Benchmark: TurboPFor Integer Compression on APPLE M1 #105
Description
icapp ZIPF
zipf alpha=1.50 range[0..255].n=1000000
bits histogram:
00:######################################## 40%
01:############## 14%
02:############# 13%
03:########## 10%
04:######## 7.7%
05:###### 5.7%
06:#### 4.1%
07:### 2.9%
08:## 2.1%
size ratio E MB/s D MB/s function integer size=32 bits (lz=,0) unsorted -1
633599 15.84% 1525.51 8032.58 1:p4nenc32 TurboPFor
633599 15.84% 1664.21 10515.40 2:p4nenc128v32 TurboPFor128
2312836 57.82% 1233.51 5328.54 4:p4ndenc32 TurboPFor delta
2312836 57.82% 1276.88 7617.02 5:p4ndenc128v32 TurboPFor128 delta
2919062 72.98% 947.46 3907.80 7:p4nd1enc32 TurboPFor delta1
2919062 72.98% 968.51 5663.58 8:p4nd1enc128v32 TurboPFor128 delta1
854911 21.37% 1576.86 5321.71 10:p4nzenc32 TurboPFor zigzag
854911 21.37% 1673.48 6746.09 11:p4nzenc128v32 TurboPFor128 zigzag
1006002 25.15% 1504.31 3995.13 13:p4nzzenc128v32 TurboPFor zzag/delta
999245 24.98% 20650.55 26847.12 20:bitnpack32 TurboPack
999245 24.98% 33450.13 53614.11 21:bitnpack128v32 TurboPack128
4007810 100.20% 15807.95 11967.66 23:bitndpack32 TurboPack delta
4007810 100.20% 22977.82 10905.65 24:bitndpack128v32 TurboPack128 delta
4007810 100.20% 14041.01 11226.43 26:bitnd1pack32 TurboPack delta1
4007810 100.20% 19744.03 7713.89 27:bitnd1pack128v32 TurboPack128 delta1
1123541 28.09% 7452.47 10869.93 29:bitnzpack32 TurboPack zigzag
1123541 28.09% 12002.27 11245.86 30:bitnzpack128v32 TurboPack128 zigzag
999390 24.98% 25560.31 11432.08 36:bitnxpack128v32 TurboPack128 xor
763425 19.09% 1253.75 3460.91 38:vsenc32 TurboVSimple
892659 22.32% 1157.96 2462.85 39:vszenc32 TurboVSimple zigzag
1014044 25.35% 12650.66 9266.57 40:vbenc32 TurboVByte scalar
1072600 26.82% 5488.71 4834.31 41:vbzenc32 TurboVByte zigzag
2622116 65.55% 1419.19 1308.16 42:vbdenc32 TurboVByte delta
3399564 84.99% 1312.15 1192.74 43:vbd1enc32 TurboVByte delta1
1178847 29.47% 2198.02 2365.85 44:vbddenc32 TurboVByte zzag delt
1250000 31.25% 16719.93 28492.57 45:v8enc32 TurboByte SIMD
2457230 61.43% 15023.68 11310.72 46:v8denc32 TurboByte delta
3040463 76.01% 14286.63 7841.42 47:v8d1enc32 TurboByte delta1
1250000 31.25% 15928.90 28202.45 48:v8xenc32 TurboByte xor
1286873 32.17% 13426.32 11363.56 49:v8zenc32 TurboByte zigzag
999245 24.98% 31691.54 53733.09 50:v8nenc128v32 TurboBitByte
1123541 28.09% 11768.05 11316.69 51:v8nzenc128v32 TurboBitByte zigzag
2465043 61.63% 10652.29 9210.60 52:v8ndenc128v32 TurboBitByte delta
3048273 76.21% 9734.20 6745.48 53:v8nd1enc128v32 TurboBitByte delta1
999390 24.98% 24531.59 11481.15 54:v8nxenc128v32 TurboBitByte xor
1192016 29.80% 1040.83 1204.86 60:bvzzenc32 Zigzag/delta bitio
983330 24.58% 1081.34 1169.94 61:bvzenc32 Zigzag bitio
954893 23.87% 1752.12 1619.43 62:fpgenc32 TurboGorilla bitio
1103692 27.59% 1640.11 1184.05 63:fphenc32 Chimp algo bitio
998253 24.96% 801.77 762.01 64:fpcenc32 TurboFloat LzX bitio
946461 23.66% 1013.91 6449.42 65:fpxenc32 TurboFloat XOR
946461 23.66% 828.29 1258.08 66:fpfcmenc32 TurboFloat FCM
3877388 96.93% 726.71 1100.40 67:fpdfcmenc32 TurboFloat DFCM
3882833 97.07% 723.76 1053.24 68:fp2dfcmenc32 TurboFloat DFCM 2D
2393070 59.83% 535.98 1727.24 70:trle TurboRLE
2633131 65.83% 457.12 1080.35 71:trlex TurboRLE xor
2633662 65.84% 362.66 1015.61 72:trlez TurboRLE zigzag
3844565 96.11% 1691.25 10432.85 73:srle32 TurboRLE ESC
3920520 98.01% 1451.95 5953.91 74:srlex32 TurboRLE ESC xor
3937690 98.44% 2065.79 5474.20 75:srlez32 TurboRLE ESC zigzag
1038578 25.96% 2908.18 3643.03 110:vlc TurboVLC bitio
1110239 27.76% 1236.72 972.80 111:vlcz TurboVLC zzag bitio
509274 12.73% 2128.08 1198.80 113:bitgenc Gamma coding bitio
645384 16.13% 1349.74 763.01 114:bitrenc Rice coding bitio
4000000 100.00% 41016.43 59778.60 117:tpenc Byte transpose
4000000 100.00% 15246.53 12684.75 118:tp4enc Nibble transpose
4000000 100.00% 55021.25 49051.97 120:memcpy memcpy
Best methods =113,2,1,114,38,10,11,39,66,65,62,61,64,50,21,20,36,54,13,40,110,41,63,111,30,29,51,44,60,48
Many thanks to Maris Ruskulis @fr3akX for providing the results of this benchmark