{"schema_version":"onlylabs.public_signal.v1","title":"Makora Repo: makora-ai/flash-moe","description":"Makora repo signal with public source context, captured evidence pages, related signals, and category-scoped analysis context.","url":"https://onlylabs.fyi/signals/373a0500-eaea-4c6b-ac9f-ed1ebfa9355e","json_url":"https://onlylabs.fyi/signals/373a0500-eaea-4c6b-ac9f-ed1ebfa9355e/signal.json","generated_at":"2026-06-11T04:08:24.39503+00:00","org":{"slug":"makora","name":"Makora","category":"neocloud","category_label":"Neocloud","dossier_url":"https://onlylabs.fyi/labs/makora","dossier_json_url":"https://onlylabs.fyi/labs/makora/dossier.json"},"related_urls":{"signal":"https://onlylabs.fyi/signals/373a0500-eaea-4c6b-ac9f-ed1ebfa9355e","signal_json":"https://onlylabs.fyi/signals/373a0500-eaea-4c6b-ac9f-ed1ebfa9355e/signal.json","source":"https://github.com/makora-ai/flash-moe","lab_dossier":"https://onlylabs.fyi/labs/makora","lab_dossier_json":"https://onlylabs.fyi/labs/makora/dossier.json","analysis":"https://onlylabs.fyi/analysis/makora","analysis_json":"https://onlylabs.fyi/analysis/makora/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/makora/evidence.json","category":"https://onlylabs.fyi/neoclouds","category_json":"https://onlylabs.fyi/neoclouds.json","category_feed":"https://onlylabs.fyi/neoclouds/feed.xml","category_signals_json":"https://onlylabs.fyi/signals.json?category=neocloud","topic":null,"topic_signals_json":null,"topic_feed":null,"data_business":null},"answer_pack":{"answer":"Makora published makora-ai/flash-moe (C++). This repository signal exposes tooling, eval, infrastructure, or model-adjacent work before it may appear in a launch post. High-signal details: repo makora-ai/flash-moe · language C++ · Low stars, trivial repo. onlylabs links this event to 1 captured evidence page and 5 related repo signals.","signal_desk":"repos","source_context":{"source_url":"https://github.com/makora-ai/flash-moe","source_host":"github.com","occurred_at":"2025-09-18T16:37:05+00:00","first_seen_at":"2026-06-05T22:32:19.171838+00:00","date_source":"source","context":"C++"},"context_markers":[{"label":"Lab","value":"Makora","source":"signal"},{"label":"Signal desk","value":"repos","source":"signal"},{"label":"Source host","value":"github.com","source":"source"},{"label":"Repository","value":"makora-ai/flash-moe","source":"source"},{"label":"Language","value":"C++","source":"source"},{"label":"Stars","value":"1","source":"traction"},{"label":"Notability","value":"Low stars, trivial repo","source":"signal"},{"label":"Watch term","value":"Eval methodology","source":"evidence"},{"label":"Watch term","value":"Infrastructure","source":"evidence"}],"evidence_coverage":{"target_pages":1,"captured_pages":1,"readable_pages":1,"capture_methods":["plain"],"missing_page_urls":[],"failed_page_urls":[],"blocked_page_urls":[],"page_urls":["https://github.com/makora-ai/flash-moe"],"related_signals":5,"has_source_url":true,"latest_page_fetched_at":"2026-06-11T04:08:24.39503+00:00"},"data_business":{"matches":false,"lanes":[],"matched_terms":[],"score":null,"reason":null},"agent_handoff":{"signal_json":"https://onlylabs.fyi/signals/373a0500-eaea-4c6b-ac9f-ed1ebfa9355e/signal.json","dossier_json":"https://onlylabs.fyi/labs/makora/dossier.json","analysis_json":"https://onlylabs.fyi/analysis/makora/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/makora/evidence.json","topic_signals_json":null,"topic_feed":null,"category_signals_json":"https://onlylabs.fyi/signals.json?category=neocloud","data_radar_json":null,"opportunities_json":null},"analysis_playbook":{"objective":"Turn new repository signals into early evidence of tooling, eval, infrastructure, model-adjacent, or product work before it appears in polished launch channels.","evidence_focus":["repo name","owner","description","language","stars","source URL","first seen time","data, eval, infra, safety, and product terms"],"extraction_questions":["What technical area does this repository expose?","Does the repo imply eval, data, infrastructure, agent, or deployment work?","Is the repo