Out-of-the-Box AI Supercomputing for Large-Scale AI Deployments
At GTC 2025, NVIDIA introduced the Blackwell Ultra-powered DGX SuperPOD™, providing enterprises with ready-to-deploy AI supercomputing for agentic AI reasoning, generative AI, and large-scale inference workloads.
The new DGX GB300 and DGX B300 systems, equipped with NVIDIA Grace Blackwell Ultra Superchips, offers next-generation AI performance for model training, inference, and AI-powered automation. Built for scalability and efficiency, these rack-scale, liquid-cooled or air-cooled solutions deliver next-level computational power for AI factories.
Scalable AI Supercomputing With DGX GB300 and DGX B300
DGX GB300 System
Designed for real-time AI agent responses and advanced reasoning models, DGX GB300 features:
- 72 NVIDIA Blackwell Ultra GPUs and 36 Grace CPUs per system.
- Up to 70x more AI performance than previous-generation Hopper-based AI factories.
- 38TB of fast HBM3e memory, enabling large-scale AI workloads.
- Fifth-Generation NVLink Switch System, creating a massive shared memory space across GPUs.
- 800Gb/s accelerated networking with 72 NVIDIA ConnectX-8 SuperNICs and 18 BlueField-3 DPUs.
DGX B300 System
Built for scalable, energy-efficient AI, the DGX B300 system:
- Delivers 11x faster AI inference and 4x speedup in training compared to NVIDIA Hopper.
- Provides 2.3TB of HBM3e memory for optimized AI model execution.
- Features 8 ConnectX-8 SuperNICs and 2 BlueField-3 DPUs for high-speed AI networking.
End-to-End NVIDIA DGX Deployment by AMAX
As an NVIDIA Elite Partner, AMAX specializes in full service deployment including: racking, cabling, integration, and optimization for large-scale AI infrastructure, ensuring seamless deployment of NVIDIA DGX SuperPOD systems for AI-driven enterprises.
From power and cooling optimization to on-site installation and testing, AMAX delivers end-to-end AI infrastructure solutions designed for maximum performance and efficiency.
Visit AMAX at GTC 2025
Visit AMAX at GTC 2025, Booth #2018, from March 17-21, showcasing the latest advancements in AI and liquid cooling technologies. Attendees can experience a live pumped two-phase liquid cooling demonstration and learn how to build a secure on-prem AI chatbot with NVIDIA AI Workbench, presented by AMAX’s engineering team.