neon-sunset / related_post_gen

Data Processing benchmark featuring Rust, Go, Swift, Zig etc.

Problem:

Given a list of posts, compute the top 5 related posts for each post based on the number of shared tags.

Steps:

• Read the posts JSON file.
• Iterate over the posts and populate a map containing: tag -> List<int>, with the int representing the post index of each post with that tag.
• Iterate over the posts and for each post:
  • Create a map: PostIndex -> int to track the number of shared tags
  • For each tag, Iterate over the posts that have that tag
  • For each post, increment the shared tag count in the map.
• Sort the related posts by the number of shared tags.
• Write the top 5 related posts for each post to a new JSON file.

Run Benchmark

./run.sh go | rust | python | all

# windows (powershell)
./run.ps1 go | rust | python | all

or
pwsh ./run.ps1 go | rust | python | all

Rules

No:

• FFI (including assembly inlining)
• Unsafe code blocks
• Custom benchmarking
• Disabling runtime checks (bounds etc)
• Specific hardware targeting

Must:

• Parse json at runtime
• Not hardcode number of posts
• Support up to 100 tags
• Use a stable release of the compiler/runtime

Updated Results from github workflow (raw data)

VM Specs

NB: The benchmark runs on the free tier of github workflow.

• CPU: 2 vCPUs
• RAM: 7GB
• OS: Ubuntu 22.04

Source

Language	Processing Time	Total (PT + I/O)
Go	26.39 ms	58.1 ms
Zig	38.00 ms	79.5 ms
Rust	38.91 ms	56.7 ms
Java (GraalVM)	40.00 ms	70.4 ms
Julia	42.67 ms	2.717 s
F#	45.86 ms	304.2 ms
Odin	46.52 ms	294.3 ms
Nim	56.00 ms	85.6 ms
Vlang	59.34 ms	400.6 ms
Swift	65.42 ms	442.3 ms
Crystal	68.53 ms	126.0 ms
C#	73.48 ms	283.7 ms
Dart VM	103.75 ms	573.1 ms
LuaJIT	118.00 ms	400.1 ms
Dart AOT	141.38 ms	283.8 ms
JS (Deno)	183.20 ms	272.4 ms
JS (Node)	202.20 ms	278.4 ms
Java (JIT)	256.69 ms	548.1 ms
Numpy	0.41 s	640.1 ms
JS (Bun)	764.40 ms	838.6 ms
Lua	2362.72 ms	3.037 s
Python	2.83 s	2.904 s

Multicore Results

Language	Processing Time	Total (PT + I/O)
Go Concurrent	18.89 ms	49.6 ms
Rust Concurrent	23.64 ms	41.8 ms
Swift Concurrent	40.02 ms	421.3 ms
F# Concurrent	40.57 ms	854.3 ms

Old Results with details (on my machine)

Language	Processing Time	Total (+ I/O)	Details
Rust	-	4.5s	Initial
Rust v2	-	2.60s	Replace std HashMap with fxHashMap by phazer99
Rust v3	-	1.28s	Preallocate and reuse map and unstable sort by vdrmn and Darksonn
Rust v4	-	0.13s	Use Post index as key instead of Pointer and Binary Heap by RB5009
Rust v5	38ms	52ms	Rm hashing from loop and use vec[count] instead of map[index]count by RB5009
Rust v6	23ms	36ms	Optimized Binary Heap Ops by scottlamb
Rust Rayon	9ms	22ms	Parallelize by masmullin2000
Rust Rayon	8ms	22ms	Remove comparison out of hot loop
⠀	⠀	⠀	⠀
Go	-	1.5s	Initial
Go v2	-	80ms	Add rust optimizations
Go v3	56ms	70ms	Use goccy/go-json
Go v3	34ms	55ms	Use generic binaryheap by DrBlury
Go v4	26ms	50ms	Replace binary heap with custom priority queue
Go v5	20ms	43ms	Remove comparison out of hot loop
Go Con	10ms	33ms	Go concurrency by tirprox and DrBlury
Go Con v2	5ms	29ms	Use arena, use waitgroup, rm binheap by DrBlury
⠀	⠀	⠀	⠀
Python	-	7.81s	Initial
Python v2	1.35s	1.53s	Add rust optimizations by dave-andersen
Numpy	0.57s	0.85s	Numpy implementation by Copper280z
⠀	⠀	⠀	⠀
Crystal	50ms	96ms	Inital w/ previous optimizations
Crystal v2	33ms	72ms	Replace binary heap with custom priority queue
⠀	⠀	⠀	⠀
Odin	110ms	397ms	Ported from golang code
Odin v2	104ms	404ms	Remove comparison out of hot loop
⠀	⠀	⠀	⠀
Dart VM	125ms	530ms	Ported frog golang code
Dart bin	274ms	360ms	Compiled executable
⠀	⠀	⠀	⠀
Vlang	339ms	560ms	Ported from golang code
⠀	⠀	⠀	⠀
Zig	80ms	110ms	Provided by akhildevelops