Faster.Map 6.1.0

There is a newer version of this package available.
See the version list below for details.
dotnet add package Faster.Map --version 6.1.0                
NuGet\Install-Package Faster.Map -Version 6.1.0                
This command is intended to be used within the Package Manager Console in Visual Studio, as it uses the NuGet module's version of Install-Package.
<PackageReference Include="Faster.Map" Version="6.1.0" />                
For projects that support PackageReference, copy this XML node into the project file to reference the package.
paket add Faster.Map --version 6.1.0                
#r "nuget: Faster.Map, 6.1.0"                
#r directive can be used in F# Interactive and Polyglot Notebooks. Copy this into the interactive tool or source code of the script to reference the package.
// Install Faster.Map as a Cake Addin
#addin nuget:?package=Faster.Map&version=6.1.0

// Install Faster.Map as a Cake Tool
#tool nuget:?package=Faster.Map&version=6.1.0                

Faster.Map

The goal of Faster.Map is to create a more efficient and performant alternative to the Dictionary and ConcurrentDictionary classes provided by .NET. These standard collections are widely used for key-value pair storage in .NET applications, but they have certain limitations, particularly in high-performance and concurrent scenarios.

Available Implementations:

  • DenseMap with SIMD Instructions: Harnesses SIMD (Single Instruction, Multiple Data) instructions for parallel processing, resulting in accelerated lookup times. Ideal for scenarios demanding high throughput and optimal CPU utilization.

  • RobinHoodMap with Linear Probing: Employs linear probing to resolve hash collisions, reducing the likelihood of clustering and improving access speed. Suitable for applications where a balance between performance and simplicity is required.

  • CMap is a high-performance, thread-safe, lockfree concurrent hashmap that uses open addressing, quadratic probing, and Fibonacci hashing to manage key-value pairs. The default load factor is set to 0.5, meaning the hash map will resize when it is half full. Note: this hashmap will only allocate once while resizing.

  • Installation:

You can include Faster.Map in your C# project via NuGet Package Manager:

Install-Package Faster.Map

How to use

DenseMap Example

// Example usage in C# (using DenseMap with SIMD Instructions)
using Faster.Map.DenseMap;

// Create a DenseMapSIMD object
var map = new DenseMap<string, string>();

// Add key-value pairs
map.Emplace("key1", "value1");

// Retrieve values
var result = map.Get("key1", out var retrievedValue);

Console.WriteLine(retrievedValue); // Output: "value1"

Tested on platforms:

  • x86
  • x64
  • arm
  • arm64

Benchmark

BenchmarkDotNet v0.13.12, Windows 11 (10.0.22631.3880/23H2/2023Update/SunValley3)
12th Gen Intel Core i5-12500H, 1 CPU, 16 logical and 12 physical cores
.NET SDK 9.0.100-preview.2.24157.14
  [Host]     : .NET 9.0.0 (9.0.24.12805), X64 RyuJIT AVX2
  DefaultJob : .NET 9.0.0 (9.0.24.12805), X64 RyuJIT AVX2

