- Dell’s Pro Max GB10 remains unavailable while Asus ships its own powerful GB10 system
- Asus Ascent GX10 offers petaflop performance in a compact desktop package
- AI developers can buy Asus’s Grace Blackwell PC now for $4100
Dell’s upcoming Pro Max GB10 AI workstation looks like a stylish piece of kit, but sadly it’s not currently available to buy.
The system, built around Nvidia’s new Grace Blackwell GB10 Superchip, is listed with a “notify me when available” status on Dell’s website, with no indication of when that might be.
If you can’t wait for it to go on sale, and you don’t mind looking elsewhere, Asus is already offering its own AI-focused desktop, the Ascent GX10.
First out of the gate
Asus’s mini beast uses the same GB10 hardware and can be ordered right now from Viperatech for $4100. The retailer says it will ship within ten days.
Both systems, and others from the likes of Acer, are designed for researchers and developers who want data center-level power in a workstation-sized package.
The GB10 chip merges CPU and GPU resources into a single unit and delivers up to one petaflop of FP4 computing performance.
It comes with 128GB of LPDDR5x unified memory and supports models with up to 200 billion parameters, a scale once limited to large server clusters.
It features an ARM v9.2-A CPU paired with Nvidia’s integrated Blackwell GPU, running on the Grace Blackwell architecture.
Storage options range from 1TB PCIe 4.0 NVMe SSDs up to 4TB PCIe 5.0 drives, providing ample space for datasets and project files.
Connectivity comes in the form of Wi-Fi 7, Bluetooth 5, and 10G Ethernet.
Ports include multiple USB 3.2 Gen 2×2 Type-C ports, one of which supports 180W power delivery, along with HDMI 2.1 for external displays.
Measuring only 150mm square and 51mm tall, and weighing 1.48kg, the GX10 is tiny but mighty and comes with advanced thermal management to keep things cool and running smoothly even under heavy loads.
It also supports dual-system stacking through Nvidia’s ConnectX-7 networking and NVLink-C2C interface, allowing local compute expansion.
For developers eager to experiment with large AI models, Asus offers one of the few ready-to-ship systems based on Nvidia’s latest architecture.
Follow TechRadar on Google News and add us as a preferred source to get our expert news, reviews, and opinion in your feeds. Make sure to click the Follow button!
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form, and get regular updates from us on WhatsApp too.
