Module vectorized 

Vectorized Batch Processing

High-performance batch operations leveraging SIMD-optimized crates:

• CRC32: Hardware-accelerated checksums via crc32fast (SSE4.2/AVX2/ARMv8)
• Memory search: SIMD byte search via memchr (AVX2/SSE2/NEON)
• Batch message encoding: Encode/decode multiple messages in single pass
• Columnar RecordBatch: Cache-friendly columnar layout for filtering
• Parallel BatchProcessor: Thread-pool batch map/filter

The SIMD acceleration is provided by the underlying crates, which auto-detect and use the best available instruction set at runtime. No hand-written SIMD intrinsics are needed — the libraries handle platform-specific dispatch.

Performance characteristics:

• 4-8x faster checksums vs naive loop (via crc32fast)
• 2-4x faster batch encoding vs sequential
• Near-zero allocation hot path

Structs

BatchDecoder

Batch decoder for high-throughput message decoding

BatchEncoder

Batch encoder for high-throughput message encoding

BatchMessage

Message for batch encoding

BatchProcessor

Batch processor for parallel operations

DecodedMessage

Decoded message

DecoderStatsSnapshot

EncoderStatsSnapshot

ProcessorStatsSnapshot

RecordBatch

Record batch for columnar storage/processing

RecordBatchIter

Iterator over RecordBatch

Functions

crc32_batch

Batch CRC32 calculation for multiple buffers

crc32_fast

Fast CRC32 calculation using hardware acceleration when available

memchr_fast

Vectorized memory search

memcmp_fast

Vectorized memory comparison

memmem_fast

Vectorized pattern search