AI Briefing
KT Cloud news
The latest AI and developer news from KT Cloud, with short summaries.
Visit website →KT Cloud Demonstrates Automated Kubernetes Add-on Upgrades and GitOps Migration Using Chkk MCP
KT Cloud automated the upgrade of three add-ons, including ArgoCD, and the transition to GitOps using Chkk MCP.
Analysis of Data Center Cooling Pump Principles, Construction Requirements, VFD Control, and AI Predictive Maintenance Technology
This article summarizes the types and principles of data center cooling pumps, construction requirements, VFD control, and AI predictive maintenance solutions.
Preparing for the 2030 Quantum Threat: SSL/TLS Migration with Hybrid ML-KEM Practical Guide
To counter HNDL attacks, PQC migration must be executed at the edge using OpenSSL 3.5 and Nginx.
Resolving AI Inference Memory Bottlenecks: Shifting Server Design to Hierarchical Memory Architectures
To resolve memory bottlenecks caused by increasing KV Cache, a hierarchical structure utilizing layers from HBM to NVMe SSD is required.
KT Cloud to unveil OpenStack internalization and Observability monitoring services at December 2 webinar
KT Cloud will share insights on platform technology internalization and the use of Observability monitoring services at a webinar on December 2.
GovWare 2026: AI Agent Prompt Injection Threats and the Need for Execution Context Security
The risk of external input traversing multiple AI agents to result in code changes must be mitigated through execution context security.
Single-VM OpenStack Sandbox: Cinder NFS Storage Reboot Verification Completed
Verified the reboot integrity of Cinder NFS storage in a single-VM OpenStack sandbox.
Kubernetes 1.34: DRA reaches GA, enabling dynamic GPU allocation and default CDI activation
In Kubernetes 1.34, DRA transitioned to GA, allowing flexible, attribute-based requests for GPUs.
Analysis of Power Density and Liquid Cooling Infrastructure Changes in North American AIDC Driven by NVIDIA GPU Evolution
From Ampere to Rubin, rising GPU TDP has increased per-rack power density to 150kW or more.
KT Cloud Aims to Secure 1GW AI Data Center Supply Capacity by 2031
The company is advancing the construction of 50MW-class AI data centers in Ansan and Cheongna.
KT Cloud shifts from public-sector DR backup focus to tiered A1–A4 recovery architectures
Following the 2025 National Information Society Agency fire, public-sector DR standards have been strengthened, requiring recovery designs tailored to A1–A4 tiers.
Principles of Data Center Fire Protection Systems for Battery Thermal Runaway Response and Service Continuity
Data center fire protection systems are availability infrastructure designed to address battery thermal runaway and ensure service continuity.
The real bottleneck of AI transformation is network infrastructure
AI training and inference traffic exceeds the limits of existing network designs, necessitating a restructuring of WAN architecture.
KT Cloud survey confirms 32.4% AI adoption in data centers, with incident response as the top challenge
Half of respondents use operational data only for post-hoc analysis, and security concerns outweighed costs in AI adoption.
Tech Week Singapore 2026: Liquid Cooling and Modular Design Are Key for AI Data Centers
Securing power and land for AI data centers and cultivating specialized talent have emerged as key challenges.
Fail-Safe Design Analysis of Latching Relays for Cooling Signal Failure in AI Data Centers
Latching relays are applied to prevent simultaneous shutdown of cooling equipment caused by PLC signal loss.
RFM requires GPUs simultaneously across three stages: training, simulation, and edge inference
RFM demands more GPUs than LLMs because data generation through simulation is essential.
0.003–0.005kWh per AI Token: Direct Liquid Cooling (DLC) and PUE 1.1 Design for 80kW per Rack
Explains the energy required for AI token generation and the principles of Direct Liquid Cooling (DLC) and PUE 1.1 design for handling 80kW power density per rack.
KubeCon China 2026: AI Infrastructure Competitiveness Lies in 'Operations,' Not Hardware
The conference showcased case studies on efficient AI infrastructure operations, ranging from GPU resource optimization to fault resilience.
Kubernetes SIG Apps Unveils Agent Sandbox Project for AI Agents
Released as v0.4.2, Agent Sandbox reduces operational complexity compared to StatefulSets and supports kernel isolation.