Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Benchmarks for different variable parsing techniques #229

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
Open

Benchmarks for different variable parsing techniques #229

wants to merge 4 commits into from

Conversation

boopathi
Copy link
Member

@boopathi boopathi commented Sep 15, 2023 •
edited
Loading

Compare benchmarks between #82 and #228

yarn benchmark --bench variablesWithNewJit,variablesWithOldJit,variablesWithGraphQLJS

Results on my Mac -

size of function for variablesWithNewJit: 6727
size of function for variableCompilation variablesWithNewJit: 14847
size of function for variablesWithOldJit: 6727
size of function for variableCompilation variablesWithOldJit: 69500
size of function for variablesWithGraphQLJS: 6727
size of function for variableCompilation variablesWithGraphQLJS: undefined
Starting variablesWithNewJit
graphql-js x 148,906 ops/sec ±0.77% (234 runs sampled)
graphql-jit x 118,267 ops/sec ±0.17% (236 runs sampled)
Starting variablesWithOldJit
graphql-js x 159,093 ops/sec ±0.13% (236 runs sampled)
graphql-jit x 545,095 ops/sec ±0.11% (237 runs sampled)
Starting variablesWithGraphQLJS
graphql-js x 153,436 ops/sec ±0.14% (235 runs sampled)
graphql-jit x 237,255 ops/sec ±0.10% (235 runs sampled)

@boopathi boopathi mentioned this pull request Sep 15, 2023
Open
@boopathi
Copy link
Member Author

Other observations -

Though the old jit (without support for cyclic input type definitions) seems to perform well for shallower input nestedness, the amount of generated code increases rapidly as we increase more levels to the nestedness.

i.e. Adding one more level of input filters in the benchmark example - L3 to L4, the length of the generated code increases from 69500 to 344204, and the bench time / performance decreases by Half - 531,561 ops/sec to 265,269 ops/sec.

While the new jit (with support for cyclic input types), we don't add this complexity because it supports cyclic input type definitions.

@boopathi
Copy link
Member Author

Sending non-cyclic schema to all benchmarks, we see the following results -

nestedness at L4 - 4 levels of filters (refer variables-parsing-shallow.ts)

Starting variablesWithNewJit
graphql-jit x 148,555 ops/sec ±0.18% (235 runs sampled)
Starting variablesWithOldJit
graphql-jit x 262,649 ops/sec ±0.11% (236 runs sampled)
Starting variablesWithGraphQLJS
graphql-jit x 242,091 ops/sec ±0.11% (235 runs sampled)

@boopathi boopathi mentioned this pull request Feb 9, 2024
Open
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@ruiaraujo ruiaraujo Awaiting requested review from ruiaraujo

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
1 participant
@boopathi