Ascent GX10 is accelerated by the NVIDIA GB10 Grace Blackwell Superchip.
It comes integrated with the NVIDIA AI software stack to provide a full-stack solution for artificial intelligence development and deployment.
The system features built-in NVIDIA® ConnectX®-7 technology to enhance your AI capabilities.
It features an ultra-small, compact design that facilitates seamless integration and deployment while delivering powerful AI performance.
It is engineered for innovators and developers who demand excellence and require a robust, small-scale server for unique AI solutions.
Yes, it provides a comprehensive full-stack solution for both AI development and deployment through hardware and software integration.
AI-Generated Content: These FAQs are generated by an automated AI system and may contain inaccuracies or omissions. Evetech makes no warranties, express or implied, regarding their completeness or reliability. For definitive information, please refer to the manufacturer's official specifications.