{"schema_version":"2026-09-15.story-v1","story":{"slug":"what-runs-between-agent-model-calls","title":"What runs between an agent’s model calls?","description":"Connect industrial agents to their execution environments, CPU measurements, and processor architecture.","type":"reading-guide","editorial_question":"Is an agent waiting for its model, executing a tool, or waiting for another system?","limits":"This reading path is editorial analysis. It does not establish a universal CPU bottleneck or forecast processor sales. Industrial deployment claims and experimental computing results describe different populations.","canonical_url":"https://www.manufacturing.ai/stories/what-runs-between-agent-model-calls"},"policy":{"editorial_prompts_are_not_evidence":true,"publication_is_not_human_verification":true,"preserve_assertion_assessments":true,"evidence_entries_may_share_a_source":true,"linked_documents_do_not_add_verification":true,"do_not_treat_a_connection_as_causation":true,"full_topic_records":"https://www.manufacturing.ai/api/v1/topics/{slug}"},"sections":[{"title":"Start with the work the agent does","editorial_prompt":"Distinguish a vendor's description of an autonomous workflow from measured production outcomes.","topic_slug":"agentic-ai","concept_slug":"agentic-ai","topic_url":"https://www.manufacturing.ai/topics/agentic-ai","evidence_url":"https://www.manufacturing.ai/topics/agentic-ai#agentic-ai","assertions":[{"id":"218463e3-593d-41bc-b542-84cd4992f019","statement":"Industrial agentic AI executes workflows rather than answering questions — monitoring process data, deciding and acting — as distinct from conversational copilots; Siemens describes its agents as independently executing complete industrial workflows.","assessment":"reported","editor_reviewed":false,"reference_only_count":0,"evidence":[{"id":"bb371d79-a8ae-4eae-bec2-59748ee99c32","stance":"supports","is_primary_for_assertion":true,"origin_independence":"unknown","note":null,"quote":null,"source":{"title":"Siemens introduces AI agents for industrial automation","publisher":"Siemens","url":"https://press.siemens.com/global/en/pressrelease/siemens-introduces-ai-agents-industrial-automation","sourceType":"press_release","retrievalStatus":"fetched","inspected":true,"publishedAt":"2025-05-12","content_inspected":true,"documents":[]}},{"id":"b5007bf3-f86b-48f0-a74e-d4ddd9072a5a","stance":"qualifies","is_primary_for_assertion":false,"origin_independence":"unknown","note":null,"quote":null,"source":{"title":"Siemens introduces AI agents for industrial automation","publisher":"Siemens","url":"https://press.siemens.com/global/en/pressrelease/siemens-introduces-ai-agents-industrial-automation","sourceType":"press_release","retrievalStatus":"fetched","inspected":true,"publishedAt":"2025-05-12","content_inspected":true,"documents":[]}}]},{"id":"58ee52cc-8cea-4f2b-adbd-3e7a5fc579cd","statement":"At Hannover Messe 2026, twelve IoT Analytics analysts counted 29 industrial agentic AI solutions of which 72% were commercially available rather than pilots and 78% were model-agnostic — a measure of what vendors are selling, not what manufacturers are running.","assessment":"reported","editor_reviewed":false,"reference_only_count":0,"evidence":[{"id":"e24f9ac6-64e8-4314-b23a-cd83121810c7","stance":"supports","is_primary_for_assertion":false,"origin_independence":"unknown","note":null,"quote":null,"source":{"title":"Mid-2026 industrial AI pulse check: Is this the year of agentic AI?","publisher":"IoT Analytics","url":"https://iot-analytics.com/mid-2026-industrial-ai-pulse-check-is-this-the-year-of-agentic-ai/","sourceType":"other","retrievalStatus":"fetched","inspected":true,"publishedAt":"2026-08-03","content_inspected":true,"documents":[]}},{"id":"65e5ae67-307b-4335-85b7-a3327f9e80d6","stance":"qualifies","is_primary_for_assertion":false,"origin_independence":"unknown","note":null,"quote":null,"source":{"title":"Mid-2026 industrial AI pulse check: Is this the year of agentic AI?","publisher":"IoT Analytics","url":"https://iot-analytics.com/mid-2026-industrial-ai-pulse-check-is-this-the-year-of-agentic-ai/","sourceType":"other","retrievalStatus":"fetched","inspected":true,"publishedAt":"2026-08-03","content_inspected":true,"documents":[]}}]}]},{"title":"Locate the execution environment","editorial_prompt":"Separate the model, harness, and tools before assigning resource usage to a processor.","topic_slug":"agentic-cpu","concept_slug":"agent-runtime-and-tools","topic_url":"https://www.manufacturing.ai/topics/agentic-cpu","evidence_url":"https://www.manufacturing.ai/topics/agentic-cpu#agent-runtime-and-tools","assertions":[{"id":"30776e8e-6d5c-4ee2-b5ee-7a21ec984cda","statement":"Anthropic describes Managed Agents as separate session, harness, and sandbox components. The harness calls the model and routes tool requests; the sandbox executes code and edits files. This is an operator's architecture description, not a measurement of