BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//cfp.ducttape.events//ethprague-2026//speaker//PWP99T
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-ethprague-2026-T99VTV@cfp.ducttape.events
DTSTART;TZID=CET:20260509T150000
DTEND;TZID=CET:20260509T152000
DESCRIPTION:Zero-knowledge proofs are now operational infrastructure for ro
 llups\, private payments\, verifiable execution\, and zkVM-based applicati
 ons\, yet published benchmark results often compare different workloads\, 
 different proof modes\, and different hardware paths. This paper presents 
 a unified host--agent framework for benchmarking zero-knowledge proving sy
 stems under a common execution and reporting methodology. Each prover is i
 solated in a containerized agent\, selected through a scheme-adapter layer
 \, and required to report canonical JSON artifacts linked to a run manifes
 t and to workload-specific semantic contracts. We use the framework to com
 pare representative circuit-based\, transparent\, GPU-accelerated\, foldin
 g\, and zkVM systems: Groth16\, ICICLE\, Halo2\, Plonky3\, Winterfell\, No
 va\, RISC Zero\, RISC Zero CUDA/Groth16 variants\, SP1\, and SP1 without A
 VX. The benchmark suite separates six primary workloads with comparable se
 mantics from secondary hash workloads that expose specialization and accel
 eration effects. Across the primary workloads\, no prover family dominates
  all deployment criteria. At 8 CPU cores\, Plonky3 and Winterfell achieve 
 the lowest median proving latency and memory footprint\; Groth16 and ICICL
 E produce sub-kilobyte proofs and the lowest median verification latency\;
  and zkVM systems trade larger proofs and higher memory use for programmab
 ility. GPU acceleration is most beneficial once fixed initialization and s
 etup costs are amortized\, while CPU-oriented provers show strong scaling 
 from 1 to 8 cores. The results support a deployment-oriented view of ZKP s
 election: benchmarks should disclose security profiles\, distinguish compa
 rable from specialization-driven workloads\, and report resource limits as
  primary outputs rather than treating proof generation time alone as the r
 anking criterion.
DTSTAMP:20261017T230355Z
LOCATION:ETHRes
SUMMARY:Benchmarking Zero-Knowledge Proving Systems for Ethereum-Style Depl
 oyments in a Unified Framework - Matěj Hůlek and Martin Perešíni
URL:https://cfp.ducttape.events/ethprague-2026/talk/T99VTV/
END:VEVENT
END:VCALENDAR