{"id":92693,"date":"2025-05-02T13:03:00","date_gmt":"2025-05-02T04:03:00","guid":{"rendered":"https:\/\/kr-dev.rebellions.ai\/?p=92693"},"modified":"2025-09-16T10:38:19","modified_gmt":"2025-09-16T01:38:19","slug":"peta-scale-soc-for-massive-ai-serving-rebel-quad","status":"publish","type":"post","link":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693","title":{"rendered":"Peta-Scale SoC for Massive AI Serving: REBEL-Quad"},"content":{"rendered":"\n

REBEL-Quad is an advanced AI SoC built on a UCIe-Advanced chiplet architecture, designed to meet the extreme compute and memory demands of serving frontier LLMs at massive scale for hyperscalers, AI data centers, and enterprises.<\/p>\n\n\n

\n
\"\"
[Peta-Scale SoC for Massive AI Serving: REBEL-Quad]<\/figcaption><\/figure>\n<\/div>\n\n\n

It features a unified hardware-software stack that maximizes utilization and delivers exceptional performance-per-watt across both compute-bound prefill and memory-bound decoding phases. The chiplet-based design enables hyper scalability without compromising latency or coherence.<\/p>\n\n\n

\n
\"\"
[Figure 1. Block diagram of REBEL-Quad with four homogeneous chiplets]<\/figcaption><\/figure>\n<\/div>\n\n\n

Solving Challenges of LLM<\/strong><\/h2>\n\n\n\n

REBEL-Quad introduces architectural solutions designed for energy-efficient AI inference at massive scale.<\/p>\n\n\n\n

\n
\n

Unified Mixed Precision Compute Engine<\/strong><\/h3>\n\n\n\n

Combines FP8\/FP16\/FP32 arithmetic in a single core, boosting compute density by 2.8\u00d7.<\/p>\n<\/div>\n\n\n\n

\n

Predictive DMA & On-Chip Mesh<\/strong><\/h3>\n\n\n\n

2.7 \u202fTB\/s effective bandwidth with software-tuned DMA, backed by a 3.3\u00d7 faster mesh fabric.<\/p>\n<\/div>\n<\/div>\n\n\n\n

REBEL-Quad\u2019s exceptional performance-per-watt, starts from its proprietary IP cores. A high-bandwidth on-chip mesh interconnect ensures seamless communication across all cores.<\/p>\n\n\n\n

\n\n\n\n
\n
\n

Holistic Synchronization<\/strong><\/h3>\n\n\n\n

Hardware-accelerated peer-to-peer and hierarchical communication enabling seamless distributed execution.<\/p>\n<\/div>\n\n\n\n

\n

Custom Die-to-Die Protocol<\/strong><\/h3>\n\n\n\n

1 TB\/s per channel with 11ns inter-chiplet latency, while preserving modularity and enabling various chiplet expansions.<\/p>\n<\/div>\n<\/div>\n\n\n\n

The software stack is tightly coupled to the underlying hardware to enable frictionless scalability to serve frontier models. Backed by a custom die-to-die protocol, REBEL-Quad reliably supports the massive inference demands for hyperscalers and AI data centers.<\/p>\n\n\n\n

Unified Mixed Precision Compute \u2028with Chiplet-Scale Scalability<\/strong><\/h2>\n\n\n\n

Traditional NPUs rely on separate arithmetic blocks for different precision formats (e.g., FP8, FP16, BF16), leading to inefficiencies in both area and dataflow scheduling. REBEL-Quad introduces a unified arithmetic engine supporting per-operand configurable precision, eliminating the need for separate functional units.<\/p>\n\n\n\n

This design enables:<\/p>\n\n\n\n

    \n
  • 2.8\u00d7 higher compute density compared to legacy implementations<\/strong><\/li>\n\n\n\n
  • Reduced instruction dependency via hardware-managed wide-issue execution<\/strong><\/li>\n<\/ul>\n\n\n\n
    <\/div>\n\n\n
    \n
    \"\"
    [Figure 2. Unified multi-\/mixed-precision arithmetic core]<\/figcaption><\/figure>\n<\/div>\n\n\n

    The wide-issue mechanism balances memory bandwidth across tensor and vector cores, ensuring simultaneous access to registers and scratchpad memory. These improvements are particularly beneficial in compute-heavy stages of LLM inference, where sustained FP8 throughput is essential. REBEL-Quad achieves 2 PFLOPS (FP8) compute within a single-node four-chiplet package, significantly enhancing performance-per-watt.<\/p>\n\n\n\n

