{"schema_version":"onlylabs.public_signal.v1","title":"Together AI Writing: ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","description":"Together AI writing signal with public source context, captured evidence pages, related signals, and category-scoped analysis context.","url":"https://onlylabs.fyi/signals/85c5ec42-d179-4a01-9ea9-2d182bc2737f","json_url":"https://onlylabs.fyi/signals/85c5ec42-d179-4a01-9ea9-2d182bc2737f/signal.json","generated_at":"2026-06-27T00:35:02.717Z","evidence_latest_fetched_at":"2026-06-23T20:04:16.95523+00:00","signal_first_seen_at":"2026-06-23T20:00:28.992447+00:00","org":{"slug":"together-ai","name":"Together AI","category":"neocloud","category_label":"Neocloud","dossier_url":"https://onlylabs.fyi/labs/together-ai","dossier_json_url":"https://onlylabs.fyi/labs/together-ai/dossier.json"},"related_urls":{"signal":"https://onlylabs.fyi/signals/85c5ec42-d179-4a01-9ea9-2d182bc2737f","signal_json":"https://onlylabs.fyi/signals/85c5ec42-d179-4a01-9ea9-2d182bc2737f/signal.json","source":"https://www.together.ai/blog/parallelkernelbench","lab_dossier":"https://onlylabs.fyi/labs/together-ai","lab_dossier_json":"https://onlylabs.fyi/labs/together-ai/dossier.json","analysis":"https://onlylabs.fyi/analysis/together-ai","analysis_json":"https://onlylabs.fyi/analysis/together-ai/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/together-ai/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":"https://onlylabs.fyi/topics/talking","topic_signals_json":"https://onlylabs.fyi/topics/talking/signals.json?category=neocloud","topic_feed":"https://onlylabs.fyi/topics/talking/feed.xml?category=neocloud","data_business":null},"answer_pack":{"answer":"Together AI published ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet). This talking signal gives public context for research themes, product direction, policy, or launch framing. High-signal details: Substantive benchmark post revealing LLM limitations. · ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) 🚀 Now serving MiniMax-M3 for efficient inference → ⚡ On-demand B200s now available on.... onlylabs links this event to 1 captured evidence page and 6 related writing signals.","signal_desk":"talking","source_context":{"source_url":"https://www.together.ai/blog/parallelkernelbench","source_host":"together.ai","occurred_at":"2026-06-23T00:00:00+00:00","first_seen_at":"2026-06-23T20:00:28.992447+00:00","date_source":"rss.item_date","context":null},"context_markers":[{"label":"Lab","value":"Together AI","source":"signal"},{"label":"Signal desk","value":"talking","source":"signal"},{"label":"Source host","value":"together.ai","source":"source"},{"label":"Notability","value":"Substantive benchmark post revealing LLM limitations.","source":"signal"},{"label":"Watch term","value":"Eval methodology","source":"evidence"},{"label":"Watch term","value":"Infrastructure","source":"evidence"},{"label":"Watch term","value":"Agents and tool use","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://www.together.ai/blog/parallelkernelbench"],"related_signals":6,"has_source_url":true,"latest_page_fetched_at":"2026-06-23T20:04:16.95523+00:00"},"data_business":{"matches":false,"lanes":[],"matched_terms":[],"score":null,"reason":null},"agent_handoff":{"signal_json":"https://onlylabs.fyi/signals/85c5ec42-d179-4a01-9ea9-2d182bc2737f/signal.json","dossier_json":"https://onlylabs.fyi/labs/together-ai/dossier.json","analysis_json":"https://onlylabs.fyi/analysis/together-ai/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/together-ai/evidence.json","topic_signals_json":"https://onlylabs.fyi/topics/talking/signals.json?category=neocloud","topic_feed":"https://onlylabs.fyi/topics/talking/feed.xml?category=neocloud","category_signals_json":"https://onlylabs.fyi/signals.json?category=neocloud","data_radar_json":null,"opportunities_json":null},"analysis_playbook":{"objective":"Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.","evidence_focus":["post title","source URL","captured page text","HN traction","linked model or paper references","publication date"],"extraction_questions":["Which themes are labs choosing to explain publicly?","Which posts are attracting outside discussion?","Which writing reframes a recent release, model, hiring wave, or policy stance?","Which posts mention data, evals, infrastructure, safety, or deployment workflows?"],"signal_questions":["What public theme, launch framing, or research direction does this writing signal expose?","Which themes are labs choosing to explain publicly?","Which posts are attracting outside discussion?","Do the 6 related writing signals show a repeated pattern?"],"output_fields":["org","theme","public_framing","traction","evidence_url"],"data_business_relevance":"Data-business lane extraction is scoped to frontier labs; for this category, keep conclusions tied to category-specific strategy, source evidence, and follow-up questions.","required_sources":[{"label":"signal_json","url":"https://onlylabs.fyi/signals/85c5ec42-d179-4a01-9ea9-2d182bc2737f/signal.json","required":true},{"label":"source","url":"https://www.together.ai/blog/parallelkernelbench","required":true},{"label":"dossier_json","url":"https://onlylabs.fyi/labs/together-ai/dossier.json","required":true},{"label":"analysis_evidence_json","url":"https://onlylabs.fyi/analysis/together-ai/evidence.json","required":true},{"label":"topic_signals_json","url":"https://onlylabs.fyi/topics/talking/signals.json?category=neocloud","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 Together AI's writing signal \"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)\" for neocloud strategy."