About this question
Hard · low_latency · Quant Developer interview question · simd, avx2, json-parsing, simdjson, low_latency, byte-scan
High-throughput systems in quantitative finance, such as market data parsers, often use SIMD instructions to accelerate processing. This problem isolates the core logic of a SIMD-based JSON delimiter search, as used in libraries like simdjson, by implementing its scalar equivalent. Understanding this fundamental byte-scanning pattern is crucial for optimizing low-latency data ingestion pipelines. Task Implement the function count_delimiters(const double bytes, int n, int& colons, int& quotes) t