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) Sat, 16 May 2026 00:00:00 GMT https://www.youngju.dev/blog/culture/2026-05-16-edge-ai-tinyml-2026-litert-executorch-edge-impulse-jetson-coral-hailo-sipeed-k230-llama-cpp-deep-dive.en https://www.youngju.dev/blog/culture/2026-05-16-edge-ai-tinyml-2026-litert-executorch-edge-impulse-jetson-coral-hailo-sipeed-k230-llama-cpp-deep-dive.en A full-stack map of the 2026 Edge AI / TinyML ecosystem — the dual standard formed after TFLite Micro was rebranded as LiteRT and ExecuTorch reached GA, the TinyML cloud workflow created by Edge Impulse, the accelerator showdown between NVIDIA Jetson Orin Nano/NX/Thor/AGX and Coral Dev Board, the rise of Israeli Hailo-15/8 and Chinese Sipeed K230 (RISC-V + NPU), the SBC standardization of Rockchip RK3588, MaixPy / Arduino Nano 33 BLE Sense / Seeed Wio AI, MicroPython for ML, and on-phone MLC LLM / llama.cpp / Whisper.cpp / GGUF — covering Phi-3/3.5/4, Gemma 2/3, Llama 3.2 1B/3B — and how IoT / mobile / automotive engineers should pick tools in the age of Always-on AI, including case studies from Korea (ETRI / Samsung / LG) and Japan (Sony AI / NTT). Sat, 16 May 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) edge-aitinymltflite-microlitertexecutorchedge-impulsejetson-orincoral-dev-boardhailosipeed-k230rockchip-rk3588maixpymicropython-mlonnx-runtime-mobilecore-mltensorrtapache-tvmmlc-llmwhisper-cppllama-cppggufphi-3phi-4gemmallama-3-22026deep-diveenglish https://www.youngju.dev/blog/culture/2026-05-16-edge-ai-tinyml-2026-litert-executorch-edge-impulse-jetson-coral-hailo-sipeed-k230-llama-cpp-deep-dive.ja https://www.youngju.dev/blog/culture/2026-05-16-edge-ai-tinyml-2026-litert-executorch-edge-impulse-jetson-coral-hailo-sipeed-k230-llama-cpp-deep-dive.ja 2026 年のエッジ AI / TinyML エコシステムをフルスタックで俯瞰 — TFLite Micro が LiteRT へリブランドされ ExecuTorch が GA となって形成された二大標準、Edge Impulse が築いた TinyML クラウドワークフロー、NVIDIA Jetson Orin Nano/NX/Thor/AGX と Coral Dev Board のアクセラレータ対決、イスラエルの Hailo-15/8 と中国の Sipeed K230(RISC-V+NPU)の台頭、Rockchip RK3588 の SBC 標準化、MaixPy / Arduino Nano 33 BLE Sense / Seeed Wio AI、ML 向け MicroPython、そしてスマホ上で動く MLC LLM / llama.cpp / Whisper.cpp / GGUF — Phi-3/3.5/4、Gemma 2/3、Llama 3.2 1B/3B まで — Always-on AI の時代に IoT / モバイル / 自動車エンジニアがどう道具を選ぶべきかを、韓国(ETRI / Samsung / LG)と日本(Sony AI / NTT)の事例とともにまとめます。 Sat, 16 May 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) edge-aitinymltflite-microlitertexecutorchedge-impulsejetson-orincoral-dev-boardhailosipeed-k230rockchip-rk3588maixpymicropython-mlonnx-runtime-mobilecore-mltensorrtapache-tvmmlc-llmwhisper-cppllama-cppggufphi-3phi-4gemmallama-3-22026deep-dive日本語 https://www.youngju.dev/blog/culture/2026-05-16-edge-ai-tinyml-2026-litert-executorch-edge-impulse-jetson-coral-hailo-sipeed-k230-llama-cpp-deep-dive https://www.youngju.dev/blog/culture/2026-05-16-edge-ai-tinyml-2026-litert-executorch-edge-impulse-jetson-coral-hailo-sipeed-k230-llama-cpp-deep-dive 2026년 엣지 AI / TinyML 생태계의 풀스택 지도 — TFLite Micro 가 LiteRT 로 리브랜드된 후 ExecuTorch 가 GA 가 되며 형성된 듀얼 표준, Edge Impulse 가 만든 TinyML 클라우드 워크플로, NVIDIA Jetson Orin Nano/NX/Thor/AGX 와 Coral Dev Board 의 가속기 격돌, 이스라엘 Hailo-15/8 과 중국 Sipeed K230(RISC-V+NPU) 의 부상, Rockchip RK3588 의 SBC 표준화, MaixPy/Arduino Nano 33 BLE Sense/Seeed Wio AI, MicroPython for ML, 그리고 폰 위에서 도는 MLC LLM / llama.cpp / Whisper.cpp / GGUF — Phi-3/3.5/4, Gemma 2/3, Llama 3.2 1B/3B 까지 — Always-on AI 의 시대에 IoT / 모바일 / 자동차 엔지니어가 어떤 도구를 어떻게 골라야 하는지 한국(ETRI/삼성/LG)과 일본(Sony AI/NTT) 사례까지 모두 정리합니다. Sat, 16 May 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) edge-aitinymltflite-microlitertexecutorchedge-impulsejetson-orincoral-dev-boardhailosipeed-k230rockchip-rk3588maixpymicropython-mlonnx-runtime-mobilecore-mltensorrtapache-tvmmlc-llmwhisper-cppllama-cppggufphi-3phi-4gemmallama-3-22026deep-dive