Computex 2026 – AI TOGETHER - Day 2 takeaways

Research evidence excerpt

Computex 2026 – AI TOGETHER - Day 2 takeaways

Taiwan Equity | Technology

Event Analysis

Figure 4: LEO satellite related products

Source: BizLink, Fubon Research

Gigabyte (2376 TT) – from server design to data center deployment

Gigabyte comprehensively showcased its depth of vertical integration in the AI server

sector, with a product line spanning from cloud data centers to the edge, providing

customers with one-stop services from L6 to L13. Gigabyte has transformed from a pure

hardware assembly OEM into a provider of complete AI infrastructure solutions with both

hardware and software integration capabilities.

In terms of servers, Gigabyte exhibited the nVidia Vera Rubin NVL72 rack and the HGX

Rubin NVL8 featuring a full liquid cooling architecture. In addition to versions equipped

with x86 CPUs, HGX/MGX will also allow customers to choose nVidia's Vera CPU in the

future. For AMD solutions, Gigabyte showcased AI servers equipped with eight MI355X

and two AMD EPYC 9005/9004 processors. As for general-purpose servers, a 6U blade

server with an inner manifold capable of supporting 20 CPUs was displayed.

To address the pain point that data center deployment often takes 1 to 2 years, Gigabyte

showcased the GADU modular container data center, integrating its direct liquid cooling

(DLC) or immersion cooling solutions, along with its self-developed POD management

platform (GPM) and the physical GAIFA (AI Verification Factory), enabling enterprises to

shorten data center deployment timelines to within one year and achieve a so-called ‘plug-

and-play’ delivery model.

In addition to server rack-level products, Gigabyte also showcased the DGX

supercomputer equipped with GB300 and the AI TOP ATOM featuring GB10,

demonstrating its deployment in on-premises AI Agent applications.