Nebula Launches the “Nebula Compute Cluster” Powered by NVIDIA H200 GPUs — Delivering Exceptional Computing Performance and Reliability for the Next Generation of AI Development    

Nebula Inc. (“the Company”) is pleased to announce the launch of the “Nebula Compute Cluster,” a high-performance computing cluster powered by the latest NVIDIA H200 Tensor Core GPUs, starting in May 2026. The service is designed to support the rapidly growing demands of advanced AI development and large language model (LLM) training.

■ Background and Purpose

With the rapid advancement of generative AI technologies, development teams are increasingly required to train models with significantly larger parameter sizes. As a result, shortages in computing resources, power efficiency concerns, and rising deployment costs have become major challenges for businesses.    
To address these issues, Nebula has integrated cutting-edge hardware architecture with its extensive infrastructure operation expertise. Through this service, we aim to democratize access to computing resources and accelerate AI development for organizations of all sizes, from startups to enterprise companies.

■ Key Features of the “Nebula Compute Cluster”

1. Exceptional Memory Bandwidth and Computing Performance    
Equipped with NVIDIA H200 GPUs, the cluster offers significantly enhanced memory capacity and bandwidth compared to the previous H100 generation. This enables faster LLM inference speeds and shorter training times for complex datasets.

2. High Availability with Integrated On-Site Support    
Beyond providing cloud-based computing resources, Nebula’s dedicated on-site support team continuously monitors the physical infrastructure layer 24/7. In the event of hardware failures, engineers can respond immediately on-site, minimizing downtime and ensuring stable operations.

3. Optimized Cost Efficiency    
The service adopts a rental-based model with no large upfront investment required. Clients can flexibly scale computing resources up or down according to project demands, reducing fixed asset risks and optimizing R&D expenditures.

■ Future Outlook

Going forward, Nebula will expand beyond computing resource services by providing optimization support for software stacks that maximize computational efficiency, as well as developing next-generation data centers with enhanced cooling technologies.    
Guided by our vision of “freeing technological progress from infrastructure limitations,” Nebula will continue to deliver innovative solutions as a computing partner supporting the foundation of a sustainable AI-driven society.