https://www.youngju.dev/blog/culture/2026-05-16-webgpu-wgsl-ecosystem-2026-chrome-safari-firefox-wgpu-naga-tint-three-js-babylon-webllm-transformers-deep-dive.en https://www.youngju.dev/blog/culture/2026-05-16-webgpu-wgsl-ecosystem-2026-chrome-safari-firefox-wgpu-naga-tint-three-js-babylon-webllm-transformers-deep-dive.en In 2026 WebGPU has finally gone GA on all three major browsers: Chrome 113 (May 2023), Safari 18 (Sept 2024) and Firefox 141 (Aug 2025). The W3C spec has crossed from Working Draft into Candidate Recommendation and WGSL has settled in as the successor to GLSL. wgpu (Rust, Mozilla/gfx-rs) provides the native backend; naga and Google Dawn's Tint handle WGSL compilation. Three.js r163+ WebGPU renderer, Babylon.js 5.x, Filament Web, regl, gpu.js and PlayCanvas cover graphics, while WebLLM (MLC LLM), Transformers.js (Xenova), ONNX Runtime WebGPU and Web Stable Diffusion enable in-browser ML inference. Compute shaders have matured and Storage Texture, Subgroups and mesh shaders are now formal 2026 proposals. With deployment cases from Toss data visualisation, Kakao Games, pixiv and CyberAgent, this deep dive paints the full 2026 WebGPU landscape through Korean, Japanese and Western lenses. Sat, 16 May 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) webgpuwgslweb-graphicswgpu-rsnagatintdawnthree-jsbabylon-jsfilament-webreglgpu-jsplaycanvaswebllmmlc-llmtransformers-jsonnx-runtimeweb-stable-diffusioncompute-shadersray-tracing2026deep-diveenglish https://www.youngju.dev/blog/culture/2026-05-16-webgpu-wgsl-ecosystem-2026-chrome-safari-firefox-wgpu-naga-tint-three-js-babylon-webllm-transformers-deep-dive.ja https://www.youngju.dev/blog/culture/2026-05-16-webgpu-wgsl-ecosystem-2026-chrome-safari-firefox-wgpu-naga-tint-three-js-babylon-webllm-transformers-deep-dive.ja 2026 年、WebGPU はついに 3 大ブラウザすべてで GA となった。Chrome 113 (2023 年 5 月)、Safari 18 (2024 年 9 月)、Firefox 141 (2025 年 8 月)。W3C 仕様は Working Draft から Candidate Recommendation に進み、WGSL シェーディング言語は GLSL の後継としてその地位を固めた。wgpu (Rust, Mozilla/gfx-rs) はネイティブバックエンドを、naga と Google Dawn の Tint は WGSL コンパイルを担う。Three.js r163+ の WebGPU レンダラ、Babylon.js 5.x、Filament Web、regl、gpu.js、PlayCanvas がグラフィックスを担当し、WebLLM (MLC LLM)、Transformers.js (Xenova)、ONNX Runtime WebGPU、Web Stable Diffusion がブラウザでの ML 推論を可能にする。コンピュートシェーダは成熟し、Storage Texture と Subgroups、そして mesh shaders は 2026 年の正式提案段階にある。トス、カカオゲームズ、ピクシブ、CyberAgent の導入事例まで含め、WebGPU 2026 の風景を韓国・日本・西欧の視点で総合的にまとめる。 Sat, 16 May 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) webgpuwgslweb-graphicswgpu-rsnagatintdawnthree-jsbabylon-jsfilament-webreglgpu-jsplaycanvaswebllmmlc-llmtransformers-jsonnx-runtimeweb-stable-diffusioncompute-shadersray-tracing2026deep-dive日本語 https://www.youngju.dev/blog/culture/2026-05-16-webgpu-wgsl-ecosystem-2026-chrome-safari-firefox-wgpu-naga-tint-three-js-babylon-webllm-transformers-deep-dive https://www.youngju.dev/blog/culture/2026-05-16-webgpu-wgsl-ecosystem-2026-chrome-safari-firefox-wgpu-naga-tint-three-js-babylon-webllm-transformers-deep-dive 2026년, WebGPU는 마침내 3대 브라우저(Chrome 2023.5, Safari 2024.9, Firefox 2025.8)에서 모두 GA가 되었다. W3C 사양은 Working Draft에서 Candidate Recommendation으로 넘어가고, WGSL 셰이딩 언어는 GLSL의 후계자로 자리를 굳혔다. wgpu(Rust/Mozilla/gfx-rs)는 네이티브 백엔드를, naga와 Google Dawn의 Tint는 WGSL 컴파일을 책임진다. Three.js r163+ WebGPU 렌더러와 Babylon.js 5.x, Filament Web, regl, gpu.js, PlayCanvas는 그래픽을, WebLLM(MLC LLM)과 Transformers.js(Xenova), ONNX Runtime WebGPU, Web Stable Diffusion은 브라우저 ML 추론을 가능하게 한다. 컴퓨트 셰이더는 성숙했고, Storage Texture와 Subgroups, 그리고 mesh shaders는 2026년 제안 단계에 있다. 토스 데이터 시각화, 카카오게임즈, 픽시브, CyberAgent의 도입 사례까지 — WebGPU 2026의 풍경을 한국·일본·서구의 시각에서 종합 정리한다. Sat, 16 May 2026 00:00:00 GMT fjvbn2003@gmail.com (Youngju Kim) webgpuwgslweb-graphicswgpu-rsnagatintdawnthree-jsbabylon-jsfilament-webreglgpu-jsplaycanvaswebllmmlc-llmtransformers-jsonnx-runtimeweb-stable-diffusioncompute-shadersray-tracing2026deep-dive