BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//cfp.riscv-europe.org//eu-summit-2026//talk//WWT8EV
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-eu-summit-2026-WWT8EV@cfp.riscv-europe.org
DTSTART;TZID=CET:20260610T171500
DTEND;TZID=CET:20260610T173000
DESCRIPTION:PELT (Per-Entity Load Tracking) is an exponential decay-based p
 er-entity load tracking algorithm in the Linux kernel. It significantly en
 hances the scheduler’s load awareness accuracy\, response latency\, and 
 energy efficiency. However\, there are still drawbacks in load tracking: T
 he load metrics that are used are not CPU-frequency invariant. The advent 
 of CPU frequency scaling causes task physical runtime to fluctuate with fr
 equency\, which\, if uncorrected\, distorts util_avg and leads to scheduli
 ng misjudgments. To address this\, the kernel employs hardware counters (e
 .g.\, Intel APERF/MPERF\, ARMv8.4-AMU) to implement frequency invariance a
 ccounting\, ensuring util_avg remains anchored to the CPU’s maximum capa
 city\, thereby maintaining load statistics accuracy and scheduling optimal
 ity in dynamic frequency environments.\nTargeting RISC-V architectures\, t
 his proposal introduces State sensitive Counters to fill the gap in PELT\n
 frequency invariance support. Together\, these counters enable the derivat
 ion of real-time operating frequency and normalized utilization without co
 stly synchronous queries.
DTSTAMP:20260522T163211Z
LOCATION:Plenary
SUMMARY:Proposal of State Sensitive Counter (Sssscnt) - Fengxue Zhang\, Boh
 ua Kou
URL:https://cfp.riscv-europe.org/eu-summit-2026/talk/WWT8EV/
END:VEVENT
END:VCALENDAR