Benchmark: msgpack-lite vs msgpack5 vs. JSON-decode

HTML Preparation code:

<script src="https://rawgit.com/kawanet/msgpack-lite/master/dist/msgpack.min.js"></script>
<script src="https://rawgit.com/mcollina/msgpack5/master/dist/msgpack5.min.js"></script>

Script Preparation code:

var stringData = {"glossary":{"title":"example glossary","GlossDiv":{"title":"S","GlossList":{"GlossEntry":{"ID":"SGML","SortAs":"SGML","GlossTerm":"Standard Generalized Markup Language","Acronym":"SGML","Abbrev":"ISO 8879:1986","GlossDef":{"para":"A meta-markup language, used to create markup languages such as DocBook.","GlossSeeAlso":["GML","XML"]},"GlossSee":"markup"}}}}};
var numberData = [
  [1, 2, 3, 4, -1, true, null],
  [3, 6, 5, 4,  1, false, 7],
  [3, 2, 8, 1,  0, true, 0],
  [10, 11, 12, 13,  14, false, true],
  [15, 16, 17, 18, 19, true, null],
  [20, 21, 22, 23,  24, false, 7],
  [25, 26, 27, 28,  29, true, 0],
  100, 200, 300,
  [
    [1, 0, 0, 0, 0],
    [0, 1, 0, 0, 0],
    [0, 0, 1, 0, 0],
    [0, 0, 0, 1, 0],
    [0, 0, 0, 0, 1]
  ]
];
var jsonEncodedStrings = JSON.stringify(stringData),
    jsonEncodedNumbers = JSON.stringify(numberData),
    msgpackEncodedStrings = msgpack.encode(stringData),
    msgpackEncodedNumbers = msgpack.encode(numberData);
var msgpack55 = msgpack5();

Tests:

MsgPack Strings Encode
var result = msgpack.encode(stringData);
JSON Strings Encode
var result = JSON.stringify(stringData);
MsgPack Numbers Encode
var result = msgpack.encode(numberData);
JSON Numbers Encode
var result = JSON.stringify(numberData);
JSON string decode
var result = JSON.parse(jsonEncodedStrings);
msgpack string decode
var result = msgpack.decode(msgpackEncodedStrings);
MsgPack5 Strings Encode
var result = msgpack55.encode(stringData);
MsgPack5 Numbers Encode
var result = msgpack55.encode(numberData);
msgpack5 string decode
var result = msgpack55.decode(msgpackEncodedStrings);

Rendered benchmark preparation results:

Suite status: <idle, ready to run>

Previous results

Test case name	Result
MsgPack Strings Encode
JSON Strings Encode
MsgPack Numbers Encode
JSON Numbers Encode
JSON string decode
msgpack string decode
MsgPack5 Strings Encode
MsgPack5 Numbers Encode
msgpack5 string decode

Fastest: N/A

Slowest: N/A

Autogenerated LLM Summary (model llama3.2:3b, generated 27 days ago):

LLMs can make mistakes. Check important info.

Let's break down the benchmark and explain what's being tested.

**Benchmark Overview**

The benchmark compares the performance of three different serialization formats: MessagePack, JSON, and a custom implementation (MsgPack5). The test measures how quickly each format can serialize and deserialize complex data structures, such as arrays and objects.

**Test Data**

The test data consists of a single object with two properties: `a` (an array) and `b` (another object with multiple properties). This is repeated for each test run to generate multiple measurements.

**Serialization Formats**

There are three serialization formats being tested:

1. **JSON**: The built-in JavaScript Object Notation format, widely used for exchanging data between web servers and web applications.
2. **MessagePack**: A binary-based serialization format designed for efficient and compact data exchange. It's commonly used in games, real-time systems, and other high-performance applications.
3. **MsgPack5**: A custom implementation of MessagePack, optimized for performance and flexibility.

**Benchmark Results**

The benchmark results show the number of executions per second (e.g., "ExecutionsPerSecond") for each test run, measured on a Chrome 126 browser on a Windows desktop platform. The results indicate that:

* JSON is the fastest serialization format, with the highest executions per second.
* MessagePack and MsgPack5 are slower than JSON but closer in performance.

**Key Takeaways**

The benchmark highlights the trade-offs between different serialization formats:

* **JSON** is widely supported, easy to implement, and fast for simple data exchange. However, it may not be suitable for high-performance applications or systems with limited memory.
* **MessagePack** offers a good balance between compactness and performance, making it suitable for many use cases. However, its implementation can be more complex than JSON's.
* **MsgPack5** is an optimized custom implementation that aims to outperform MessagePack in certain scenarios. Its results are mixed, suggesting it may not offer significant benefits over the standard MessagePack implementation.

Keep in mind that this benchmark is specific to Chrome 126 on Windows desktop, and results might vary across other platforms, browsers, or architectures.

Latest run results:

Run details: (Test run date: 19 days ago)

User agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/130.0.0.0 Safari/537.36

Browser/OS: Chrome 130 on Windows

View result in a separate tab

Test name	Executions per second
MsgPack Strings Encode	256957.8 Ops/sec
JSON Strings Encode	1832978.6 Ops/sec
MsgPack Numbers Encode	291653.6 Ops/sec
JSON Numbers Encode	1054591.6 Ops/sec
JSON string decode	1104551.8 Ops/sec
msgpack string decode	375386.4 Ops/sec
MsgPack5 Strings Encode	16436.6 Ops/sec
MsgPack5 Numbers Encode	7314.8 Ops/sec
msgpack5 string decode	89817.2 Ops/sec

Related benchmarks:

msgpack-lite vs msgpack5 vs. JSON-decode (version: 0)

Comparing performance of: MsgPack Strings Encode vs JSON Strings Encode vs MsgPack Numbers Encode vs JSON Numbers Encode vs JSON string decode vs msgpack string decode vs MsgPack5 Strings Encode vs MsgPack5 Numbers Encode vs msgpack5 string decode

Created: 4 months ago by: Guest

Jump to the latest result

MsgPack Strings Encode

JSON Strings Encode

MsgPack Numbers Encode

JSON Numbers Encode

JSON string decode

msgpack string decode

MsgPack5 Strings Encode

MsgPack5 Numbers Encode

msgpack5 string decode

Suite status: <idle, ready to run>

Experimental features:

Fastest: N/A

Slowest: N/A

Autogenerated LLM Summary (model llama3.2:3b, generated 27 days ago):