    Predictive DMA and High-Bandwidth Memory Access<\/strong><\/h2>\n\n\n\n

    In the decoding phase, token-by-token generation is limited by KV cache memory bandwidth, which scales poorly with longer context windows. To mitigate this, REBEL-Quad implements a predictive, software-configurable DMA engine capable of:<\/p>\n\n\n\n

      \n
    • 2.7 TB\/s effective memory bandwidth<\/strong><\/li>\n\n\n\n
    • Simultaneous local and remote HBM access<\/strong><\/li>\n\n\n\n
    • Multi-path routing for bandwidth interleaving<\/strong><\/li>\n<\/ul>\n\n\n\n
      <\/div>\n\n\n\n

      This DMA is tightly integrated with REBEL-Quad\u2019s custom on-chip mesh interconnect, providing 3.3\u00d7 higher per-core bandwidth over previous architectures. The DMA engine also supports per-task QoS, minimizing latency spikes and dependency stalls across long-tail workloads.<\/p>\n\n\n\n

      Hierarchical Synchronization and Peer Communication<\/strong><\/h2>\n\n\n\n

      To sustain performance across distributed workloads, especially in models with complex attention patterns and long-range dependencies, REBEL-Quad employs a full-chip synchronization and communication mechanism.<\/p>\n\n\n\n

      Key mechanisms include:<\/p>\n\n\n\n

        \n
      • A dedicated virtual channel for control signals across the mesh network<\/strong><\/li>\n\n\n\n
      • A centralized synchronization manager that orchestrates execution flow<\/strong><\/li>\n\n\n\n
      • Hardware-accelerated peer-to-peer communication between cores, DMAs, and synchronization units<\/strong><\/li>\n<\/ul>\n\n\n\n
        <\/div>\n\n\n\n

        The hierarchical communication protocol supports both fine-grained intra-chiplet coordination and scalable inter-chiplet dependency resolution, ensuring maximum utilization across all neural cores during concurrent prefill and decoding. This architecture avoids traditional synchronization bottlenecks and minimizes software overhead, enabling compute-dense execution, high utilization rate, and consequently, maximum performance-per-watt.<\/p>\n\n\n

        \n
        \"\"
        [Figure 3. Full-chip data transfer utilizing neural cores and DMA engines]<\/figcaption><\/figure>\n<\/div>\n\n
        \n
        \"\"
        [Figure 4. Full-chip peer-to-peer and hierarchical synchronization scheme using the per-chiplet centralized sync manager]<\/figcaption><\/figure>\n<\/div>\n\n\n

        Scalable Die-to-Die Protocol for Modular Expansion<\/strong><\/h2>\n\n\n\n

        REBEL-Quad\u2019s modular design is enabled by a custom die-to-die protocol based on UCIe-Advanced, offering:<\/p>\n\n\n\n

          \n
        • 1 TB\/s bi-directional per-channel throughput and 11ns full-path inter-chiplet latency<\/strong><\/li>\n\n\n\n
        • Load-store memory semantics across dies<\/strong><\/li>\n\n\n\n
        • Future-proofed via flexible scale-up and scale-out<\/strong><\/li>\n<\/ul>\n\n\n\n
          <\/div>\n\n\n\n

          This interconnect transforms the multi-chip system into a virtually monolithic unit, while preserving modular scalability for future system expansion. Each chiplet communicates via three UCIe channels, with topology-aware die rotation ensuring horizontal mesh continuity. The protocol is paired with a robust switch network and real-time debug mechanisms to support high-reliability, error-free operation, essential for supporting massive AI serving demands of frontier models, especially for hyperscalers, AI data centers, and enterprises. Future extensibility is planned through I\/O and memory expander chiplets, enabling even broader system configurations with minimal redesign.<\/p>\n\n\n\n

          REBEL-Quad delivers the performance, efficiency, and scalability needed to serve next-generation LLMs\u2014without compromise. Its chiplet-based design enables modular upgrades and long-term adaptability, making it an ideal foundation for enterprise-scale AI systems.<\/p>\n","protected":false},"excerpt":{"rendered":"

