System builders like ASUS and Dell customize NVIDIA DGX systems for different users by tailoring the hardware and software configurations to meet specific needs. Here's a detailed overview of how they might approach customization:
Hardware Customization
1. Processor and GPU Selection: Both ASUS and Dell can customize the DGX systems by selecting appropriate NVIDIA GPUs and CPUs that align with the user's performance requirements. For instance, the DGX Station uses the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip, which includes a high-performance NVIDIA Blackwell Ultra GPU and a Grace CPU[6][7].
2. Memory and Storage Options: Users can choose from various memory and storage configurations to optimize their systems for specific tasks. For example, the DGX Station offers 784GB of coherent memory space, which can be adjusted based on user needs[6].
3. Networking and Connectivity: Customization may also involve selecting the right networking components, such as the NVIDIA ConnectX-8 SuperNIC, to ensure optimal connectivity and data transfer speeds for AI workloads[6].
Software Customization
1. NVIDIA AI Enterprise Software: Both ASUS and Dell can integrate the NVIDIA AI Enterprise software platform, which provides access to NVIDIA NIM microservices. This allows users to manage and optimize AI workflows efficiently[6].
2. Model Customization Tools: For AI model customization, Dell has collaborated with NVIDIA to create scalable and modular solutions that support techniques like fine-tuning, instruction tuning, and transfer learning[1]. ASUS, while not directly mentioned in AI model customization, can leverage similar NVIDIA technologies to offer customized AI solutions.
3. User Interface and Control: Customization might also involve tailoring the user interface and control software to meet specific user preferences. For example, ASUS is known for its user-friendly software interfaces in other products, such as audio cards[2][5].
Integration and Support
1. Integration with Other Systems: Both ASUS and Dell can help integrate DGX systems with existing infrastructure, ensuring seamless compatibility and workflow optimization. This might involve configuring systems to work with cloud services or other AI platforms[6].
2. Professional Services: Dell offers professional services for generative AI, including advisory, implementation, adoption, scale, and managed services. These services can help users optimize their DGX systems for specific AI applications[1].
In summary, system builders like ASUS and Dell customize DGX systems by offering a range of hardware and software configurations tailored to meet the specific needs of different users, from AI developers to researchers and enterprises. This customization ensures that users can maximize the performance and utility of their DGX systems for various AI applications.
Citations:
[1] https://www.delltechnologies.com/asset/en-us/solutions/infrastructure-solutions/technical-support/gen-ai-model-customization-design-guide.pdf
[2] https://dlcdnet.asus.com/pub/ASUS/Audio_Card/Xonar_DX/e3859b_Xonar_DX.pdf
[3] https://www.pcmag.com/how-to/bios-basics-how-to-configure-your-pcs-firmware-for-first-use
[4] https://www.asus.com/support/faq/1038387/
[5] https://dlsvr04.asus.com.cn/pub/ASUS/Audio_Card/Xonar_DSX/E11099_Xonar_DSX_UM_V5_web.pdf
[6] https://www.investing.com/news/company-news/nvidia-launches-personal-ai-supercomputers-for-desktops-93CH-3934971
[7] https://investor.nvidia.com/news/press-release-details/2025/NVIDIA-Announces-DGX-Spark-and-DGX-Station-Personal-AI-Computers/default.aspx
[8] https://www.tech-critter.com/unboxing-review-asus-xonar-dgx-pci-e-51/