{"type":"post","id":"project-maya-v1-1-35-keeps-every-glm-2a210a","slug":"project-maya-v1-1-35-keeps-every-glm-2a210a","title":"Project Maya v1.1.35 keeps every GLM expert in NVIDIA VRAM when it fits, 2.5x decode on 4x A100","description":"Project Maya v1.1.35 changes how GLM experts are kept on NVIDIA GPUs: when the full expert set fits in VRAM, it stays resident through long prompts rather than being evicted.","retrieval_nugget":"Project Maya v1.1.35 changes how GLM experts are kept on NVIDIA GPUs: when the full expert set fits in VRAM, it stays resident through long prompts rather than being evicted. The changelog reports 2.5x decode on one 4x A100 split, a result relevant to teams tuning local inference but not a general performance guarantee.","published_at":"2026-11-04","updated_at":"2026-11-07","record_date":"2026-11-04","date_kind":"published_at","topics":["prompts"],"entities":["Project Maya","GLM","NVIDIA","A100"],"related_patterns":["harness-architecture-outlives-model-choice"],"novelty_label":"same_event","prior_refs":[],"source_url":"https://github.com/mw00/project-maya/blob/main/CHANGELOG.md","source_urls":["https://github.com/mw00/project-maya/blob/main/CHANGELOG.md"],"schema_version":"newruntime-agent-readable-v0.2","stable_id":"post:project-maya-v1-1-35-keeps-every-glm-2a210a","status":"published","visuals":[],"editor":{"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]},"routes":{"html":"https://newruntime.com/posts/project-maya-v1-1-35-keeps-every-glm-2a210a/","markdown":"https://newruntime.com/posts/project-maya-v1-1-35-keeps-every-glm-2a210a.md","json":"https://newruntime.com/posts/project-maya-v1-1-35-keeps-every-glm-2a210a.json"},"markdown_url":"https://newruntime.com/posts/project-maya-v1-1-35-keeps-every-glm-2a210a.md","next_reads":[{"type":"topic","path":"/topics/prompts/","url":"https://newruntime.com/topics/prompts/","title":"Prompts","reason":"Explore the prompts topic hub.","media_type":"text/html","markdown_url":"https://newruntime.com/topics/prompts.md","json_url":"https://newruntime.com/topics/prompts.json"},{"type":"collection","path":"/posts/","url":"https://newruntime.com/posts/","title":"Field Notes","reason":"Open the latest editorial analysis.","media_type":"text/html","json_url":"https://newruntime.com/posts.json"},{"type":"collection","path":"/sources/","url":"https://newruntime.com/sources/","title":"Sources","reason":"See the original sources this archive cites.","media_type":"text/html"}]}