          REBEL-Quad is an advanced AI SoC built on a UCIe-Advanced chiplet architecture, designed to meet the extreme compute and memory…<\/p>\n","protected":false},"author":8,"featured_media":92723,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[145],"tags":[],"class_list":["post-92693","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-white-papers"],"acf":[],"yoast_head":"\nPeta-Scale SoC for Massive AI Serving: REBEL-Quad - Rebellions<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693\" \/>\n<meta property=\"og:locale\" content=\"ko_KR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Peta-Scale SoC for Massive AI Serving: REBEL-Quad - Rebellions\" \/>\n<meta property=\"og:description\" content=\"REBEL-Quad is an advanced AI SoC built on a UCIe-Advanced chiplet architecture, designed to meet the extreme compute and memory...\" \/>\n<meta property=\"og:url\" content=\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693\" \/>\n<meta property=\"og:site_name\" content=\"Rebellions\" \/>\n<meta property=\"article:published_time\" content=\"2025-05-02T04:03:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2025-09-16T01:38:19+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2025\/05\/rebel-quad.png\" \/>\n\t<meta property=\"og:image:width\" content=\"1054\" \/>\n\t<meta property=\"og:image:height\" content=\"998\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/png\" \/>\n<meta name=\"author\" content=\"jiwon.kwak\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@RebellionsAI\" \/>\n<meta name=\"twitter:site\" content=\"@RebellionsAI\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#article\",\"isPartOf\":{\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693\"},\"author\":{\"name\":\"jiwon.kwak\",\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/#\/schema\/person\/1b946c50a99f04d7b7193c47b212b6c5\"},\"headline\":\"Peta-Scale SoC for Massive AI Serving: REBEL-Quad\",\"datePublished\":\"2025-05-02T04:03:00+00:00\",\"dateModified\":\"2025-09-16T01:38:19+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693\"},\"wordCount\":732,\"publisher\":{\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/#organization\"},\"image\":{\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#primaryimage\"},\"thumbnailUrl\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2025\/05\/rebel-quad.png\",\"articleSection\":[\"White Papers\"],\"inLanguage\":\"ko-KR\"},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693\",\"url\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693\",\"name\":\"Peta-Scale SoC for Massive AI Serving: REBEL-Quad - Rebellions\",\"isPartOf\":{\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#primaryimage\"},\"image\":{\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#primaryimage\"},\"thumbnailUrl\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2025\/05\/rebel-quad.png\",\"datePublished\":\"2025-05-02T04:03:00+00:00\",\"dateModified\":\"2025-09-16T01:38:19+00:00\",\"breadcrumb\":{\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#breadcrumb\"},\"inLanguage\":\"ko-KR\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"ko-KR\",\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#primaryimage\",\"url\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2025\/05\/rebel-quad.png\",\"contentUrl\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2025\/05\/rebel-quad.png\",\"width\":1054,\"height\":998},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Peta-Scale SoC for Massive AI Serving: REBEL-Quad\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/#website\",\"url\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/\",\"name\":\"\\bRebellions\",\"description\":\"Drive AI Innovation. Simple. Fast. At Scale.\",\"publisher\":{\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/#organization\"},\"alternateName\":\"\ub9ac\ubca8\ub9ac\uc628\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"ko-KR\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/#organization\",\"name\":\"Rebellions\",\"alternateName\":\"\ub9ac\ubca8\ub9ac\uc628\",\"url\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ko-KR\",\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2023\/02\/Web_Logo_Typing_05.gif\",\"contentUrl\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2023\/02\/Web_Logo_Typing_05.gif\",\"width\":1920,\"height\":1080,\"caption\":\"Rebellions\"},\"image\":{\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/x.com\/RebellionsAI\",\"https:\/\/www.youtube.com\/@Rebellions_inc\",\"https:\/\/www.linkedin.com\/company\/rebellions-ai\/\"]},{\"@type\":\"Person\",\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/#\/schema\/person\/1b946c50a99f04d7b7193c47b212b6c5\",\"name\":\"jiwon.kwak\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ko-KR\",\"@id\":\"https:\/\/dev-rbln-kr.locomotion.co.kr\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/bad9ff0489ff35bd54e60b39bd7b60aae7eb336d677d5991a83975ec819e19a4?s=96&d=blank&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/bad9ff0489ff35bd54e60b39bd7b60aae7eb336d677d5991a83975ec819e19a4?s=96&d=blank&r=g\",\"caption\":\"jiwon.kwak\"}}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Peta-Scale SoC for Massive AI Serving: REBEL-Quad - Rebellions","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693","og_locale":"ko_KR","og_type":"article","og_title":"Peta-Scale SoC for Massive AI Serving: REBEL-Quad - Rebellions","og_description":"REBEL-Quad is an advanced AI SoC built on a UCIe-Advanced chiplet architecture, designed to meet the extreme compute and