ASUS unveils a creator-focused workstation lineup headlined by the ProArt GR1X, powered by NVIDIA’s RTX Spark and a 20-core Grace CPU paired with a Blackwell RTX GPU, offering up to 1 petaFLOP of FP4 AI performance and up to 128GB of unified memory for local AI work and 3D workloads.

Storage solutions include UF920-E3-RS24 AI-native STX-based storage, OJ340A-RS60 object storage, and VS320D-RS26N storage to meet capacity, availability, and data-management needs from design validation to production-scale AI workloads.

Edge systems span the ESC8000-E12P for enterprise inference and vision AI with RTX PRO 6000/4500 Blackwell GPUs and the compact PE3000N using Jetson Thor T5000 for real-time inference, sensor fusion, and autonomous control in industrial settings.

Rubin systems leverage NVIDIA Vera Rubin and HGX Rubin architectures, featuring AI POD XA VR721-E3 for rack-scale deployments and XA NR1I-E12LR/NR1I-E12L for training and inference, with liquid or hybrid cooling to boost efficiency.

The rollout coincides with NVIDIA AI Day Singapore 2026, underscoring ASUS’s broad NVIDIA-powered AI portfolio across enterprise, desktop, and edge installations.

ASUS showcased a full spectrum of NVIDIA-powered AI systems at NVIDIA AI Day Singapore 2026, spanning rack-scale data center gear, desktop workstations, edge devices, and storage for training, inference, agentic AI, and industrial automation.

The XA P2N-E2 2U MGX platform targets agentic AI workloads with dual Vera CPUs and up to two dual-slot GPUs to enable scalable AI reasoning, data processing, and orchestration.

Desktop options include the ExpertCenter Pro ET900N G3, a DGX Station-based system with GB300 Grace Blackwell Ultra Desktop Superchip, delivering up to 20 PFLOPS and 748GB memory, alongside the Ascent GX10 (DGX Spark) with GB10 Grace Blackwell Superchip, up to 1 petaFLOP and 128GB unified memory for local development and inference.