GIGABYTE adds 64 GB AI TOP ATOM for local desktop AI development
GIGABYTE has made a 64 GB Unified Memory version of its AI TOP ATOM desktop system available, adding a lower-capacity option alongside the existing 128 GB configuration. Both systems use the NVIDIA DGX Spark platform and the same compact chassis for local AI development.
- GIGABYTE's 64 GB AI TOP ATOM is now available.
- The 64 GB and 128 GB versions use the NVIDIA DGX Spark platform.
- The 64 GB model offers 1 TB or 2 TB M.2 NVMe storage options.
- Up to four systems can be connected for supported distributed AI workloads.
- Regional availability, prices and configurations may vary.
A second memory configuration for AI TOP ATOM
GIGABYTE said on November 11, 2026, that the 64 GB version of the AI TOP ATOM is now available in addition to the 128 GB model. The two configurations are based on the NVIDIA DGX Spark platform and share a compact desktop design.
The company is positioning the choice around different model sizes and workloads. AI developers, machine learning engineers, researchers, teachers and AI enthusiasts can select the Unified Memory capacity that fits their work. The 64 GB system can be configured with an M.2 NVMe SSD using PCIe 4.0, with a storage capacity of either 1 TB or 2 TB.
Designed for local AI workloads
The AI TOP ATOM is designed to provide dedicated AI computing resources in a compact desktop system. GIGABYTE says it supports local model inference, prototype development, data analysis and application validation in offices, laboratories and educational institutions.
Users can run models and process project data within their own IT environment. This gives the system a role in development and testing workflows where the work is performed locally, rather than relying on an external environment. The source does not provide performance figures or specify which models are supported.
For supported distributed AI workloads, users can connect up to four AI TOP ATOM systems into a cluster. The systems use integrated NVIDIA ConnectX-7 networking and NVIDIA Sync's Cluster Assistant. Depending on the application and software configuration, a cluster can provide additional memory and computing resources.
Software supports inference and RAG workflows
The hardware is paired with the NVIDIA CUDA accelerated AI software ecosystem and GIGABYTE AI TOP Utility. According to GIGABYTE, the utility supports workflows including downloading supported models, inference and retrieval-augmented generation, or RAG.
These tools allow developers to try supported open models, test AI assistants and create knowledge-based question and answer applications using their own documents. GIGABYTE presents the functions as support for developing, testing and validating AI applications in a local environment.
What the new option means for buyers
The 64 GB model expands the AI TOP ATOM range without changing the platform or compact chassis used by the 128 GB version. It gives prospective users another memory configuration to consider when matching the system to their model sizes and workloads.
GIGABYTE cautions that product availability, sales channels, prices and storage configurations may vary by region. The company directs customers to the product page for the AI TOP ATOM series for further information.