Retrieving a million pre-generated keys

Method Length Mean Error StdDev Ratio RatioSD Code Size Allocated Alloc Ratio
DenseMap 1000 0.001561 ms 0.0000298 ms 0.0000306 ms 0.55 0.01 312 B - NA
RobinhoodMap 1000 0.001453 ms 0.0000072 ms 0.0000063 ms 0.52 0.00 182 B - NA
Dictionary 1000 0.002817 ms 0.0000157 ms 0.0000123 ms 1.00 0.01 412 B - NA
DenseMap 10000 0.018443 ms 0.0001868 ms 0.0001560 ms 0.25 0.01 312 B - NA
RobinhoodMap 10000 0.048846 ms 0.0000682 ms 0.0000570 ms 0.65 0.02 182 B - NA
Dictionary 10000 0.074840 ms 0.0013760 ms 0.0034011 ms 1.00 0.06 412 B - NA
DenseMap 100000 0.302329 ms 0.0018030 ms 0.0016865 ms 0.28 0.00 289 B - 0.00
RobinhoodMap 100000 0.525929 ms 0.0020816 ms 0.0018453 ms 0.48 0.00 182 B 1 B 1.00
Dictionary 100000 1.091758 ms 0.0026861 ms 0.0022430 ms 1.00 0.00 393 B 1 B 1.00
DenseMap 400000 1.966361 ms 0.0391485 ms 0.0366195 ms 0.40 0.01 289 B 3 B 0.50
RobinhoodMap 400000 2.377139 ms 0.0238691 ms 0.0223271 ms 0.48 0.01 182 B 3 B 0.50
Dictionary 400000 4.901921 ms 0.0584053 ms 0.0517748 ms 1.00 0.01 412 B 6 B 1.00
DenseMap 900000 9.743361 ms 0.0631023 ms 0.0559385 ms 0.60 0.01 289 B 12 B 0.52
RobinhoodMap 900000 8.845731 ms 0.0784927 ms 0.0734222 ms 0.55 0.01 182 B 12 B 0.52
Dictionary 900000 16.156572 ms 0.1954894 ms 0.1632425 ms 1.00 0.01 412 B 23 B 1.00
DenseMap 1000000 11.216461 ms 0.1142519 ms 0.0954055 ms 0.65 0.01 289 B 12 B 0.52
RobinhoodMap 1000000 11.368053 ms 0.0775996 ms 0.0687900 ms 0.66 0.01 182 B 12 B 0.52
Dictionary 1000000 17.346513 ms 0.1926637 ms 0.1504191 ms 1.00 0.01 412 B 23 B 1.00

Adding a million keys

Method Length Mean Error StdDev Median Allocated
DenseMap 1000 0.04696 ms 0.00105 ms 0.00301 ms 0.04580 ms 784 B
RobinhoodMap 1000 0.05695 ms 0.00139 ms 0.00407 ms 0.05695 ms 448 B
Dictionary 1000 0.05925 ms 0.00302 ms 0.00886 ms 0.05710 ms 448 B
DenseMap 10000 0.10247 ms 0.00204 ms 0.00412 ms 0.10170 ms 784 B
RobinhoodMap 10000 0.11880 ms 0.00235 ms 0.00412 ms 0.11740 ms 784 B
Dictionary 10000 0.28064 ms 0.00560 ms 0.00872 ms 0.28010 ms 784 B
DenseMap 100000 0.74044 ms 0.01424 ms 0.01462 ms 0.73490 ms 784 B
RobinhoodMap 100000 1.09080 ms 0.02139 ms 0.03393 ms 1.08320 ms 448 B
Dictionary 100000 2.99360 ms 0.05919 ms 0.07269 ms 2.98355 ms 784 B
DenseMap 400000 5.01900 ms 0.34001 ms 0.99719 ms 5.46250 ms 784 B
RobinhoodMap 400000 4.83217 ms 0.10687 ms 0.30317 ms 4.76510 ms 784 B
Dictionary 400000 7.67889 ms 0.15666 ms 0.42620 ms 7.59280 ms 784 B
DenseMap 900000 10.63200 ms 0.20614 ms 0.18274 ms 10.60900 ms 784 B
RobinhoodMap 900000 14.09060 ms 0.28137 ms 0.78899 ms 14.05040 ms 784 B
Dictionary 900000 24.58053 ms 0.89271 ms 2.61815 ms 24.80390 ms 784 B
DenseMap 1000000 10.93189 ms 0.21667 ms 0.55930 ms 10.74225 ms 784 B
RobinhoodMap 1000000 17.13302 ms 0.37145 ms 1.05975 ms 16.94610 ms 784 B
Dictionary 1000000 28.22181 ms 0.84126 ms 2.45399 ms 27.72925 ms 784 B

