● Covered by 2 sources · 2 reportsMedium impact2 neutral
NVIDIA Introduces 64GB DGX Spark for Local AI Development
Key points
- NVIDIA released a 64GB DGX Spark configuration.
- It features the GB10 SoC and unified memory.
- The system supports local AI development and models up to 100B parameters.
- Available from partners like Acer, ASUS, Dell, HP, MSI in October.
- Two units can cluster using NVIDIA Sync Cluster Assistant.
New DGX Spark Configuration
NVIDIA has introduced a new 64GB unified memory configuration for its DGX Spark platform. This version is designed to provide a more accessible option for local AI development, featuring the GB10 SoC and the NVIDIA CUDA-accelerated AI software stack.
The new SKU will be available from manufacturing partners including Acer, ASUS, Dell, Gigabyte, HP, and MSI, starting in October. It comes with DGX OS and the NVIDIA AI software stack pre-installed.
Addressing Evolving AI Needs
The introduction of the 64GB DGX Spark addresses the changing landscape of AI models and silicon supply. While earlier assumptions suggested a need for 128GB of RAM or more for intelligent models, current efficient AI models, such as Qwen 3.8 27B, can operate effectively within 32GB of RAM.
This reduced memory requirement, combined with rising RAM prices, makes a 64GB configuration a more practical and affordable entry point for developers who do not require the full 128GB capacity.
Capabilities and Scalability
The 64GB DGX Spark is capable of running local AI agents on-device, supporting models with up to 100 billion parameters. It offers a platform for experimenting with models and personal data without relying on cloud instances for every task.
For increased workloads, two DGX Spark units can be clustered together using the NVIDIA Sync Cluster Assistant, without requiring additional setup.
Personal AI Supercomputing
The DGX Spark system integrates NVIDIA Grace Blackwell compute, unified memory, NVIDIA ConnectX-7 networking, and a CUDA-accelerated AI software stack. It functions as a complete local AI platform suitable for agents, inference, fine-tuning, data science, and edge development.
This compact system provides a personal AI supercomputing environment, allowing developers and researchers to work with AI models locally and privately.
✨ This summary was generated by AI from the outlets' reporting listed below. It is not independently verified and may contain errors — check the original sources. How BrevFeed works →
The daily brief
One email each morning: the day's tech stories, clustered across outlets and summarized. No account needed.
One email a day. Unsubscribe in one click, any time.
Today's brief
Spend a few minutes, get the whole day. Every topic's top stories in one hands-free rundown — listen, watch, or read the transcript.
▶ Play today's brief
New every morning, and the back catalogue is archived by date.
How outlets covered it
NVIDIA is releasing a 64GB unified memory configuration for its DGX Spark system through manufacturing partners. This new configuration provides a more accessible option for local AI development, supporting up to 100-billion-parameter models and offering scalability through clustering.
Nvidia released a 64GB version of its DGX Spark platform, featuring the GB10 SoC, to offer a more affordable option for local AI development. This new configuration addresses the rising cost of RAM and the reduced memory requirements of current efficient AI models, making the platform more accessible. The 64GB DGX Spark models will be available from multiple vendors starting October 23.