Giga Computing Launches NVIDIA HGX B300-Powered Flagship Server

September 18, 2025 – Giga Computing, a pioneer in generative AI and high-performance computing (HPC) servers, today announced the launch of its new flagship server, GIGABYTE G894-SD3-AAX7. This 8U dual-processor platform supports NVIDIA HGX B300 and is designed to accelerate AI training, inference, and complex HPC workloads, delivering unparalleled performance and scalability.

Leading-Edge AI Reasoning and High-Performance Computing Platform

The G894-SD3-AAX7 combines the latest hardware innovations in a compact 8U, air-cooled chassis, maximizing both system performance and expansion capabilities:

●         NVIDIA HGX B300 Support:  The 8-GPU HGX B300 platform integrates 8x NVIDIA Blackwell Ultra GPUs with NVIDIA NVLink and NVLink Switch to deliver 1.8TBps GPU-GPU bandwidth and 14.4 TB/s total bandwidth for scalable AI performance.

●         Powerful Processor Configuration: Dual Intel® Xeon® 6700/6500 series processors with 32x DDR5 DIMM slots, delivering robust compute for next-generation AI training and HPC applications.

●         Ultra-High-Speed Networking: Equipped with 8x 800Gb/s OSFP InfiniBand or  16x 400Gb/s Ethernet ports via NVIDIA ConnectX-8 SuperNICs, enabling high-throughput performance for large-scale data center workloads.

●         Flexible Expansion: Supports NVIDIA BlueField-3 DPUs and 4x PCIe Gen5 x16 slots, providing maximum data throughput to storage and other clusters plus room for future system expansion.

●         Data Center-Grade Reliability: Features 12x 3000W 80 PLUS Titanium level redundant power supplies, ensuring consistent performance and enterprise-level stability.

Unmatched AI Computing and Memory Scale

The NVIDIA HGX B300 platform delivers transformative AI compute capabilities and memory capacity:

●        7x AI performance over the Hopper platform

●        Over 2 TB HBM3E memory

●        Integrated NVIDIA ConnectX-8 SuperNICs, providing 8x 800Gb/s OSFP InfiniBand or  16x 400Gb/s Ethernet connections

By leveraging NVIDIA Blackwell architecture, NVIDIA HGX B300 accelerates not only traditional AI training and inference but also unlocks large language model reasoning capabilities, supporting the most demanding scenarios from AI agents and complex decision-making to real-time multimedia generation, setting a new standard for data center performance.

"The G894-SD3-AAX7 is designed for emerging AI computing cloud service providers like neoclouds, while also meeting the innovation needs of multiple industries,” said Vicent Wang, Sales VP at Giga Computing. “With the power of NVIDIA HGX B300, enterprises can accelerate generative AI, reasoning services, and HPC workloads, enabling breakthroughs across healthcare, manufacturing, finance, and research."

Built for Next-Generation Workloads

The G894-SD3-AAX7 drives innovative applications across industries, including healthcare and scientific research, financial modeling, automation and smart manufacturing, and generative AI. Advanced cooling and optimized power efficiency ensure stable and efficient operation under the most demanding AI and HPC workloads, helping customers reduce total cost of ownership (TCO) while delivering high-performance, sustainable, and reliable systems.

With the launch of the G894-SD3-AAX7, Giga Computing also expands its GIGAPOD high-performance AI cluster solutions portfolio. The addition of the B300 flagship server not only provides more powerful compute node options for GIGAPOD but also enhances overall performance and scalability for generative AI, scientific simulation, and enterprise AI factory deployments. By tightly integrating with the NVIDIA technology ecosystem, GIGAPOD continues to provide enterprises with end-to-end solutions — from single servers to ultra-large-scale clusters — accelerating the adoption and deployment of AI innovations.