ASUS is presenting its full-stack AI infrastructure at Ai4 2026 in Las Vegas. The company is showing systems that cover data-center training through to deskside inference at Booth 1348. The lineup includes the ESC8000A-E13P server based on NVIDIA MGX architecture. It supports up to eight dual-slot NVIDIA H200 or RTX PRO Blackwell GPUs. Another rack option is the RS720A-E13-RS8G DC-MHS server. For smaller setups, ASUS is displaying the Ascent GX10 AI Supercomputer based on NVIDIA DGX Spark. The deskside highlight is the ExpertCenter Pro ET900N G3, built on NVIDIA DGX Station architecture. It delivers up to 20 petaFLOPS of AI performance with 748 GB of coherent memory and can run models up to one trillion parameters locally. ASUS is a global OEM partner for both DGX Spark and DGX Station. The company describes the approach as trusted AI with total flexibility, letting enterprises move from rack-scale training to local development without changing vendors.
ASUS is presenting its full-stack AI infrastructure at Ai4 2026 in Las Vegas. The company is showing systems that cover data-center training through to deskside inference at Booth 1348. The lineup includes the ESC8000A-E13P server based on NVIDIA MGX architecture. It supports up to eight dual-slot NVIDIA H200 or RTX PRO Blackwell GPUs. Another rack option is the RS720A-E13-RS8G DC-MHS server. For smaller setups, ASUS is displaying the Ascent GX10 AI Supercomputer based on NVIDIA DGX Spark. The deskside highlight is the ExpertCenter Pro ET900N G3, built on NVIDIA DGX Station architecture. It delivers up to 20 petaFLOPS of AI performance with 748 GB of coherent memory and can run models up to one trillion parameters locally. ASUS is a global OEM partner for both DGX Spark and DGX Station. The company describes the approach as trusted AI with total flexibility, letting enterprises move from rack-scale training to local development without changing vendors.