Shopify’s journey to faster breadth-first GraphQL execution

Summary

Shopify describes how high-cardinality, deeply nested GraphQL queries exposed execution overhead beyond data loading, with field resolvers assembling responses consuming much of request time. Its GraphQL Cardinal engine replaces conventional depth-first traversal with breadth-first execution, resolving each selected field once across an aggregated set of objects and reducing repeated field-level work. For large list queries, Shopify reports up to 15x faster execution, 90% less memory use, and reductions of many seconds in P50 latency. The migration included a tuned interpreter, cheaper tracer behavior, shadow verification, benchmarks, and ongoing conversion of tens of thousands of legacy resolvers, while future work includes asynchronous patterns and lower-level C bindings.

Takeaways

1. 1

   Conventional depth-first GraphQL traversal processes each object's subtree independently, so field execution and its associated overhead scale linearly with the response breadth.

2. 2

   Cardinal gives field resolvers sets of objects and mapped results, making breadth batching native to the execution engine instead of layering it on depth traversal with dataloaders.

3. 3

   Shopify reports up to 15x faster execution and 90% less memory for large list queries; regressions found during resolver migration have so far been attributed to translation errors.