Xiilab Unveils GPU Optimization Solution ‘AstraGo’ at HPE Partner Seminar

Reduce TCO of AI Infrastructure by Maximizing GPU Resource Utilization
HPE Server Interworking Technology Optimizes Large-Scale AI Infrastructure

[Daily Korea Reporter Min Byung-moo]
Xiilab, a Vision AI company, announced on the 9th that it had introduced AstraGo, a GPU optimization solution, and presented real data-center application cases at the “KIES-HPE Partners Day” seminar.

The event, hosted by Korea Information Engineering, a domestic distributor of HPE, brought together major domestic partners and IT experts to share the latest trends in the AI market, server roadmaps, and GPU utilization strategies.

Xiilab’s AstraGo is a GPU cluster management solution optimized for AI workloads. It automatically partitions and allocates GPU resources essential for AI infrastructure, improving project speed while reducing costs. Its flexible Kubernetes-based architecture allows easy expansion and management of thousands of nodes and containers.

During the presentation, Xiilab highlighted its interworking technology with HPE servers and introduced methods for integrating and managing large-scale AI infrastructure.
AstraGo integrates software and hardware on a single platform through HPE OneView monitoring and a Redfish-standard interface, greatly enhancing operational convenience and efficiency. It also features auto-scaling that can expand thousands of nodes with a few clicks and real-time failure-prediction alerts, significantly reducing operational complexity.

Following this year’s distribution agreement with Korea Information Engineering, Xiilab has been expanding the supply of AstraGo solutions for data centers and research institutes. The company continues to increase AI infrastructure adoption across industries such as semiconductors and electronics.

“AstraGo is the most practical solution for simultaneously achieving economic efficiency and performance in AI infrastructure,” said Yoon Se-hyuk, CEO of Xiilab. “Through close collaboration with HPE, we will continue to advance technologies that support global hyperscale AI infrastructure and lower the barriers for companies adopting AI.”

https://daily.hankooki.com/news/articleView.html?idxno=1240832