Benchmark

The current Benchmark tested a bunch of different iterations and access techniques. However, the most interesting one is the QueryBenchmark. In the benchmark, a set of entities was iterated over using the framework to access their transform and velocity and calculate a new position each iteration.

Their components looked like this:

public struct Transform{ float x; float y; float z; }
public struct Velocity { float x; float y; float z; }

The following performance was achieved with Arch for the scenario under heavy load and different amount of entities.

NET.7

Method	Amount	Mean	Error	StdDev	CacheMisses/Op	Allocated
WorldEntityQuery	10000	147.660 us	13.2838 us	0.7281 us	746	-
Query	10000	20.159 us	1.4188 us	0.0778 us	103	-
EntityQuery	10000	17.711 us	1.1311 us	0.0620 us	49	-
StructQuery	10000	7.767 us	0.1572 us	0.0086 us	7	-
StructEntityQuery	10000	7.338 us	1.7188 us	0.0942 us	12	-
WorldEntityQuery	100000	1,726.959 us	3,058.5935 us	167.6518 us	11,761	-
Query	100000	203.555 us	4.6038 us	0.2523 us	2,977	-
EntityQuery	100000	228.222 us	17.4030 us	0.9539 us	2,708	-
StructQuery	100000	115.466 us	8.8355 us	0.4843 us	2,726	-
StructEntityQuery	100000	76.823 us	2.1875 us	0.1199 us	2,544	-
WorldEntityQuery	1000000	20,419.798 us	4,491.2760 us	246.1820 us	90,624	-
Query	1000000	2,679.153 us	35.1696 us	1.9278 us	28,579	-
EntityQuery	1000000	2,462.296 us	322.4767 us	17.6760 us	28,113	-
StructQuery	1000000	1,514.479 us	296.5311 us	16.2539 us	29,723	-
StructEntityQuery	1000000	1,483.142 us	329.9446 us	18.0854 us	31,272	-

Legend

Legends

• amount : Value of the 'amount' parameter
• Mean : Arithmetic mean of all measurements
• Error : Half of 99.9% confidence interval
• StdDev : Standard deviation of all measurements
• Median : Value separating the higher half of all measurements (50th percentile)
• Allocated : Allocated memory per single operation (managed only, inclusive, 1KB = 1024B)
• 1 μs : 1 Microsecond (0.000001 sec)