Washington, D.C., Oct. 28, 2025 — ASUS will showcase a flexible, end-to-end AI Factory at NVIDIA GTC Washington, D.C. (Walter E. Washington Convention Center; ASUS booth #738). Spanning extreme edge, deskside, and rack-scale deployments built on NVIDIA Blackwell architecture, the lineup is designed to fit a wide range of operational environments—from research labs and control rooms to on-prem data centers—with options that support disciplined lifecycle management and on-site deployment needs. Shipments of ASUS XA GB721-E2  built on the NVIDIA GB300 NVL72 platform and XA NB3I-E12 (based on NVIDIA HGX B300 system) began in September 2025.

1) A Flexible AI Factory: From Pilot to Production, On-Site or At the Edge

The ASUS AI Factory combines hardware, optimized software stacks, and services to streamline AI adoption across generative AI, VLM/LLM inference, computer vision, and predictive analytics. Organizations can standardize on common NVIDIA frameworks while choosing the right footprint—edge appliances, deskside systems, or rack-scale PODs—to meet data locality, availability, and serviceability preferences. High-density nodes such as XA NB3I-E12 (leveraging the NVIDIA HGX B300 platform), includes  8 NVIDIA ConnectX-8 InfiniBand SuperNICs, 5 PCIe® expansion slots, 32 DIMM, and 10 NVMe, helping teams scale performance without sacrificing maintainability.

2) Systems on Display: Field-Ready Edge, Deskside Acceleration, and Rack-Scale Throughput

Rack-Scale & GPU Servers

• ASUS XA GB721-E2  (Grace Blackwell Ultra system): A reference blueprint for multi-node training and large-scale inference using NVIDIA networking and software.

• RS720A-E13-RS8G (2U GPU server): Versatile 2U platform for training and high-throughput inference with thoughtful service access.

• ESC8000A-E13X (NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs and NVIDIA ConnectX-8 SuperNIC): Multi-GPU acceleration for simulation, visualization, and AI pipelines with 400G-class networking options.

Deskside & Workstations

• ASUS ASCENT GX10: Compact desktop AI supercomputer for developers, analysts delivering 1 petaFLOP of AI performance and 128GB memory for 200B model fine-tuning.

• ExpertCenter Pro ET900N G3: The ASUS ExpertCenter Pro ET900N G3 is part of a new class of deskside AI supercomputers. Designed from the ground up to build and run AI, this revolutionary system unleashes the NVIDIA® GB300 Grace Blackwell Ultra Desktop Superchip and up to a massive 784GB of large coherent memory.

Extreme Edge & Robotics

• ASUS IoT PE3000N — Compact edge-AI platform built on NVIDIA Jetson Thor T5000, delivering up to 2,070 FP4 TFLOPS via a Blackwell-class GPU, a 14-core Arm® CPU, and 128GB LPDDR5X. Its scalable, modular I/O supports PoE, GMSL, CAN, QSFP28 and more, while 12–60V DC with ignition enables vehicle and field-robot power. PTP/PPS supports precise sensor time sync; TPM 2.0 and ruggedized connectors aid reliability. Options include up to 4×25GbE and 16 GMSL cameras for high-bandwidth sensor fusion—ideal for inspection, mobility, and smart-infrastructure use cases.

Thermals & Serviceability

• XA NB3I-E12 (B300 air-cooled): Balanced thermals and straightforward maintenance for mainstream deployments where uptime and accessibility matter.

Join the ASUS 2025 GTC Washington, D.C. Theater Talk

• Don’t miss the 15-minute ASUS Infrastructure for Every Scale—from Edge to Trillion-Token AI session at Expo Hall Stage on October 29 from 4:00 PM - 4:15 PM EDT. During this insightful presentation we will share how ASUS helps customers build future-ready AI data centers. Learn how our servers, rack-scale ASUS AI PODs with NVIDIA GB200/GB300 NVL72, and high-serviceability designs address diverse AI workloads and deployment challenges.