Add and resize

Method Length Mean Error StdDev Median Gen0 Gen1 Gen2 Allocated
DenseMap 1000 0.06888 ms 0.002836 ms 0.008361 ms 0.06740 ms - - - 37.75 KB
RobinhoodMap 1000 0.08617 ms 0.003494 ms 0.010303 ms 0.08575 ms - - - 37.51 KB
Dictionary 1000 0.06571 ms 0.004702 ms 0.013865 ms 0.06270 ms - - - 72.09 KB
DenseMap 10000 0.19847 ms 0.003901 ms 0.007880 ms 0.19530 ms - - - 290.31 KB
RobinhoodMap 10000 0.34157 ms 0.006765 ms 0.007239 ms 0.33995 ms - - - 578.18 KB
Dictionary 10000 0.31703 ms 0.005693 ms 0.009668 ms 0.31630 ms - - - 657.93 KB
DenseMap 100000 1.23743 ms 0.019447 ms 0.015183 ms 1.23370 ms - - - 2306.88 KB
RobinhoodMap 100000 2.31568 ms 0.044450 ms 0.041579 ms 2.30430 ms - - - 4610.78 KB
Dictionary 100000 3.03312 ms 0.057396 ms 0.092684 ms 3.02825 ms - - - 5896.77 KB
DenseMap 400000 5.08197 ms 0.101222 ms 0.108306 ms 5.08500 ms - - - 9219.25 KB
RobinhoodMap 400000 9.40513 ms 0.186977 ms 0.337159 ms 9.31860 ms - - - 18435.23 KB
Dictionary 400000 11.62261 ms 0.232425 ms 0.595794 ms 11.54280 ms - - - 25374.92 KB
DenseMap 900000 12.07392 ms 0.238274 ms 0.222882 ms 12.07595 ms - - - 18435.44 KB
RobinhoodMap 900000 21.91546 ms 0.422548 ms 0.433926 ms 21.90840 ms - - - 36867.18 KB
Dictionary 900000 32.15119 ms 0.642537 ms 1.843559 ms 32.09600 ms 1000.0000 1000.0000 1000.0000 52626.84 KB
DenseMap 1000000 21.12683 ms 0.199052 ms 0.155406 ms 21.11290 ms - - - 36867.63 KB
RobinhoodMap 1000000 26.75000 ms 0.529428 ms 0.742185 ms 26.66140 ms - - - 36867.46 KB
Dictionary 1000000 36.88133 ms 0.731351 ms 1.510366 ms 36.99375 ms 1000.0000 1000.0000 1000.0000 52626.84 KB

Updating a million keys

Method Length Mean Error StdDev Median Allocated
DenseMap 1000 0.001497 ms 0.0000095 ms 0.0000080 ms 0.001497 ms -
RobinhoodMap 1000 0.001886 ms 0.0000369 ms 0.0000563 ms 0.001887 ms -
Dictionary 1000 0.003374 ms 0.0000594 ms 0.0000660 ms 0.003388 ms -
DenseMap 10000 0.018904 ms 0.0000967 ms 0.0000807 ms 0.018898 ms -
RobinhoodMap 10000 0.028991 ms 0.0001625 ms 0.0001520 ms 0.029036 ms -
Dictionary 10000 0.086054 ms 0.0001592 ms 0.0001329 ms 0.086069 ms -
DenseMap 100000 0.722426 ms 0.0139707 ms 0.0171572 ms 0.725357 ms 1 B
RobinhoodMap 100000 0.572789 ms 0.0107215 ms 0.0100289 ms 0.568764 ms 1 B
Dictionary 100000 1.195203 ms 0.0032038 ms 0.0029968 ms 1.196440 ms 1 B
DenseMap 400000 4.508174 ms 0.0646880 ms 0.0605092 ms 4.485129 ms 6 B
RobinhoodMap 400000 2.844626 ms 0.0536748 ms 0.0475813 ms 2.823570 ms 3 B
Dictionary 400000 5.396909 ms 0.0286768 ms 0.0268243 ms 5.405657 ms 6 B
DenseMap 900000 9.321427 ms 0.1539810 ms 0.1365001 ms 9.280112 ms 12 B
RobinhoodMap 900000 8.692224 ms 0.0805071 ms 0.0713675 ms 8.675028 ms 12 B
Dictionary 900000 16.036116 ms 0.1468814 ms 0.1226526 ms 16.054666 ms 23 B
DenseMap 1000000 10.349873 ms 0.1659302 ms 0.1385592 ms 10.316205 ms 12 B
RobinhoodMap 1000000 11.018986 ms 0.2149889 ms 0.5273715 ms 10.852050 ms 12 B
Dictionary 1000000 18.725914 ms 0.2099566 ms 0.1753233 ms 18.752681 ms 23 B

