.jpg)
The Era of 1 Million GPUs: Bottlenecks Are 'Wiring,' Not Computation… CPOs Are Changing the Data Center Landscape
AI data centers face data movement as a bottleneck rather than GPU computation. CPO (Co-Packaged Optics), which packages switch ASICs and optical engi…
2026.07.31 by 배종인 기자
Silicon Labs Completes Validation of 200-Node Thread-Based Matter Network
Silicon Labs has successfully completed the construction and operation of a Thread-based Matter network at a scale of 200 nodes. In validation using O…
2026.06.25 by 명세환 기자
Synology Launches Enterprise NVMe Storage 'PAS7700'
Synology has launched the 'PAS7700', its first active-active all-flash NVMe enterprise storage system. The PAS7700 targets AI, virtualization, and lar…
2026.05.26 by 명세환 기자
Infineon, No. 1 Automotive Semiconductor Manufacturer, Accelerates Transition to 'RISC-V', Core Pillar of SDV
Infineon Technologies, the global leader in automotive semiconductors, is positioning RISC-V (Reduced Instruction Set Computer - Five) as a core pilla…
2026.04.20 by 배종인 기자
Intel and Google Accelerate Construction of Large-Scale AI Systems Combining CPUs and IPUs
Intel and Google are pursuing a strategic partnership that strengthens the roles of CPUs and custom-built IPUs (Infrastructure Processing Units) to ef…
2026.04.10 by 배종인 기자
.jpg)
Congatec Extends Performance of AMD Ryzen AI-Based COM Modules… Expanding Edge AI Design Flexibility
Congatec has enhanced the performance and scalability of the 'conga-TCRP1', a COM Express module based on AMD Ryzen AI Embedded P100 series. The modul…
2026.03.17 by 명세환 기자
Nota selected as a leading digital company in Daejeon for three consecutive years, presenting a model for commercializing local AI.
The competitiveness of the regional AI industry is determined not by research and development results, but by the practical application of technology …
2026.03.11 by 배종인 기자
NVIDIA's Blackwell Platform Reduces AI Inference Costs by Up to 10x Per Token
NVIDIA is accelerating AI adoption across industries by reducing the cost per token for AI inference services by up to 10 times through its next-gener…
2026.02.20 by 배종인 기자