},"semantic_triples":[{"subject":"Together AI","predicate":"published","object":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","text":"Together AI published ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"is classified as","object":"writing signal","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) is classified as writing signal."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"belongs to","object":"talking desk","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) belongs to talking desk."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has evidence coverage","object":"1 captured evidence page","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has evidence coverage 1 captured evidence page."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has captured page count","object":"1","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has captured page count 1."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has readable page count","object":"1","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has readable page count 1."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has related signal count","object":"6","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has related signal count 6."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has analysis playbook objective","object":"Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has analysis playbook objective Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has source host","object":"together.ai","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has source host together.ai."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has lab","object":"Together AI","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has lab Together AI."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has signal desk","object":"talking","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has signal desk talking."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has source host","object":"together.ai","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has source host together.ai."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has notability","object":"Substantive benchmark post revealing LLM limitations.","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has notability Substantive benchmark post revealing LLM limitations.."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has watch term","object":"Eval methodology","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has watch term Eval methodology."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has watch term","object":"Infrastructure","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has watch term Infrastructure."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has watch term","object":"Agents and tool use","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has watch term Agents and tool use."}]},"intelligence":{"signal_desk":"talking","answer":"Together AI published ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet). This talking signal gives public context for research themes, product direction, policy, or launch framing. High-signal details: Substantive benchmark post revealing LLM limitations. · ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) 🚀 Now serving MiniMax-M3 for efficient inference → ⚡ On-demand B200s now available on.... onlylabs links this event to 1 captured evidence page and 6 related writing signals.","semantic_triples":[{"subject":"Together AI","predicate":"published","object":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","text":"Together AI published ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"is classified as","object":"writing signal","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) is classified as writing signal."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"belongs to","object":"talking desk","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) belongs to talking desk."},{"subject":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","predicate":"has evidence coverage","object":"1 captured evidence page","text":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) has evidence coverage 1 captured evidence page."