As more devices built around NVIDIA's GB10 chip come to market, partners may be wondering whether they represent a genuine opportunity for customers looking to develop, test and deploy AI solutions. With the ability to run large language models (LLMs) locally, these compact systems could offer a practical alternative to relying on data centre infrastructure or cloud resources during the development phase.
► Move AI modelling and testing closer to the edge.
► Gain flexibility for AI development without major infrastructure investment.
Ultra-small form factor systems powered by the NVIDIA GB10 Grace Blackwell Superchip, such as the ASUS Ascent GX10, are now becoming widely available. These devices allow organisations to run LLMs locally, supporting models with up to 200 billion parameters on a single unit and more than double that when two devices are linked together. Alongside ASUS, vendors including Dell, HP and Lenovo are introducing GB10-based systems, with laptops expected to follow.
With prices starting at around £3,700, these devices represent a significant but potentially accessible investment for organisations looking to accelerate AI initiatives. Designed as power-efficient platforms for enterprise AI workloads, they offer high-performance capabilities in a desktop-sized form factor. The ASUS Ascent GX10, for example, comes with 128GB of memory as standard.
Rather than investing in high-end servers or consuming cloud resources for every stage of development, organisations can use GB10 devices to build, test, and refine AI models closer to where they will be used. For many businesses, this provides a more flexible and cost-effective environment for experimentation, development, and proof-of-concept work.
That could prove particularly valuable as organisations look to turn AI ambitions into measurable outcomes. According to a recent report from Info-Tech Research, common barriers include rapidly changing AI technologies, difficulties demonstrating business value, unclear project objectives, conflicting stakeholder priorities, and resistance resulting from previous project failures.
Having access to a dedicated platform for testing, refining and retesting models without additional infrastructure investment may help organisations address some of these challenges more effectively. For partners, understanding where GB10 devices fit within a customer's AI strategy could open up conversations around AI readiness, development environments, and project execution.
TD SYNNEX's Destination AI programme provides partners with access to the training, tools and expertise needed to build AI knowledge, strengthen advisory capabilities, and grow their AI practice.
Destination AI
Destination AI from TD SYNNEX is our comprehensive end-to-end programme designed to help you and your customers get ahead of the AI curve.
