Giga Computing Shows Comprehensive Data Center Portfolio and Infrastructure at NVIDIA GTC 2026; Featuring New NVIDIA Rubin Platforms

Mar 17, 2026

March 16, 2026 - Giga Computing, a subsidiary of GIGABYTE and a leader in accelerated computing and infrastructure solutions, today announced new enterprise AI solutions that support NVIDIA Vera CPU and NVIDIA Vera Rubin platform, well as a new AI factory in Taiwan. The GIGABYTE booth at NVIDIA GTC shows scalable AI solutions that not only focus on performance and efficiency but also incorporate the software and infrastructure needed to build AI factories and other large GPU clusters. The GIGABYTE booth staff is ready to introduce the latest hardware and software for success in deploying accelerated computing.

Overview of The Must See at GIGABYTE Booth #1413:

Personal AI Supercomputers

With products spanning all segments of the supercomputer space, Giga Computing showcases professional desktop and deskside solutions that are ideal for AI development and accelerating AI training and inference workloads. These solutions are being used by data scientists and researchers in research institutions, government agencies, and enterprises.

GIGABTYE AI TOP ATOM is a compact solution powered by NVIDIA GB10 Grace Blackwell Superchip that is capable of up to 1 petaFLOPS of performance and has 128GB of unified memory.
GIGABYTE W775-V10-L01 is a deskside supercomputer that is built to develop and run frontier AI locally, providing up to 20 petaFLOPS of AI FP4 performance and capable of running up to 1 trillion parameter models powered by the NVIDIA GB300 Grace Blackwell Ultra Desktop Superchip with 748GB of coherent memory. This system also supports GPUs such as NVIDIA RTX PRO 6000 Blackwell Workstation Edition GPU and NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition GPU.

GIGABYTE and NVIDIA are working together on NVIDIA NemoClaw — an open source stack that simplifies running OpenClaw always-on assistants, more safely, with a single command. As part of the NVIDIA Agent Toolkit, it installs the NVIDIA OpenShell runtime — a secure environment for running autonomous agents, and open-source models like NVIDIA Nemotron.

As desktop AI compute becomes increasingly capable, NVIDIA NemoClaw enables a new class of always-on AI agents that run directly on local hardware. Use cases include personal research assistants, workflow automation agents, and private coding assistants. NemoClaw simplifies deployment of these intelligent agents for individuals and small teams, reducing reliance on enterprise scale cloud infrastructure, keeping data private and allowing full control over AI workloads. GIGABYTE's personal AI supercomputers provide the desktop compute foundation to support the growing shift towards agentic AI.

Rack-scale AI Supercomputer

NVIDIA Vera Rubin NVL72 platform is a rack-scale AI Supercomputer that is built for demanding large language models and succeeds by unifying technologies. A single rack unifies 36 NVIDIA Vera CPUs and 72 NVIDIA Rubin GPUs and is interconnected with NVIDIA NVLink 6 for exascale performance while running highly efficient with liquid cooling for CPUs, GPU, storage, NVIDIA ConnectX-9 SuperNIC, and more.
In addition, attendees will see the new evolution of compute and switch trays for this powerful AI supercomputer.

Liquid-cooled and Air-cooled Clusters

GIGAPOD is GIGABYTE’s turnkey, rack-scale AI & HPC data center solution that clusters multiple GPU servers into a high-performance, scalable compute ecosystem. It integrates cutting-edge accelerators, networking, and GIGABYTE POD Manager (GPM) software for unified monitoring, orchestration, and deployment of AI workloads across supercomputing infrastructure. In collaboration with nVent, GIGABYTE hardware and software is paired with powerful liquid cooling technologies like the nVent CX121 CDU (at the GIGABYTE booth) that is modularized for variable uses and infrastructure.