Removing a million keys

Method Length Mean Error StdDev Median Allocated
DenseMap 1000 0.02876 ms 0.001047 ms 0.003037 ms 0.02820 ms 736 B
RobinhoodMap 1000 0.04083 ms 0.001065 ms 0.003089 ms 0.03990 ms 736 B
Dictionary 1000 0.04365 ms 0.002059 ms 0.005973 ms 0.04250 ms 448 B
DenseMap 10000 0.05232 ms 0.000904 ms 0.000755 ms 0.05210 ms 736 B
RobinhoodMap 10000 0.07519 ms 0.001501 ms 0.002892 ms 0.07450 ms 736 B
Dictionary 10000 0.14230 ms 0.002837 ms 0.006686 ms 0.14110 ms 736 B
DenseMap 100000 0.41937 ms 0.030693 ms 0.087072 ms 0.38860 ms 736 B
RobinhoodMap 100000 0.79855 ms 0.023237 ms 0.067415 ms 0.82460 ms 64 B
Dictionary 100000 1.23970 ms 0.024222 ms 0.038419 ms 1.23260 ms 736 B
DenseMap 400000 3.47176 ms 0.099247 ms 0.286350 ms 3.39285 ms 736 B
RobinhoodMap 400000 4.69018 ms 0.105336 ms 0.308932 ms 4.70110 ms 736 B
Dictionary 400000 6.04550 ms 0.118591 ms 0.173829 ms 6.07010 ms 736 B
DenseMap 900000 10.99406 ms 0.212917 ms 0.493468 ms 10.93480 ms 736 B
RobinhoodMap 900000 12.97197 ms 0.358164 ms 1.050432 ms 12.51800 ms 736 B
Dictionary 900000 17.48864 ms 0.273846 ms 0.242757 ms 17.47670 ms 736 B
DenseMap 1000000 12.14540 ms 0.241315 ms 0.435141 ms 12.16020 ms 736 B
RobinhoodMap 1000000 14.51971 ms 0.284443 ms 0.252151 ms 14.44940 ms 736 B
Dictionary 1000000 19.87700 ms 0.305629 ms 0.285886 ms 19.80510 ms 736 B

Add and resize

Method Length Mean per Operation Error per Operation StdDev per Operation Median per Operation
DenseMap 1 660.9 ns 56.60 ns 159.6 ns 700.0 ns
RobinhoodMap 1 415.1 ns 53.24 ns 151.0 ns 400.0 ns
QuadMap 1 496.8 ns 58.76 ns 167.7 ns 450.0 ns
Dictionary 1 3,348.5 ns 577.13 ns 1,692.6 ns 3,100.0 ns
DenseMap 10 116.0 ns 9.99 ns 28.8 ns 110.0 ns
RobinhoodMap 10 491.9 ns 60.34 ns 174.1 ns 490.0 ns
QuadMap 10 508.1 ns 48.24 ns 134.5 ns 520.0 ns
Dictionary 10 551.6 ns 54.59 ns 158.4 ns 540.0 ns
DenseMap 100 121.8 ns 7.21 ns 20.3 ns 125.0 ns
RobinhoodMap 100 216.3 ns 17.88 ns 51.3 ns 229.0 ns
QuadMap 100 182.6 ns 12.62 ns 35.2 ns 187.0 ns
Dictionary 100 166.6 ns 15.54 ns 45.3 ns 174.0 ns
DenseMap 1000 94.5 ns 3.46 ns 9.5 ns 95.3 ns
RobinhoodMap 1000 118.8 ns 4.15 ns 11.6 ns 121.1 ns
QuadMap 1000 110.1 ns 3.74 ns 10.2 ns 110.9 ns
Dictionary 1000 96.1 ns 6.38 ns 18.5 ns 98.9 ns
DenseMap 10000 27.5 ns 0.55 ns 1.5 ns 27.4 ns
RobinhoodMap 10000 50.3 ns 1.51 ns 4.1 ns 50.2 ns
QuadMap 10000 50.3 ns 1.59 ns 4.5 ns 50.5 ns
Dictionary 10000 45.8 ns 1.18 ns 3.2 ns 45.5 ns
DenseMap 100000 16.3 ns 0.87 ns 2.5 ns 15.3 ns
RobinhoodMap 100000 32.1 ns 1.55 ns 4.6 ns 31.2 ns
QuadMap 100000 35.3 ns 2.24 ns 6.5 ns 33.3 ns
Dictionary 100000 33.1 ns 3.25 ns 9.5 ns 32.1 ns
DenseMap 1000000 28.0 ns 0.55 ns 0.7 ns 28.0 ns
RobinhoodMap 1000000 37.6 ns 0.93 ns 2.7 ns 36.7 ns
QuadMap 1000000 37.4 ns 1.11 ns 3.3 ns 36.0 ns
Dictionary 1000000 39.6 ns 0.77 ns 1.7 ns 39.7 ns

