Competitor Comparison

This page documents comparative benchmark baselines for Excalibur.Dispatch using two explicit tracks:

1. In-process parity (raw handler-dispatch style)
2. Queued/bus semantics (publish/send + consumer flow)

Baseline Artifacts

• Baseline folder: benchmarks/baselines/net10.0/dispatch-comparative-20260219/results/
• Runtime: .NET 10.0.3
• Tooling: BenchmarkDotNet v0.15.4
• Machine: Intel Core i9-14900K, Windows 11 (10.0.26200.7840)

Scope

These are microbenchmarks for framework overhead and path cost. They are not end-to-end production latency claims.

Executive Summary

Track	Summary
In-process parity (MediatR)	MediatR is faster in current baseline
In-process parity (Wolverine InvokeAsync)	Dispatch is faster
In-process parity (MassTransit Mediator)	Dispatch is faster
Queued/bus end-to-end parity (Dispatch remote route / Wolverine SendAsync / MassTransit bus)	Dispatch is faster in current baseline

Track A: In-Process Parity

Dispatch vs MediatR

Source: benchmarks/baselines/net10.0/dispatch-comparative-20260219/results/Excalibur.Dispatch.Benchmarks.Comparative.MediatRComparisonBenchmarks-report-github.md

Scenario	Dispatch	MediatR	Relative Result
Single command handler	118.79 ns	40.92 ns	MediatR ~2.9x faster
Single command strict direct-local	116.29 ns	40.92 ns	MediatR ~2.8x faster
Single command ultra-local API	47.12 ns	40.92 ns	MediatR ~1.2x faster
Notification to 3 handlers	154.47 ns	96.10 ns	MediatR ~1.6x faster
Query with return value	126.63 ns	49.29 ns	MediatR ~2.6x faster
Query ultra-local API	66.94 ns	49.29 ns	MediatR ~1.4x faster
10 concurrent commands	1,244.58 ns	497.81 ns	MediatR ~2.5x faster
100 concurrent commands	12,107.20 ns	4,797.88 ns	MediatR ~2.5x faster

Dispatch vs Wolverine (InvokeAsync parity)

Source: benchmarks/baselines/net10.0/dispatch-comparative-20260219/results/Excalibur.Dispatch.Benchmarks.Comparative.WolverineInProcessComparisonBenchmarks-report-github.md

Scenario	Dispatch	Wolverine (Invoke/local)	Relative Result
Single command	70.31 ns	193.14 ns	Dispatch ~2.7x faster
Notification to 2 handlers	75.95 ns	4,138.96 ns	Dispatch ~54.5x faster
Query with return	97.43 ns	259.53 ns	Dispatch ~2.7x faster
10 concurrent commands	747.27 ns	1,985.19 ns	Dispatch ~2.7x faster
100 concurrent commands	7,033.04 ns	19,976.37 ns	Dispatch ~2.8x faster

Dispatch vs MassTransit Mediator (in-process)

Source: benchmarks/baselines/net10.0/dispatch-comparative-20260219/results/Excalibur.Dispatch.Benchmarks.Comparative.MassTransitMediatorComparisonBenchmarks-report-github.md

Scenario	Dispatch	MassTransit Mediator	Relative Result
Single command	67.20 ns	1,185.70 ns	Dispatch ~17.7x faster
Notification to 2 consumers	83.36 ns	1,705.15 ns	Dispatch ~20.4x faster
Query with return	95.19 ns	18,771.04 ns	Dispatch ~197.2x faster
10 concurrent commands	779.68 ns	11,924.61 ns	Dispatch ~15.3x faster
100 concurrent commands	7,037.86 ns	120,396.41 ns	Dispatch ~17.1x faster

Track B: Queued/Bus End-to-End Parity

Dispatch vs Wolverine vs MassTransit (queued end-to-end parity)

Source: benchmarks/baselines/net10.0/dispatch-comparative-20260219/results/Excalibur.Dispatch.Benchmarks.Comparative.TransportQueueParityComparisonBenchmarks-report-github.md

Scenario	Dispatch (remote route)	Wolverine	MassTransit	Relative Result
Queued command end-to-end	1.317 us	4.005 us	22.655 us	Dispatch ~3.0x faster than Wolverine, ~17.2x faster than MassTransit
Queued event fan-out end-to-end	1.362 us	3.943 us	23.184 us	Dispatch ~2.9x faster than Wolverine, ~17.0x faster than MassTransit
Queued commands end-to-end (10 concurrent)	7.132 us	39.655 us	147.692 us	Dispatch ~5.6x faster than Wolverine, ~20.7x faster than MassTransit

Interpretation Guardrail

Use Track A for closest in-process handler overhead parity. Use Track B when comparing queued/bus completion semantics.

Routing-First Local + Hybrid Parity

Source: benchmarks/baselines/net10.0/dispatch-comparative-20260219/results/Excalibur.Dispatch.Benchmarks.Comparative.RoutingFirstParityBenchmarks-report-github.md

Scenario	Mean	Relative to local command
Dispatch: pre-routed local command	106.0 ns	baseline
Dispatch: pre-routed local query	141.3 ns	+33.3%
Dispatch: pre-routed remote event (AWS SQS)	183.3 ns	+72.9%
Dispatch: pre-routed remote event (Azure Service Bus)	191.8 ns	+81.0%
Dispatch: pre-routed remote event (Kafka)	189.1 ns	+78.4%
Dispatch: pre-routed remote event (RabbitMQ)	184.1 ns	+73.7%
Dispatch: pre-routed Kafka observability profile	321.2 ns	+203.0%
Dispatch: pre-routed RabbitMQ observability profile	321.4 ns	+203.2%

Running These Comparisons

# Build once
dotnet build benchmarks/Excalibur.Dispatch.Benchmarks/Excalibur.Dispatch.Benchmarks.csproj -c Release --nologo -v minimal

# Track A (in-process parity)
pwsh ./eng/run-benchmark-matrix.ps1 -NoBuild -NoRestore -Classes MediatRComparisonBenchmarks,WolverineInProcessComparisonBenchmarks,MassTransitMediatorComparisonBenchmarks

# Track B (queued/bus end-to-end parity)
pwsh ./eng/run-benchmark-matrix.ps1 -NoBuild -NoRestore -Classes TransportQueueParityComparisonBenchmarks

Results are written to BenchmarkDotNet.Artifacts/results/ unless -ArtifactsPath is provided.