Incorporated Into GIGAPOD and at the GIGABYTE Booth:

GIGABYTE POD Manager (GPM), a powerful infrastructure and workflow management platform that centralizes monitoring, orchestration, and automation across AI and HPC data centers. GPM streamlines deployment, resource utilization, and real-time insights, boosting efficiency for modern enterprise workloads.
GIGABYTE G4L4-SD3 is a powerful, liquid-cooled AI inference server. Supporting Intel Xeon CPUs paired with NVIDIA HGX B300 system, , this server is compute dense in a 4U design.
GIGABYTE G2L4-SD4 proves performance can increase while continuing to push the limits of compute density. Powered by NVIDIA Vera Rubin platform, NVIDIA HGX Rubin NVL8 platform supercharges training and inference for AI and HPC, enabling enterprises to break through performance and efficiency bottlenecks and deliver intelligence at scale. and is built for agentic AI and reasoning models. Notably, this new 2U GPU server incorporates an OCP busbar design.
GIGABYTE XN24-VC0 was the server selected by the RIKEN Center for Computational Science to advance scientific AI and quantum computing. Based on NVIDIA MGX modular architecture, it is a liquid-cooled solution for NVIDIA GB200 NVL4 system.
GIGABYTE XLS4-SX2 is an NVIDIA RTX PRO Server with liquid cooling technology powered by NVIDIA RTX PRO 6000 Blackwell Server Edition GPUs, NVIDIA ConnectX-8 SuperNIC switch, Intel Xeon CPUs, NVIDIA BlueField-3 DPU, and system memory. Designed for demanding AI and graphics workloads, this server is built for AI factories and enterprise deployments.
GIGABYTE XL44-SX2 is the air-cooled version of the liquid-cooled XLS4. As a NVIDIA RTX PRO server, it was the first server to come to market with this hardware configuration mentioned above. The SuperNIC board supports 800Gb/s bandwidth and PCIe Gen6 connectivity to deliver high throughput and better scale-out capabilities.
GIGABYTE TO86-SD1 follows OCP ORV3 architecture and is ideal for scientific HPC and is great for double-precision workloads and compute-bound tasks with FP32 throughput. The system integrates NVIDIA Blackwell GPUs with Intel Xeon CPUs.

New AI Factory in Taiwan

To power the era of AI reasoning, Giga Computing is establishing the Giga Computing AI Factory Accelerator (GAIFA) in Taiwan. This project will become a centralized AI computing hub, where AI factories are built at scale and optimized for efficiency and high-performance. Combining engineering expertise and infrastructure capabilities, GAIFA delivers a full-stack solution powered by NVIDIA AI infrastructure (ex. NVIDIA Blackwell Ultra GPU), NVIDIA Quantum-X800 InfiniBand and Spectrum-X Ethernet networking, and NVIDIA AI Enterprise with GIGABYTE’s management software. The Taiwan-based facility will accelerate testing, validation, and deployment timelines, enabling faster and more seamless AI factory implementation for customers worldwide.

Other notable NVIDIA technologies at the GIGABYTE Booth

NVIDIA Vera CPUs integrated into motherboards for NVIDIA MGX for reinforcement learning, agentic AI, analytics and cloud as well as and NVIDIA HGX designs for AI and HPC workloads.
NVIDIA ConnectX-9 SuperNIC that delivers 1.6 Tb/s throughput per GPU
NVIDIA BlueField-4 DPU for secure multi-tenant networking, rapid data access, and more
NVIDIA Quantum-X800 Q3450 CPO Switch is liquid-cooled and supports 144 800Gb/s ports
NVIDIA RTX PRO 4500 Blackwell Server Edition GPU delivers breakthrough performance for demanding data processing, AI, video, and visual computing workloads in a single-slot, power-efficient design.

With a portfolio spanning personal AI supercomputers, rack-scale systems, and integrated liquid & air-cooled clusters, Giga Computing is determined in its commitment to scalable AI infrastructure. By combining NVIDIA Vera CPUs and Rubin GPUs with advanced networking, cooling, and management software, Giga Computing enables enterprises, researchers, and cloud providers to accelerate innovation with confidence. Visit Booth #1413 at NVIDIA GTC 2026 to see how Giga Computing is powering AI from edge to cloud at scale.

For queries or more information, please contact sales.