memory...","og_url":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693","og_site_name":"Rebellions","article_published_time":"2025-05-02T04:03:00+00:00","article_modified_time":"2025-09-16T01:38:19+00:00","og_image":[{"width":1054,"height":998,"url":"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2025\/05\/rebel-quad.png","type":"image\/png"}],"author":"jiwon.kwak","twitter_card":"summary_large_image","twitter_creator":"@RebellionsAI","twitter_site":"@RebellionsAI","schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#article","isPartOf":{"@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693"},"author":{"name":"jiwon.kwak","@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/#\/schema\/person\/1b946c50a99f04d7b7193c47b212b6c5"},"headline":"Peta-Scale SoC for Massive AI Serving: REBEL-Quad","datePublished":"2025-05-02T04:03:00+00:00","dateModified":"2025-09-16T01:38:19+00:00","mainEntityOfPage":{"@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693"},"wordCount":732,"publisher":{"@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/#organization"},"image":{"@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#primaryimage"},"thumbnailUrl":"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2025\/05\/rebel-quad.png","articleSection":["White Papers"],"inLanguage":"ko-KR"},{"@type":"WebPage","@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693","url":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693","name":"Peta-Scale SoC for Massive AI Serving: REBEL-Quad - Rebellions","isPartOf":{"@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/#website"},"primaryImageOfPage":{"@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#primaryimage"},"image":{"@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#primaryimage"},"thumbnailUrl":"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2025\/05\/rebel-quad.png","datePublished":"2025-05-02T04:03:00+00:00","dateModified":"2025-09-16T01:38:19+00:00","breadcrumb":{"@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#breadcrumb"},"inLanguage":"ko-KR","potentialAction":[{"@type":"ReadAction","target":["https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693"]}]},{"@type":"ImageObject","inLanguage":"ko-KR","@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#primaryimage","url":"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2025\/05\/rebel-quad.png","contentUrl":"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2025\/05\/rebel-quad.png","width":1054,"height":998},{"@type":"BreadcrumbList","@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?p=92693#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/dev-rbln-kr.locomotion.co.kr"},{"@type":"ListItem","position":2,"name":"Peta-Scale SoC for Massive AI Serving: REBEL-Quad"}]},{"@type":"WebSite","@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/#website","url":"https:\/\/dev-rbln-kr.locomotion.co.kr\/","name":"\bRebellions","description":"Drive AI Innovation. Simple. Fast. At Scale.","publisher":{"@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/#organization"},"alternateName":"\ub9ac\ubca8\ub9ac\uc628","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/dev-rbln-kr.locomotion.co.kr\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"ko-KR"},{"@type":"Organization","@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/#organization","name":"Rebellions","alternateName":"\ub9ac\ubca8\ub9ac\uc628","url":"https:\/\/dev-rbln-kr.locomotion.co.kr\/","logo":{"@type":"ImageObject","inLanguage":"ko-KR","@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/#\/schema\/logo\/image\/","url":"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2023\/02\/Web_Logo_Typing_05.gif","contentUrl":"https:\/\/dev-rbln-kr.locomotion.co.kr\/wp-content\/uploads\/2023\/02\/Web_Logo_Typing_05.gif","width":1920,"height":1080,"caption":"Rebellions"},"image":{"@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/x.com\/RebellionsAI","https:\/\/www.youtube.com\/@Rebellions_inc","https:\/\/www.linkedin.com\/company\/rebellions-ai\/"]},{"@type":"Person","@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/#\/schema\/person\/1b946c50a99f04d7b7193c47b212b6c5","name":"jiwon.kwak","image":{"@type":"ImageObject","inLanguage":"ko-KR","@id":"https:\/\/dev-rbln-kr.locomotion.co.kr\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/bad9ff0489ff35bd54e60b39bd7b60aae7eb336d677d5991a83975ec819e19a4?s=96&d=blank&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/bad9ff0489ff35bd54e60b39bd7b60aae7eb336d677d5991a83975ec819e19a4?s=96&d=blank&r=g","caption":"jiwon.kwak"}}]}},"_links":{"self":[{"href":"https:\/\/dev-rbln-kr.locomotion.co.kr\/index.php?rest_route=\/wp\/v2\/posts\/92693","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/dev-rbln-kr.locomotion.co.kr\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/dev-rbln-kr.locomotion.co.kr\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/dev-rbln-kr.locomotion.co.kr\/index.php?rest_route=\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/dev-rbln-kr.locomotion.co.kr\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=92693"}],"version-history":[{"count":0,"href":"https:\/\/dev-rbln-kr.locomotion.co.kr\/index.php?rest_route=\/wp\/v2\/posts\/92693\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/dev-rbln-kr.locomotion.co.kr\/index.php?rest_route=\/wp\/v2\/media\/92723"}],"wp:attachment":[{"href":"https:\/\/dev-rbln-kr.locomotion.co.kr\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=92693"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/dev-rbln-kr.locomotion.co.kr\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=92693"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/dev-rbln-kr.locomotion.co.kr\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=92693"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}