[Award] "gcube" Wins CES 2026 Innovation Award in the "AI" Category… Setting a New Standard for AI Cloud with Distributed GPU Infrastructure

Data Alliance's GPU (graphics processing unit) cloud platform "gcube" has won an Innovation Award (CES Innovation Awards) in the AI category at CES 2026, the world's largest IT and consumer-electronics exhibition.

The CES Innovation Award is a global technology award presented annually by the U.S. Consumer Technology Association (CTA) to the most innovative technologies and services. This award signifies that a distributed GPU infrastructure technology built in-house by a Korean company has been officially recognized as competitive on the international stage.

GPU resources — essential for AI model development and service operation — have long acted as a barrier to entry for many players, including startups, research institutions, and content-creation platforms, due to cost burdens and the difficulty of securing infrastructure.

The conventional cloud model, which concentrates resources in large-scale data centers, has featured an industrial imbalance in which a small number of large operators hold the advantage in both pricing and supply.

gcube drew attention for breaking away from this structure and proposing an AI infrastructure model based on a "distributed network" that connects idle GPU resources scattered around the world.

Data Alliance, which develops and operates gcube, consolidates previously unused GPU equipment — companies' idle servers, high-performance desktop GPUs sitting idle across the country, and personal gaming rigs — into a single cloud environment. The pooled resources are made available for AI developers to use immediately through a standardized container execution environment and a distributed GPU scheduler. This allows AI developers to carry out large-scale model training and inference at up to 90% lower cost than conventional cloud services, while GPU providers can earn revenue simply by sharing their resources, without any separate technical management.

This structure goes beyond mere cost optimization to achieve environmental sustainability. Large-scale data-center expansion drives enormous power consumption and equipment-disposal problems; gcube reduces this environmental burden by reusing resources that already exist.

gcube's innovation is also evident in how it operates technically. Binding GPU resources distributed across multiple locations into a single compute pool requires complex control technologies, including performance-variance adjustment, network-latency minimization, and automatic fault detection and failover. gcube is designed to combine its own container-orchestration engine with a Web3 settlement system, transparently distributing contribution-based rewards to resource providers while allowing users to maintain a stable compute environment. The CES judging panel rated this highly.

gcube has already proven its effectiveness in real-world service settings. On the supply side, small, mid-sized, and large IT-infrastructure subsidiaries across the country participate; on the demand side, AI development academies, university research labs, AI startups, and generative-content creation platforms use gcube to perform high-performance computing tasks such as large-scale model training, image and video generation, and data processing.

The company previously won a special "Global AI Impact Award" at "Try Everything 2025," a global startup festival hosted by the Seoul Metropolitan Government, and with this CES award it has now demonstrated both its technological innovation and its potential for social application.

"With this CES award as a springboard, we plan to strengthen the building of global partnerships centered on North America and Southeast Asia, and to expand our distributed GPU network on a global scale," said CEO Kwangbeom Lee of Data Alliance.

Reporter Jisun Kim, ET News (river@etnews.com) Source: https://www.etnews.com/20251114000287