}]},"signal":{"id":"85c5ec42-d179-4a01-9ea9-2d182bc2737f","url":"https://onlylabs.fyi/signals/85c5ec42-d179-4a01-9ea9-2d182bc2737f","json_url":"https://onlylabs.fyi/signals/85c5ec42-d179-4a01-9ea9-2d182bc2737f/signal.json","source_url":"https://www.together.ai/blog/parallelkernelbench","title":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","summary":"Together AI published a writing signal. onlylabs watches public writing for research themes, product direction, and model-launch context.","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2026-06-23T00:00:00+00:00","first_seen_at":"2026-06-23T20:00:28.992447+00:00","date_source":"rss.item_date","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://www.together.ai/blog/parallelkernelbench"]},"facets":{},"traction":{"github_stars":null,"hn_points":null,"hn_comments":null,"hn_story_id":null,"hf_downloads":null,"hf_likes":null},"data_radar":null},"primary_evidence_page":{"is_primary":true,"source_match":true,"url":"https://www.together.ai/blog/parallelkernelbench","final_url":"https://www.together.ai/blog/parallelkernelbench","title":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet)","http_status":200,"content_type":"text/html; charset=utf-8","capture_method":"plain","fetched_at":"2026-06-23T20:04:16.95523+00:00","bytes":345995,"raw_path":"1cd48943110558194b1e73480c0e2ba97b69627b38934b38c5d7525784f7590a.html","content_hash":"d9149bed732471ee0931401c00f10f019f05b480f704493ae51532046ec8ef5d","excerpt_chars":1200,"truncated":true,"excerpt":"ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) 🚀 Now serving MiniMax-M3 for efficient inference → ⚡ On-demand B200s now available on Together GPU Clusters → 📊 Delivering 31% more TPS than the next-fastest OSS engine for production coding agent workloads → 💬 How Together built the world's fastest speech-to-text stack → 🇫🇷 Join us at RAISE 2026 in Paris → All blog posts Research Published 6/23/2026 ParallelKernelBench: Frontier LLMs can't write fast multi-GPU kernels (yet) The best frontier model solves under a third of 87 real-world problems — but a few generated kernels beat anything publicly available. Authors Willy Chan, Nathan Paek, Simon Guo, Simran Arora, Daniel Y. Fu Table of contents 40+ Models Chosen for Production...40+ Models Chosen for Production...40+ Models Chosen for Production... Links in this article Paper HuggingFace Code Summary LLMs have gotten surprisingly good at writing GPU kernels [1][2][3] , but almost all current benchmarks measuring that progress are single-GPU. In production, communication is often the bottleneck: communication overhead can account for over 20% of inference latency [4] , and that gap keeps..."},"evidence_pages":[],"related_signals":[{"id":"5a4fd1e2-4560-4992-8745-899df77ecad5","url":"https://onlylabs.fyi/signals/5a4fd1e2-4560-4992-8745-899df77ecad5","source_url":"https://www.together.ai/blog/kimi-k2-7-code-vs-claude-fable-5","title":"Kimi K2.7 Code vs Claude Fable 5: Landing pages that cost 94% less","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2026-06-17T00:00:00+00:00","first_seen_at":"2026-06-17T20:00:28.497487+00:00","date_source":"rss.item_date"},{"id":"9294f377-1f3d-4b21-8078-53ecff3e7406","url":"https://onlylabs.fyi/signals/9294f377-1f3d-4b21-8078-53ecff3e7406","source_url":"https://www.together.ai/blog/iso-27001-2022-certification","title":"Building trust in enterprise AI: Together AI earns ISO 27001:2022 certification","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2026-06-10T00:00:00+00:00","first_seen_at":"2026-06-11T07:01:27.070847+00:00","date_source":"rss.item_date"},{"id":"33644a67-d468-44ed-8255-6990f9054eec","url":"https://onlylabs.fyi/signals/33644a67-d468-44ed-8255-6990f9054eec","source_url":"https://www.together.ai/blog/serving-minimax-m3-for-efficient-inference-unlocking-1m-token-context-and-multimodality-without-regrets","title":"Serving MiniMax-M3 for efficient inference: Unlocking 1M-Token Context and Multimodality Without Regrets ","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2026-06-02T00:00:00+00:00","first_seen_at":"2026-06-05T22:32:06.025484+00:00","date_source":"rss.item_date"},{"id":"56ba412f-f785-4495-a0c4-bec800f64fd3","url":"https://onlylabs.fyi/signals/56ba412f-f785-4495-a0c4-bec800f64fd3","source_url":"https://www.together.ai/blog/how-together-ai-built-the-worlds-fastest-speech-to-text-stack","title":"How Together AI built the world’s fastest speech-to-text stack","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2026-05-29T00:00:00+00:00","first_seen_at":"2026-06-05T22:32:06.025484+00:00","date_source":"rss.item_date"},{"id":"3c08a1c0-235e-42b0-b347-d52e39d12ee1","url":"https://onlylabs.fyi/signals/3c08a1c0-235e-42b0-b347-d52e39d12ee1","source_url":"https://www.together.ai/blog/coding-agent-benchmarks","title":"Benchmarking inference at scale: coding agents","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2026-05-19T00:00:00+00:00","first_seen_at":"2026-06-05T22:32:06.025484+00:00","date_source":"rss.item_date"},{"id":"49734867-446a-4524-963f-4812d706b5eb","url":"https://onlylabs.fyi/signals/49734867-446a-4524-963f-4812d706b5eb","source_url":"https://www.together.ai/blog/together-ai-partners-with-pearl-research-labs","title":"Together AI and Pearl Research Labs Team Up to Reduce the Cost of AI Inference","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"together-ai","name":"Together AI","category":"neocloud"},"occurred_at":"2026-05-15T00:00:00+00:00","first_seen_at":"2026-06-05T22:32:06.025484+00:00","date_source":"rss.item_date"}]}