-12%

ASUS Ascent GX10 128GB


The ASUS Ascent GX10 128GB is a compact AI supercomputer powered by the NVIDIA GB10 Grace Blackwell Superchip, combining a 20-core CPU and integrated Blackwell GPU with 128GB LPDDR5X unified memory. Delivering up to 1 petaFLOP of AI performance in a 15 cm chassis, it’s ideal for AI development, training, and edge computing.

ASUS Ascent GX10 AI Supercomputer, NVIDIA GB10 Grace Blackwell Superchip, 128GB Coherent Unified System Memory, NVLink-C2C Technology, Integrated NVIDIA ConnectX-7 Network | Ascent-GX10

Min. Quantity – 5 Nos

Note: Below are the approximate and promotional prices. For the latest pricing and further details, please WhatsApp or call us at +91-8903657999.

195,999 223,000

The ASUS Ascent GX10 128GB is a compact desktop AI supercomputer powered by the NVIDIA GB10 Grace Blackwell Superchip, delivering up to 1 petaFLOP (1,000 TOPS) of AI performance. It features 128GB unified LPDDR5X memory, a 20-core Arm Grace CPU, and a high-performance Blackwell GPU with next-gen Tensor Cores, enabling the training and deployment of generative AI models up to 200 billion parameters. Designed for researchers and developers, the GX10 offers exceptional compute density, advanced NVLink-C2C CPU-GPU connectivity, and scalability to handle massive AI workloads directly from a desktop.

ASUS Ascent GX10, ASUS Ascent GX10 128GB, ASUS GX10 Workstation, ASUS GX10 Server, ASUS GX10 High Performance, ASUS GX10 128GB RAM, ASUS Ascent GX10 Desktop, ASUS Ascent GX10 PC, ASUS GX10 High-End Workstation, ASUS GX10 Professional Computer, ASUS GX10 for AI, ASUS GX10 Data Science

Product Name ASUS Ascent GX10
Form Factor  Ultra-compact mini AI supercomputer
Weight  ~1.6 kg
Processor (CPU)  NVIDIA GB10 Grace Arm CPU: 20-core
Graphics (GPU) Integrated NVIDIA Blackwell GPU with 5th-gen Tensor Cores and FP4 support
Memory 128 GB LPDDR5X unified system memory
Performance Up to 1 petaFLOP (1,000 TOPS) AI compute (FP4, sparsity)
Memory Bandwidth  Unified CPU+GPU memory via NVLink®-C2C, ~5× bandwidth of PCIe 5.0
Storage M.2 NVMe slots (2242/2230): PCIe 5.0 x4 (up to 4 TB) or PCIe 4.0 (1 TB)
Networking 10 Gb Ethernet (LAN) + NVIDIA ConnectX-7 NIC for 200 GbE RDMA clustering