Skip to content

Latest commit

 

History

History
78 lines (50 loc) · 2.85 KB

README.md

File metadata and controls

78 lines (50 loc) · 2.85 KB
Raw

puerts_node_performance

Tests

  • napi_perf: Implemented using napi

  • puerts_fastcall_perf: Implemented using puerts binding, and enabling PES_EXTENSION_WITH_V8_API and WITH_V8_FAST_CALL macros. Please note, to run this example, you need to manually download the v8-fast-api-calls.h for the corresponding local nodejs version and place it in the node-gyp header file directory.

  • puerts_perf: Implemented using puerts binding

  • puerts_v8_perf: Implemented using puerts binding, and enabling PES_EXTENSION_WITH_V8_API macro.

  • v8api_perf: Implemented directly using v8 api

  • v8pp_pref: Implemented using v8pp binding

How to manually download v8-fast-api-calls.h

First, check the node.js version, download the corresponding source code package, unzip it and copy v8-fast-api-calls.h to the node-gyp header file directory. Generally speaking, the node-gyp header file directory on Linux, macOS, and Windows are as follows:

  • Linux: ~/.cache/node-gyp/<node_version>/include/node, please replace <node_version> with the locally installed nodejs version, the same below.

  • macOS: ~/Library/Caches/node-gyp/<node_version>/include/node/

  • Windows: %userprofile%\AppData\Local\node-gyp\Cache<node_version>\include\node

Test Results

Environments

Windows

  • os: Windows 10 22H2
  • node version: v16.16.0
  • cpu: 12th Gen Intel(R) Core(TM) i7-1260P 2.10 GHz
  • memory: 32.0 GB

Ubuntu

  • os: Ubuntu 22.04.2 LTS (run on wsl)
  • node version: v18.14.1
  • cpu: 12th Gen Intel(R) Core(TM) i7-1260P 2.10 GHz
  • memory: 32.0 GB

macOS@intel

  • os: macOS Monterey 12.6.8
  • node version: v20.5.0
  • cpu: 10 core 10th-generation Intel Core i9 3.6G
  • memory: 32.0 GB

macOS@m1

  • os: macOS Ventura 13.5
  • node version: v18.9.0
  • cpu: Apple M1 Pro
  • memory: 32.0 GB

Test Results

1000000 static function calls

napi_perf puerts_fastcall_perf puerts_perf puerts_v8_perf v8api_perf v8pp_pref
Windows 25ms 4ms 24ms 16ms 13ms 30ms
Ubuntu 20ms 4ms 28ms 12ms 11ms 21ms
macOS@intel 27ms 5ms 25ms 21ms 15ms 30ms
macOS@m1 39ms 5ms 29ms 19ms 15ms 64ms

1000000 member method calls

puerts_fastcall_perf puerts_perf puerts_v8_perf v8pp_pref
Windows 11ms 29ms 19ms 138ms
Ubuntu 10ms 39ms 20ms 98ms
macOS@intel 10ms 37ms 25ms 88ms
macOS@m1 12ms 31ms 21ms 78ms