new evidence for a lab direction that is not yet in writing or releases?","Which related signals should an analyst inspect next?"],"signal_questions":["What does this new repository reveal before a formal announcement exists?","What technical area does this repository expose?","Does the repo imply eval, data, infrastructure, agent, or deployment work?","Do the 5 related repo signals show a repeated pattern?"],"output_fields":["org","repo","technical_theme","evidence_url"],"data_business_relevance":"Data-business lane extraction is scoped to frontier labs; for this category, interpret the repository as source-grounded category strategy evidence.","required_sources":[{"label":"signal_json","url":"https://onlylabs.fyi/signals/373a0500-eaea-4c6b-ac9f-ed1ebfa9355e/signal.json","required":true},{"label":"source","url":"https://github.com/makora-ai/flash-moe","required":true},{"label":"dossier_json","url":"https://onlylabs.fyi/labs/makora/dossier.json","required":true},{"label":"analysis_evidence_json","url":"https://onlylabs.fyi/analysis/makora/evidence.json","required":true},{"label":"topic_signals_json","url":null,"required":false},{"label":"data_radar_json","url":null,"required":false}],"expected_output":["one-paragraph source-grounded interpretation","category-specific implication","confidence and missing evidence","recommended next source to inspect"],"prompt_seed":"Using only the linked onlylabs JSON, captured source context, and cited evidence, analyze Makora's repo signal \"makora-ai/flash-moe\" for neocloud strategy."},"semantic_triples":[{"subject":"Makora","predicate":"published repo","object":"makora-ai/flash-moe","text":"Makora published repo makora-ai/flash-moe."},{"subject":"makora-ai/flash-moe","predicate":"is classified as","object":"repo signal","text":"makora-ai/flash-moe is classified as repo signal."},{"subject":"makora-ai/flash-moe","predicate":"belongs to","object":"repos desk","text":"makora-ai/flash-moe belongs to repos desk."},{"subject":"makora-ai/flash-moe","predicate":"has context","object":"C++","text":"makora-ai/flash-moe has context C++."},{"subject":"makora-ai/flash-moe","predicate":"has evidence coverage","object":"1 captured evidence page","text":"makora-ai/flash-moe has evidence coverage 1 captured evidence page."},{"subject":"makora-ai/flash-moe","predicate":"has captured page count","object":"1","text":"makora-ai/flash-moe has captured page count 1."},{"subject":"makora-ai/flash-moe","predicate":"has readable page count","object":"1","text":"makora-ai/flash-moe has readable page count 1."},{"subject":"makora-ai/flash-moe","predicate":"has related signal count","object":"5","text":"makora-ai/flash-moe has related signal count 5."},{"subject":"makora-ai/flash-moe","predicate":"has analysis playbook objective","object":"Turn new repository signals into early evidence of tooling, eval, infrastructure, model-adjacent, or product work before it appears in polished launch channels.","text":"makora-ai/flash-moe has analysis playbook objective Turn new repository signals into early evidence of tooling, eval, infrastructure, model-adjacent, or product work before it appears in polished launch channels.."},{"subject":"makora-ai/flash-moe","predicate":"has source host","object":"github.com","text":"makora-ai/flash-moe has source host github.com."},{"subject":"makora-ai/flash-moe","predicate":"has lab","object":"Makora","text":"makora-ai/flash-moe has lab Makora."},{"subject":"makora-ai/flash-moe","predicate":"has signal desk","object":"repos","text":"makora-ai/flash-moe has signal desk repos."},{"subject":"makora-ai/flash-moe","predicate":"has source host","object":"github.com","text":"makora-ai/flash-moe has source host github.com."},{"subject":"makora-ai/flash-moe","predicate":"has repository","object":"makora-ai/flash-moe","text":"makora-ai/flash-moe has repository makora-ai/flash-moe."},{"subject":"makora-ai/flash-moe","predicate":"has language","object":"C++","text":"makora-ai/flash-moe has language C++."},{"subject":"makora-ai/flash-moe","predicate":"has stars","object":"1","text":"makora-ai/flash-moe has stars 1."