Serving Agentic Robot Fleets: Workload Characterization, KV-Cache Thrashing, and Admission Control
Overview
This research abstract details the characterization of workload constraints for agentic robot fleets, focusing on key-value (KV) cache management and action deadlines.
Generated from attributed reports · Updated 1 hours ago
Event evidence and corrections
0 attributed source owners. Ownership does not establish independent confirmation. Quantities are reported separately and are never added together.
Reported quantity · tokens: 215000 other · Basis not reported
“nerated by GPT-6-Astra, RPent, and pi0.5 (1,862 brain turns and 1,582 VLA inference chunks), profile both layers on H800 graphics processing units (GPUs), and replay the resulting agent programs on a vLLM serving engine. Brain contexts reach 215k tokens; the median Astra request takes 7.78 s, while the intervening environment-step gap is only 1.0 s. Under a 0.25 s action-chunk deadline, one VLA GPU meets the 1% starv”
Exact source · revision 1Source owner not reported
Reported quantity · request latency: 7.78 other · Basis not reported
“nerated by GPT-6-Astra, RPent, and pi0.5 (1,862 brain turns and 1,582 VLA inference chunks), profile both layers on H800 graphics processing units (GPUs), and replay the resulting agent programs on a vLLM serving engine. Brain contexts reach 215k tokens; the median Astra request takes 7.78 s, while the intervening environment-step gap is only 1.0 s. Under a 0.25 s action-chunk deadline, one VLA GPU meets the 1% starv”
Exact source · revision 1Source owner not reported
Reported quantity · environment-step gap: 1 other · Basis not reported
“nerated by GPT-6-Astra, RPent, and pi0.5 (1,862 brain turns and 1,582 VLA inference chunks), profile both layers on H800 graphics processing units (GPUs), and replay the resulting agent programs on a vLLM serving engine. Brain contexts reach 215k tokens; the median Astra request takes 7.78 s, while the intervening environment-step gap is only 1.0 s. Under a 0.25 s action-chunk deadline, one VLA GPU meets the 1% starv”
Exact source · revision 1Source owner not reported
Reported quantity · action-chunk deadline: 0.25 other · Basis not reported
“nerated by GPT-6-Astra, RPent, and pi0.5 (1,862 brain turns and 1,582 VLA inference chunks), profile both layers on H800 graphics processing units (GPUs), and replay the resulting agent programs on a vLLM serving engine. Brain contexts reach 215k tokens; the median Astra request takes 7.78 s, while the intervening environment-step gap is only 1.0 s. Under a 0.25 s action-chunk deadline, one VLA GPU meets the 1% starv”
Exact source · revision 1Source owner not reported
Report timeline
Follow attributed reports and material updates.
- arXiv Robotics — research abstractsSignalServing Agentic Robot Fleets: Workload Characterization, KV-Cache Thrashing, and Admission Control
This research abstract details the characterization of workload constraints for agentic robot fleets, focusing on key-value (KV) cache management and action deadlines.
Event coverage history
There is not enough continuous observation data to show a trend.
Timezone · UTC
Article dates follow your selected timezone. Briefing editions use Hong Kong time (UTC+8).