Urgent.News

What's breaking now, across thousands of outlets.

AI

An Asus homage to legendary Apple Mac Pro: GB300 DGX workstation gets two thumbs up from reviewer who calls $120,000 AI supercomputer 'the most capable thing we’ve ever put on a desk'

ASUS's ET900N G3 workstation delivers data center-level AI performance, running dozens of agents at once for roughly $120,000.

An Asus homage to legendary Apple Mac Pro: GB300 DGX workstation gets two thumbs up from reviewer who calls $120,000 AI supercomputer 'the most capable thing we’ve ever put on a desk'

ASUS has unveiled the ExpertCenter Pro ET900N G3 workstation, an impressive AI supercomputer boasting Nvidia's GB300 Grace Blackwell Ultra Superchip. This 27kg, 584 x 232 x 565 mm tower is engineered for a stationary desktop setup, lacking portability. It boasts a brushed silver exterior, dark mesh front panel, and two metal handles for easy relocation.

A sealed liquid cooling system with copper cold plates cools the Superchip, memory, and network card, while a 1,600-watt power supply minimizes energy waste. Two 2TB NVMe drives protect the OS and software, with spaces for three PCIe slots and two additional expansion bays for future upgrades. Benchmarks show it delivering 4,793 tokens per second on AI workloads, handling large models like GLM-5.2 with 215GB of expert weight data.

Reviewers compare its performance to Apple's Mac Pro, but the GB300 DGX Station stands in a class of its own. It features two 400-gigabit networking ports, enabling up to 1,496GB of coherent memory when linked with another unit, allowing it to outperform even high-end desktop computers. NVIDIA GB300 DGX Station's capabilities have been demonstrated with dozens of AIs working simultaneously, achieving over 45,000 tokens per second and scaling smoothly from a single agent to 128 without slowdowns.

Written by urgent.news from TechRadar's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at techradar.com →

More in AI

More from Sunday 27 September →