For processing 2MB files per minute, which option minimizes Snowflake cost?

• A. CSV with Snowpipe
• B. AVRO with Snowpipe
• C. Create a dedicated virtual warehouse and use JSON file format
• D. Create a dedicated virtual warehouse and use AVRO file format

Answer: (A)

For this scenario, the goal is to minimize compute and keep ingestion costs low for a steady, light stream. Snowpipe uses on-demand, auto-scaling compute to ingest files, so you only pay for the actual work needed to load the 2MB per minute, without keeping a dedicated warehouse running continuously. A simple CSV format tends to parse more quickly and with less CPU overhead than more verbose or complex formats, reducing the time Snowflake spends loading each file. Using a dedicated virtual warehouse means paying for that compute continuously, even during idle periods, and processing more complex formats like AVRO or JSON often requires more CPU for parsing, increasing cost. Put together, CSV with Snowpipe delivers the lowest ongoing ingestion cost for this rate.