BigQuery beyond SQL and JS: Running C and Rust code at scale with wasm

1 row in 3 seconds is not impressive. 5 billion rows in 56 seconds is. Let’s see how we went from a to b: BigQuery strengths: Throughput, not latencyIf you are expecting results in less than one second, BigQuery is not the right tool: Usually BigQuery won’t return results in less than one second. This because BigQuery was built for throughput, not for latency. A race car will always be faster than

[sh19910711]

2018年に書かれた記事なのでLegacy SQLの話だけどBigQueryのUDFで処理するときにレコードをまとめてバッチ形式にすると速く捌けるケースがあるっぽい / "BigQuery UDF tricks: For slow UDFs, batch the requests"

[gfx]

BigQueryはJSを実行できて、その中でWasmを使うこともできるから、BQ用にCやRustで任意の処理をかける、と…！！