Add string benchmark

Method Length Mean Error StdDev Allocated
DenseMap 1000 0.009561 ms 0.0001123 ms 0.0000995 ms -
RobinhoodMap 1000 0.009880 ms 0.0001149 ms 0.0001075 ms -
Dictionary 1000 0.007242 ms 0.0000712 ms 0.0000595 ms -
DenseMap 10000 0.108277 ms 0.0012752 ms 0.0010649 ms -
RobinhoodMap 10000 0.127488 ms 0.0006820 ms 0.0006046 ms -
Dictionary 10000 0.140610 ms 0.0007416 ms 0.0006574 ms -
DenseMap 100000 1.527491 ms 0.0102525 ms 0.0090886 ms 1 B
RobinhoodMap 100000 1.971168 ms 0.0137025 ms 0.0128173 ms 3 B
Dictionary 100000 1.796192 ms 0.0107030 ms 0.0100116 ms 1 B
DenseMap 400000 15.195286 ms 0.0608646 ms 0.0539549 ms 12 B
RobinhoodMap 400000 20.500458 ms 0.1949961 ms 0.1628305 ms 23 B
Dictionary 400000 10.485305 ms 0.2214228 ms 0.6528703 ms 12 B
DenseMap 900000 51.631312 ms 0.6341017 ms 0.4950649 ms 74 B
RobinhoodMap 900000 68.940244 ms 1.0120403 ms 1.4187392 ms 8 B
Dictionary 900000 37.085791 ms 0.7275504 ms 1.0664340 ms 9 B
DenseMap 1000000 62.505928 ms 1.2287320 ms 2.4820994 ms 14 B
RobinhoodMap 1000000 84.237352 ms 1.6354160 ms 2.2926251 ms 123 B
Dictionary 1000000 45.689701 ms 0.9896159 ms 2.8867563 ms 67 B

Create StringWrapperBenchmark (cached hashcode)

