Benchmarks

Every measured warm Reservoir path allocated 0 B per operation.

Results below used BenchmarkDotNet 0.15.8 MediumRun, .NET 10.0.11, Windows 11, AMD EPYC 9V74. Nanosecond timings vary by machine; compare methods within a table.

Automated results

Generated 2026-08-11 23:33 UTC from commit d32ff7f69c75. See the GitHub Actions run for logs and downloadable artifacts.

Core pool

The payload owns a 256-byte buffer. Lower ratio is better; new is the baseline.

Method	Mean	Ratio	Allocated
new	11.83 ns	1.00	304 B
Reservoir	10.36 ns	0.88	0 B
Microsoft.Extensions.ObjectPool	11.57 ns	0.98	0 B
ConcurrentBag<T> pool	25.19 ns	2.13	0 B

Warm allocation guarantee

Pool	Mean	Allocated
ObjectPool	10.29 ns	0 B
ListPool	10.97 ns	0 B
DictionaryPool	11.90 ns	0 B
HashSetPool	12.01 ns	0 B
QueuePool	11.32 ns	0 B
StackPool	11.14 ns	0 B
StringBuilderPool	15.91 ns	0 B

Specialized workloads

Workload	Baseline	Reservoir	Baseline allocated	Reservoir allocated
StringBuilder, append 128 chars	29.67 ns	20.82 ns	400 B	0 B
List<int>, 8 items	17.76 ns	25.01 ns	88 B	0 B
List<int>, 128 items	196.35 ns	184.15 ns	568 B	0 B
List<int>, 2,048 items	3,036.25 ns	2,750.70 ns	8,248 B	0 B

The single-thread TLS StringBuilder cache measured 10.35 ns and 0 B; it gives up cross-thread reuse and bounded shared capacity.

Choosing a rental API

For synchronous hot paths on .NET 10, prefer RentScoped(out T). On .NET 8, manual Rent() and Return() remain faster. Both warm paths allocate 0 B. Always use manual rental when ownership crosses an await, and measure representative workloads on target hardware.

Reproduce

dotnet run -c Release -f net10.0 --project benchmarks/Reservoir.Benchmarks -- --filter "*" --job Short --runtimes net8.0 net10.0 --apples

Raw Markdown, CSV, and HTML exports—including 1–32 worker contention results—are available from the GitHub Actions run.