NVIDIA DGX Spark Adds 64GB Configuration for Local AI Development

New lower-memory SKU targets developers running large models without cloud dependency.

Editorial illustration: A round lidded glass jar sits on a rough stone surface against a dark brown background, its interior filled with branching blue-white lightning network patterns spreading across a dark rocky substrate topped with green moss and small purple flowers.

AI-generated editorial illustration.

NVIDIA announced a 64GB unified memory configuration of DGX Spark, available October 23 from Acer, ASUS, Dell, Gigabyte, HP and MSI starting at $4,999. The system retains the GB10 Grace Blackwell Superchip, DGX OS and full NVIDIA AI software stack, and supports models up to 100 billion parameters on device. Two units can cluster via NVIDIA Sync Cluster Assistant using a QSFP cable, pooling memory to 128GB and supporting up to 200 billion parameters. NVIDIA says two clustered 64GB systems delivered up to 1.7x performance versus a single unit in internal Qwen 3.8 27B testing. Supported runtimes include Ollama, vLLM and PyTorch.

Source: NVIDIA News. Original announcement: October 02, 2026.

Keep up with IT Solution Journal

Practical guidance and new coverage. You can withdraw your permission at any time.

Read our privacy and data-use policy.