Method Length Mean Error StdDev Allocated
DenseMap 1000 0.005590 ms 0.0000622 ms 0.0000582 ms -
RobinhoodMap 1000 0.004822 ms 0.0000862 ms 0.0000807 ms -
Dictionary 1000 0.006721 ms 0.0001277 ms 0.0001311 ms -
DenseMap 10000 0.072046 ms 0.0005074 ms 0.0004237 ms -
RobinhoodMap 10000 0.071678 ms 0.0010047 ms 0.0008390 ms -
Dictionary 10000 0.134088 ms 0.0004288 ms 0.0004011 ms -
DenseMap 100000 1.111280 ms 0.0174223 ms 0.0154444 ms 1 B
RobinhoodMap 100000 1.359501 ms 0.0153216 ms 0.0143318 ms 1 B
Dictionary 100000 1.866555 ms 0.0053967 ms 0.0045064 ms 1 B
DenseMap 400000 13.668025 ms 0.0525273 ms 0.0465641 ms 12 B
RobinhoodMap 400000 13.020727 ms 0.0614468 ms 0.0479736 ms 12 B
Dictionary 400000 10.846306 ms 0.0903438 ms 0.0800874 ms 12 B
DenseMap 900000 34.296921 ms 0.1181645 ms 0.0986727 ms 49 B
RobinhoodMap 900000 33.615793 ms 0.4644336 ms 0.4344315 ms 49 B
Dictionary 900000 41.753783 ms 0.8050352 ms 0.7906522 ms 4 B
DenseMap 1000000 40.145635 ms 0.7456945 ms 0.6975231 ms 57 B
RobinhoodMap 1000000 41.060782 ms 0.5800290 ms 0.5141806 ms 5 B
Dictionary 1000000 40.949160 ms 0.3808110 ms 0.3179945 ms 5 B
Product Compatible and additional computed target framework versions.
.NET net7.0 is compatible.  net7.0-android was computed.  net7.0-ios was computed.  net7.0-maccatalyst was computed.  net7.0-macos was computed.  net7.0-tvos was computed.  net7.0-windows was computed.  net8.0 is compatible.  net8.0-android was computed.  net8.0-browser was computed.  net8.0-ios was computed.  net8.0-maccatalyst was computed.  net8.0-macos was computed.  net8.0-tvos was computed.  net8.0-windows was computed.  net9.0 is compatible. 
Compatible target framework(s)
Included target framework(s) (in package)
Learn more about Target Frameworks and .NET Standard.

NuGet packages (2)

Showing the top 2 NuGet packages that depend on Faster.Map:

Package Downloads
Faster.Ioc

Package Description

UPSRestApi

UpsRestApi library for .Net used to access UPS API Catalog to browse products and view documentation.

GitHub repositories

This package is not used by any popular GitHub repositories.

Version Downloads Last updated
6.1.2 42 11/16/2024
6.1.1 93 10/31/2024
6.1.0 92 10/31/2024
6.0.4 116 10/13/2024
6.0.3 172 9/4/2024
6.0.2 107 9/3/2024
6.0.1 100 9/1/2024
6.0.0 130 8/7/2024
5.0.4 956 4/16/2024
5.0.3 489 11/18/2023
5.0.2 225 10/29/2023
5.0.1 141 10/29/2023
5.0.0 160 10/28/2023
4.2.1 308 9/2/2023
4.2.0 2,203 4/24/2023
4.1.6 245 4/24/2023
4.1.5 351 3/19/2023
4.1.4 341 3/14/2023
4.1.3 301 3/14/2023
4.1.2 272 3/14/2023
4.1.1 863 3/10/2023
4.1.0 303 3/9/2023
4.0.9 296 3/9/2023
4.0.8 293 3/9/2023
4.0.7 316 3/6/2023
4.0.6 344 3/3/2023
4.0.5 328 3/1/2023
4.0.4 410 2/24/2023
4.0.3 365 2/23/2023
4.0.2 1,149 1/7/2023
4.0.1 403 1/6/2023
4.0.0 384 1/4/2023
3.0.4.2 750 12/22/2022
3.0.3.9 1,097 12/9/2022
3.0.3.8 762 11/8/2022
3.0.3.7 523 10/27/2022
3.0.3.6 539 10/18/2022
3.0.3.5 438 10/18/2022
3.0.3.4 1,111 9/20/2022
3.0.3.3 1,262 5/19/2022
3.0.3.2 484 5/17/2022
3.0.3.1 477 5/17/2022
3.0.3 479 5/17/2022
3.0.2 494 5/15/2022

New Features:

     Enhanced emplace Behavior: The emplace function now operates with improved efficiency:
     Duplicate Management: When a duplicate entry is detected, emplace updates the existing entry instead of creating a new one.
     New Entry Insertion: If no duplicate is found, a new entry is added.

     Optimizations:

     Smart Tombstone Management: Introduced an optimized approach to manage and avoid tombstones, enhancing data integrity and reducing unnecessary overhead.