},{"subject":"makora-ai/flash-moe","predicate":"has notability","object":"Low stars, trivial repo","text":"makora-ai/flash-moe has notability Low stars, trivial repo."},{"subject":"makora-ai/flash-moe","predicate":"has watch term","object":"Eval methodology","text":"makora-ai/flash-moe has watch term Eval methodology."}]},"intelligence":{"signal_desk":"repos","answer":"Makora published makora-ai/flash-moe (C++). This repository signal exposes tooling, eval, infrastructure, or model-adjacent work before it may appear in a launch post. High-signal details: repo makora-ai/flash-moe · language C++ · Low stars, trivial repo. onlylabs links this event to 1 captured evidence page and 5 related repo signals.","semantic_triples":[{"subject":"Makora","predicate":"published repo","object":"makora-ai/flash-moe","text":"Makora published repo makora-ai/flash-moe."},{"subject":"makora-ai/flash-moe","predicate":"is classified as","object":"repo signal","text":"makora-ai/flash-moe is classified as repo signal."},{"subject":"makora-ai/flash-moe","predicate":"belongs to","object":"repos desk","text":"makora-ai/flash-moe belongs to repos desk."},{"subject":"makora-ai/flash-moe","predicate":"has context","object":"C++","text":"makora-ai/flash-moe has context C++."},{"subject":"makora-ai/flash-moe","predicate":"has evidence coverage","object":"1 captured evidence page","text":"makora-ai/flash-moe has evidence coverage 1 captured evidence page."}]},"signal":{"id":"373a0500-eaea-4c6b-ac9f-ed1ebfa9355e","url":"https://onlylabs.fyi/signals/373a0500-eaea-4c6b-ac9f-ed1ebfa9355e","json_url":"https://onlylabs.fyi/signals/373a0500-eaea-4c6b-ac9f-ed1ebfa9355e/signal.json","source_url":"https://github.com/makora-ai/flash-moe","title":"makora-ai/flash-moe","summary":"Makora published a new repository. onlylabs watches repos for tooling, eval, infra, and model-adjacent work.","context":"C++","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"makora","name":"Makora","category":"neocloud"},"occurred_at":"2025-09-18T16:37:05+00:00","first_seen_at":"2026-06-05T22:32:19.171838+00:00","date_source":"source","evidence_coverage":{"target_pages":1,"captured_pages":1,"readable_pages":1,"capture_methods":["plain"],"missing_page_urls":[],"failed_page_urls":[],"blocked_page_urls":[],"page_urls":["https://github.com/makora-ai/flash-moe"]},"facets":{"repo":"makora-ai/flash-moe","language":"C++"},"traction":{"github_stars":1,"hn_points":null,"hn_comments":null,"hn_story_id":null,"hf_downloads":null,"hf_likes":null},"data_radar":null},"primary_evidence_page":{"url":"https://github.com/makora-ai/flash-moe","final_url":"https://github.com/makora-ai/flash-moe","title":"makora-ai/flash-moe repository metadata","http_status":200,"content_type":"application/json","capture_method":"plain","fetched_at":"2026-06-11T04:08:24.39503+00:00","bytes":10336,"raw_path":"d6aaf094bcb4abfd172a640aa1bcd1e4ac666faea5b7db8bea00b39bc9ded348.json","content_hash":"ec9c7ee6d8fe7923a2d9eaf45246f5e9d9428f2d53f6bb6154cf78bb1ece3946","excerpt_chars":1200,"truncated":true,"excerpt":"makora-ai/flash-moe Language: C++ Stars: 1 Forks: 1 Open issues: 1 Created: 2025-09-18T16:37:05Z Pushed: 2026-01-30T11:08:19Z Default branch: main Fork: no Archived: no README: flash-moe Project status at last day of active development(28.01.2026). Flash-moe project aims to speedup Mixture of Experts operator for case where experts weights are distributed to different GPUs(each GPU gets different set of experts, assinged linearly) and each GPU has distict set of tokens. The initial focus was on decode phase where expected number of tokens is small. The main idea is to have a single kernel, which performs gate + experts calculations and also initalizes and transfers tokens to other GPUs at the same time. THe expected speedup comes from the fact that in this sceario kernel can overlap communication with other GPUS with computations and no all2all barriers are needed. The idea is based on https://arxiv.org/abs/2506.04667 paper, but implementation is for AMD GPUs and ecosystem. The goal of current implementation was to provide POC implementation with focus of being correct and bug-free. Becasue of that all subsystems are not optimized. Goal of POC: - proof of the usefullness of the..."