CPU consumption. It explains why model serving and tool execution should be examined separately.","assessment":"reported","editor_reviewed":false,"reference_only_count":0,"evidence":[{"id":"1f707fb5-cd90-4764-a4ec-3ae12c6bdb5b","stance":"supports","is_primary_for_assertion":false,"origin_independence":"unknown","note":null,"quote":null,"source":{"title":"Scaling Managed Agents: Decoupling the brain from the hands","publisher":"Anthropic","url":"https://www.anthropic.com/engineering/managed-agents","sourceType":"other","retrievalStatus":"fetched","inspected":true,"publishedAt":"2026-04-08","content_inspected":true,"documents":[]}},{"id":"06dba13d-c15b-4cc1-b709-f45920661756","stance":"qualifies","is_primary_for_assertion":false,"origin_independence":"unknown","note":null,"quote":null,"source":{"title":"Scaling Managed Agents: Decoupling the brain from the hands","publisher":"Anthropic","url":"https://www.anthropic.com/engineering/managed-agents","sourceType":"other","retrievalStatus":"fetched","inspected":true,"publishedAt":"2026-04-08","content_inspected":true,"documents":[]}}]}]},{"title":"Separate work from waiting","editorial_prompt":"Identify whether the reported time is elapsed latency or CPU time, and which process or machine the measurement covers.","topic_slug":"agentic-cpu","concept_slug":"agent-cpu-time-vs-elapsed","topic_url":"https://www.manufacturing.ai/topics/agentic-cpu","evidence_url":"https://www.manufacturing.ai/topics/agentic-cpu#agent-cpu-time-vs-elapsed","assertions":[{"id":"74640b66-267c-46c0-bc33-9be84b56db10","statement":"Python's process_time measures the current process's user and system CPU time and excludes sleep. perf_counter measures elapsed duration, including sleep. A slow tool call therefore does not by itself establish high CPU use. A timer around the harness also does not account for CPU work performed by a separate sandbox or remote service.","assessment":"reported","editor_reviewed":false,"reference_only_count":0,"evidence":[{"id":"685e21dc-67d5-42cb-a3f4-2505ca645fe7","stance":"supports","is_primary_for_assertion":false,"origin_independence":"unknown","note":null,"quote":null,"source":{"title":"Python time documentation: perf_counter and process_time","publisher":"Python Software Foundation","url":"https://docs.python.org/3/library/time.html","sourceType":"other","retrievalStatus":"fetched","inspected":true,"publishedAt":null,"content_inspected":true,"documents":[]}},{"id":"748b898a-2e3d-404a-b5b9-916059bd7b59","stance":"qualifies","is_primary_for_assertion":false,"origin_independence":"unknown","note":null,"quote":null,"source":{"title":"Python time documentation: perf_counter and process_time","publisher":"Python Software Foundation","url":"https://docs.python.org/3/library/time.html","sourceType":"other","retrievalStatus":"fetched","inspected":true,"publishedAt":null,"content_inspected":true,"documents":[]}}]}]},{"title":"Return to the processor architecture","editorial_prompt":"Read the host CPU's role alongside the workload evidence. A platform configuration is not a universal CPU-to-GPU requirement.","topic_slug":"cpu","concept_slug":"host-cpu-in-ai-systems","topic_url":"https://www.manufacturing.ai/topics/cpu","evidence_url":"https://www.manufacturing.ai/topics/cpu#host-cpu-in-ai-systems","assertions":[{"id":"2b583c21-226b-413b-877f-1c70e0596d29","statement":"NVIDIA's March 2026 Vera Rubin platform announces the Vera CPU as one of seven coordinated chips, with the NVL72 rack pairing 72 GPUs to 36 CPUs — the host processor shipping as a component of a rack-scale product rather than as the machine itself.","assessment":"reported","editor_reviewed":false,"reference_only_count":0,"evidence":[{"id":"09b9ed7e-941b-44cf-9c26-2e989aad8851","stance":"supports","is_primary_for_assertion":true,"origin_independence":"unknown","note":null,"quote":null,"source":{"title":"NVIDIA Vera Rubin Opens Agentic AI Frontier","publisher":"NVIDIA","url":"https://nvidianews.nvidia.com/news/nvidia-vera-rubin-platform","sourceType":"press_release","retrievalStatus":"fetched","inspected":true,"publishedAt":"2026-03-16","content_inspected":true,"documents":[]}}]},{"id":"7b349c96-95b1-4f33-9dfb-aa4da41895f3","statement":"NVIDIA's Grace CPU Superchip pairs 144 Arm Neoverse V2 cores and LPDDR5X memory at up to 1 TB/s with a 900 GB/s NVLink-C2C link to the accelerator — an Arm design in an x86 data centre, specified around the bandwidth of that link rather than around compute.","assessment":"reported","editor_reviewed":false,"reference_only_count":0,"evidence":[{"id":"dfa7d724-44db-4c3e-867a-e187c5ccb932","stance":"supports","is_primary_for_assertion":true,"origin_independence":"unknown","note":null,"quote":null,"source":{"title":"NVIDIA Grace CPU","publisher":"NVIDIA","url":"https://www.nvidia.com/en-us/data-center/grace-cpu/","sourceType":"company_website","retrievalStatus":"fetched","inspected":true,"publishedAt":null,"content_inspected":true,"documents":[]}}]}]}]}