[Reprint] Data Alliance Selected for Ministry of SMEs and Startups' 'Scale-up TIPS' Program… Launches 2 Billion Won R&D Project Over 3 Years
Secured 2 billion KRW in government R&D funding over 3 years… Developed gCube-based distributed GPU inference technology

Data Alliance has been selected for a 'Scale-up TIPS' R&D project organized by the Ministry of SMEs and Startups.
Scale-up TIPS is a program in which the government provides linked R&D funding to promising technology companies that have been invested in and recommended by private investment firms after verifying their growth potential. With this selection, Data Alliance has secured a total of 2 billion won in government R&D support funds for 36 months, from July 2026 to June 2029.
The development goal of this project is the 'Heterogeneous Distributed GPU for Asynchronous Large Language Model (LLM) Inference' technology. This technology combines GPUs with different specifications into a single resource to distribute and process LLM inference tasks, enabling the configuration of a stable inference environment without the need for a large number of high-performance GPUs.
Data Alliance plans to enhance its platform by applying this technology to its core service, 'gcube'. gcube is a DePIN (Distributed Physical Infrastructure Network)-based shared GPU cloud platform that connects scattered idle GPU resources and supplies them to users; through this R&D, the company intends to significantly improve resource utilization efficiency in distributed environments.
Data Alliance plans to successfully complete the three-year project to scale up the gcube platform and secure differentiated technological competitiveness in the rapidly growing AI infrastructure market.

![[WIS 2026] Innovative Companies Present New Technologies & Products (Data Alliance Wins Grand Prize)… A Catalyst for Investment](https://static.wixstatic.com/media/f1fac8_b811deaf39d54716bfd14b0cc92bea2c~mv2.png/v1/fill/w_980,h_766,al_c,q_90,usm_0.66_1.00_0.01,enc_avif,quality_auto/f1fac8_b811deaf39d54716bfd14b0cc92bea2c~mv2.png)

Comments