ASUS Launches 64GB Ascent GX10 to Expand AI Development Access

ASUS recently unveiled a new 64GB configuration of its ASUS Ascent GX10, a compact AI supercomputer designed to facilitate local agentic AI development. This launch, announced on October 2, 2026, is significant as it broadens access to advanced AI capabilities for developers, researchers, and data scientists. By providing a more affordable entry point into the NVIDIA DGX Spark platform, ASUS aims to democratize the development of autonomous AI agents, which are increasingly becoming integral to business operations.

The ASUS Ascent GX10 64GB is built on the NVIDIA GB10 Grace Blackwell Superchip architecture, delivering up to 1 petaflop of AI performance. This configuration is specifically tailored for running today’s most efficient open-source models, particularly those with 30 to 35 billion parameters. The introduction of this model allows a wider audience to engage in hands-on AI experimentation, moving beyond mere curiosity to practical application. The existing 128GB model continues to cater to teams with more demanding workloads, ensuring that ASUS can serve a diverse range of customer needs.

As businesses increasingly adopt agentic AI—systems that autonomously execute complex tasks—the demand for local infrastructure is rising. Companies are seeking to mitigate the costs associated with cloud inference and to maintain control over sensitive data by keeping operations in-house. The 64GB configuration addresses these concerns by enabling developers to run powerful AI models locally, thus avoiding the pitfalls of public API reliance. This shift toward local AI infrastructure is poised to reshape how organizations manage and deploy AI technologies.

The design of the ASUS Ascent GX10 64GB emphasizes scalability. It allows users to cluster multiple units, effectively pooling memory and computational power. This capability is crucial for organizations anticipating growth in their AI workloads. By connecting two units, users can achieve a combined 128GB of memory and up to 2 petaflops of compute, enhancing throughput significantly. This modular approach not only offers flexibility but also provides a clear upgrade path for teams as their needs evolve.

ASUS’s strategy, encapsulated in its "Ubiquitous AI. Incredible Possibilities." initiative, indicates a broader commitment to making advanced AI accessible to a wider range of developers. The 64GB model is positioned to attract a larger segment of the market, including those who may have previously found such technology out of reach. This move could intensify competition among AI hardware providers, as companies like NVIDIA and others in the space continue to innovate.

The thermal efficiency of the ASUS Ascent GX10 64GB is another noteworthy aspect. With a sophisticated cooling system designed for sustained high performance, the unit can handle demanding AI tasks over extended periods. This capability is particularly relevant for businesses that require reliable and consistent performance for long-running AI applications.

Looking ahead, the introduction of the ASUS Ascent GX10 64GB is likely to accelerate the adoption of local AI solutions across various industries. As more organizations recognize the benefits of running AI workloads on-premises—such as enhanced data security and reduced costs—demand for such systems will likely increase. This trend may lead to further innovations in local AI infrastructure, compelling competitors to enhance their offerings and expand their market reach. The evolution of AI technology is set to transform operational paradigms, and companies that adapt quickly will be better positioned to capitalize on the opportunities presented by this shift.