https://www.youngju.dev/blog 천천히 올바르게. AI Researcher & DevOps Engineer Youngju's tech blog. GPU/CUDA, LLM, MLOps, Kubernetes AI workloads, distributed training, and data engineering. ko fjvbn2003@gmail.com (Youngju Kim) fjvbn2003@gmail.com (Youngju Kim) Tue, 16 Jun 2026 00:00:00 GMT https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-interconnect-nvlink-ualink.en https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-interconnect-nvlink-ualink.en In large-scale AI training and inference, the real bottleneck is not compute but communication. This article walks through the principles and practice of AI interconnect — the scale-up domains built by NVLink and NVSwitch, rack-scale systems like GB200 NVL72, and open alternatives such as UALink and Ultra Ethernet. Tue, 16 Jun 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) gpu-cudanvlinknvswitchualinkinterconnectdistributed-trainingai-hardware https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-interconnect-nvlink-ualink.ja https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-interconnect-nvlink-ualink.ja 大規模なAI学習と推論における本当のボトルネックは、演算ではなく通信です。NVLinkとNVSwitchが作るスケールアップドメイン、GB200 NVL72のようなラックスケールシステム、そしてUALinkやUltra Ethernetといったオープンな代替まで、AIインターコネクトの原理と実務を整理します。 Tue, 16 Jun 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) gpu-cudanvlinknvswitchualinkinterconnectdistributed-trainingai-hardware https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-interconnect-nvlink-ualink https://www.youngju.dev/blog/gpu-cuda/2026-06-16-ai-interconnect-nvlink-ualink 대규모 AI 학습과 추론의 진짜 병목은 연산이 아니라 통신입니다. NVLink와 NVSwitch가 만드는 스케일업 도메인, GB200 NVL72 같은 랙 스케일 시스템, 그리고 UALink와 Ultra Ethernet 같은 개방형 대안까지, AI 인터커넥트의 원리와 실무를 정리합니다. Tue, 16 Jun 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) gpu-cudanvlinknvswitchualinkinterconnectdistributed-trainingai-hardware