GIGABYTE expands desktop AI options with 64GB AI TOP ATOM

GIGABYTE has made the 64GB unified-memory version of its AI TOP ATOM available, expanding the desktop AI system’s existing 128GB configuration. Both versions are based on the NVIDIA DGX Spark platform and use the same compact hardware design.
- GIGABYTE has made the 64GB AI TOP ATOM available.
- The system joins an existing 128GB configuration.
- The 64GB version offers 1TB and 2TB PCIe 4.0 SSD options.
- Users can group up to four units with NVIDIA Sync.
- AI TOP Utility supports inference, model downloads and RAG workflows.
A second memory configuration for desktop AI
GIGABYTE said the 64GB AI TOP ATOM is intended to give developers more choice when matching system capacity to the size of their models and workloads. The company’s existing 128GB configuration remains available, with both systems aimed at AI developers, machine-learning engineers, researchers, educators and AI enthusiasts.
The 64GB version also supports 1TB and 2TB M.2 NVMe PCIe 4.0 SSD options. According to GIGABYTE, those storage configurations allow users to select capacity based on their individual needs. Availability, sales channels, pricing and storage configurations may vary by region.
Designed for local AI development
GIGABYTE positions the AI TOP ATOM as a compact desktop system for AI development on site and at customer locations. The company says it provides dedicated AI computing for local model inference, prototyping, data analysis and application validation in offices, laboratories and educational environments.
Processing data locally can give users greater control over project data and development resources, according to GIGABYTE. This local approach is intended for teams that want to develop, refine and validate applications in a dedicated environment rather than relying exclusively on remote computing resources.
Software tools and multi-unit scaling
The AI TOP ATOM combines NVIDIA’s CUDA-accelerated AI software ecosystem with GIGABYTE’s AI TOP Utility. The utility supports model downloading, inference and retrieval-augmented generation, or RAG. GIGABYTE says developers can use these functions to explore open models, test AI assistants and build knowledge-based question-and-answer applications with their own documents.
For workloads that need additional computing resources, the system includes NVIDIA ConnectX-7 networking. With NVIDIA Sync, developers can group up to four AI TOP ATOM units, creating a larger pool of memory and more processing capacity, according to GIGABYTE.
The 64GB unified-memory AI TOP ATOM is available alongside the 128GB version. GIGABYTE directs users to its AI TOP ATOM product page for more information.


