SearchValues<T> in .NET 8/9 (.NET & C#)

Day 1/30 — SearchValues in .NET 8/9 (.NET & C#)

Spotted string parsing bottlenecks during a recent code review where repeated IndexOfAny calls were triggering excessive CPU cycles on hot paths.

• SIMD Vectorization: JIT compiler auto-vectorizes instructions to scan multiple bytes simultaneously across wide registers.
• Heap Allocation: Caching the SearchValues instance eliminates repeated array creation overhead inside tight loops.
• Memory Safety: Operating directly on ReadOnlySpan prevents unnecessary string slicing and Gen0 GC pressure.

Watch out: Always cache SearchValues as a static readonly field to avoid costly instance initialization overhead on every method call.

Have you benchmarked SIMD primitives in your ingestion pipelines yet?

DotNet #CSharp #Performance #BackendDevelopment #SoftwareEngineering