},"evidence_pages":[{"url":"https://github.com/makora-ai/flash-moe","final_url":"https://github.com/makora-ai/flash-moe","title":"makora-ai/flash-moe repository metadata","http_status":200,"content_type":"application/json","capture_method":"plain","fetched_at":"2026-06-11T04:08:24.39503+00:00","bytes":10336,"raw_path":"d6aaf094bcb4abfd172a640aa1bcd1e4ac666faea5b7db8bea00b39bc9ded348.json","content_hash":"ec9c7ee6d8fe7923a2d9eaf45246f5e9d9428f2d53f6bb6154cf78bb1ece3946","excerpt_chars":1200,"truncated":true,"excerpt":"makora-ai/flash-moe Language: C++ Stars: 1 Forks: 1 Open issues: 1 Created: 2025-09-18T16:37:05Z Pushed: 2026-01-30T11:08:19Z Default branch: main Fork: no Archived: no README: flash-moe Project status at last day of active development(28.01.2026). Flash-moe project aims to speedup Mixture of Experts operator for case where experts weights are distributed to different GPUs(each GPU gets different set of experts, assinged linearly) and each GPU has distict set of tokens. The initial focus was on decode phase where expected number of tokens is small. The main idea is to have a single kernel, which performs gate + experts calculations and also initalizes and transfers tokens to other GPUs at the same time. THe expected speedup comes from the fact that in this sceario kernel can overlap communication with other GPUS with computations and no all2all barriers are needed. The idea is based on https://arxiv.org/abs/2506.04667 paper, but implementation is for AMD GPUs and ecosystem. The goal of current implementation was to provide POC implementation with focus of being correct and bug-free. Becasue of that all subsystems are not optimized. Goal of POC: - proof of the usefullness of the..."}],"related_signals":[{"id":"356bb6de-a92a-4eb1-82f4-a84d0ecfee47","url":"https://onlylabs.fyi/signals/356bb6de-a92a-4eb1-82f4-a84d0ecfee47","source_url":"https://github.com/makora-ai/makora","title":"makora-ai/makora","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"makora","name":"Makora","category":"neocloud"},"occurred_at":"2026-02-26T17:36:55+00:00","first_seen_at":"2026-06-05T22:32:19.171838+00:00","date_source":"source"},{"id":"995695a2-b823-42d4-abce-17df1edd1f3b","url":"https://onlylabs.fyi/signals/995695a2-b823-42d4-abce-17df1edd1f3b","source_url":"https://github.com/makora-ai/kernels","title":"makora-ai/kernels","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"makora","name":"Makora","category":"neocloud"},"occurred_at":"2025-12-10T19:00:32+00:00","first_seen_at":"2026-06-05T22:32:19.171838+00:00","date_source":"source"},{"id":"8f136de4-95f9-4a8a-ae3a-0cc6b3e6051a","url":"https://onlylabs.fyi/signals/8f136de4-95f9-4a8a-ae3a-0cc6b3e6051a","source_url":"https://github.com/makora-ai/mako-generate-agent-playground","title":"makora-ai/mako-generate-agent-playground","context":"Shell","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"makora","name":"Makora","category":"neocloud"},"occurred_at":"2025-09-17T11:45:58+00:00","first_seen_at":"2026-06-05T22:32:19.171838+00:00","date_source":"source"},{"id":"515efaa5-8c60-41c3-9e1e-33a19cb03c14","url":"https://onlylabs.fyi/signals/515efaa5-8c60-41c3-9e1e-33a19cb03c14","source_url":"https://github.com/makora-ai/aiagent_playground","title":"makora-ai/aiagent_playground","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"makora","name":"Makora","category":"neocloud"},"occurred_at":"2025-02-10T19:36:39+00:00","first_seen_at":"2026-06-05T22:32:19.171838+00:00","date_source":"source"},{"id":"0a85d69b-a06d-441d-87f6-53cfca444e16","url":"https://onlylabs.fyi/signals/0a85d69b-a06d-441d-87f6-53cfca444e16","source_url":"https://github.com/makora-ai/gpuq","title":"makora-ai/gpuq","context":"Python","kind":{"key":"repo_new","label":"Repo"},"org":{"slug":"makora","name":"Makora","category":"neocloud"},"occurred_at":"2024-11-15T13:56:54+00:00","first_seen_at":"2026-06-05T22:32:19.171838+00:00","date_source":"source"}]}