Cytoscape.js GPU benchmarks

v4 executor 'cpu' vs 'gpu' · times are per-call p50 · 2026-09-02 01:47 · 9ed49abc (v4) · chromium v24.18.0 · algorithms-gpu profile

CPU	Intel(R) Core(TM) i9-9900K CPU @ 3.60GHz
Cores	8 physical / 16 logical · 3.6 GHz base · 5 GHz max
Memory	62.6 GiB
GPU	CoffeeLake-S GT2 [UHD Graphics 630] i915 Ellesmere [Radeon RX 470/480/570/570X/580/580X/590] (8.0 GiB) amdgpu
Adapter	amd gcn-4
Browser	chromium 149.0.7827.55
OS	Fedora Linux 43 (Workstation Edition) · 7.1.10-100.fc43.x86_64 · x64
Node	v24.18.0 (v8 13.6.233.17-node.50)
Commit	9ed49abc docs(113): the performance review recorded — what the sweep measured and what it fixed
Machine id	5cf3f79c

Speedup overview cpu p50 ÷ gpu p50, log scale · right of the 1× line = gpu faster

heatKernel @ 1.024k

2222×

heatKernel @ 512

674×

markovClustering @ 1.024k

650×

markovClustering @ 512

301×

rwrProximity @ 1.024k

277×

kMedoids @ 4.096k

159×

simRank @ 1.024k

156×

rwrProximity @ 512

124×

markovClustering @ 256

115×

kMedoids @ 2.048k

109×

simRank @ 512

90×

heatKernel @ 256

76×

fuzzyCMeans @ 65.536k

69×

fuzzyCMeans @ 16.384k

51×

closenessCentralityNormalized @ 1.024k

49×

effectiveResistance @ 1.024k

43×

floydWarshall @ 1.024k

37×

rwrProximity @ 256

35×

simRank @ 256

33×

fuzzyCMeans @ 4.096k

30×

closenessCentralityNormalized @ 512

25×

motifCensus @ 2.048k

24×

kMedoids @ 1.024k

23×

kMeans @ 65.536k

22×

effectiveResistance @ 512

21×

betweennessCentrality @ 2.048k

18×

kMeans @ 16.384k

17×

floydWarshall @ 512

16×

betweennessCentrality @ 1.024k

10×

triangleCount @ 2.048k

9.0×

motifCensus @ 1.024k

7.9×

kMeans @ 4.096k

7.7×

effectiveResistance @ 256

7.4×

neighborhoodSimilarity @ 2.048k

6.6×

closenessCentralityNormalized @ 256

6.4×

triangleCount @ 1.024k

4.7×

betweennessCentrality @ 512

3.9×

neighborhoodSimilarity @ 1.024k

3.9×

affinityPropagation @ 512

3.7×

affinityPropagation @ 1.024k

3.7×

floydWarshall @ 256

3.6×

affinityPropagation @ 256

3.2×

triangleCount @ 512

2.3×

motifCensus @ 512

2.0×

neighborhoodSimilarity @ 512

2.00×

hierarchicalClustering @ 1.024k

1.01×

hierarchicalClustering @ 4.096k

0.96×

hierarchicalClustering @ 2.048k

0.94×

pageRankDense @ 1.024k

0.19×

pageRankDense @ 2.048k

0.19×

pageRankDense @ 512

0.15×

pageRank @ 512

0.06×

katzCentrality @ 512

0.05×

pageRank @ 1.024k

0.02×

katzCentrality @ 1.024k

0.02×

pageRank @ 2.048k

0.01×

katzCentrality @ 2.048k

0.01×

0.001×0.01×0.1×1×10×100×1000×10000×

Scaling speedup vs graph size

algorithms-gpu

group	× @ N=256	× @ N=512	× @ N=1.024k	× @ N=2.048k	× @ N=4.096k	× @ N=16.384k	× @ N=65.536k
markovClustering	115×	301×	650×	–	–	–	–
floydWarshall	3.6×	16×	37×	–	–	–	–
affinityPropagation	3.2×	3.7×	3.7×	–	–	–	–
closenessCentralityNormalized	6.4×	25×	49×	–	–	–	–
effectiveResistance	7.4×	21×	43×	–	–	–	–
simRank	33×	90×	156×	–	–	–	–
rwrProximity	35×	124×	277×	–	–	–	–
heatKernel	76×	674×	2222×	–	–	–	–
pageRank	–	0.06×	0.02×	0.01×	–	–	–
pageRankDense	–	0.15×	0.19×	0.19×	–	–	–
betweennessCentrality	–	3.9×	10×	18×	–	–	–
triangleCount	–	2.3×	4.7×	9.0×	–	–	–
neighborhoodSimilarity	–	2.00×	3.9×	6.6×	–	–	–
katzCentrality	–	0.05×	0.02×	0.01×	–	–	–
motifCensus	–	2.0×	7.9×	24×	–	–	–
kMedoids	–	–	23×	109×	159×	–	–
hierarchicalClustering	–	–	1.01×	0.94×	0.96×	–	–
kMeans	–	–	–	–	7.7×	17×	22×
fuzzyCMeans	–	–	–	–	30×	51×	69×

algorithms-gpu N=256 · 9.3 sv4 (gpu) v4 (cpu)

markovClustering

115×

floydWarshall

3.6×

affinityPropagation

3.2×

closenessCentralityNormalized

6.4×

effectiveResistance

7.4×

simRank

33×

rwrProximity

35×

heatKernel

76×

1 ms10 ms100 ms1 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
markovClustering	cpu	607 ms	612 ms	615 ms	603 ms	5
markovClustering	gpu	5.30 ms	5.80 ms	5.50 ms	5.20 ms	5
markovClustering	gpu first call	25.7 ms	25.7 ms	25.7 ms	25.7 ms	1
floydWarshall	cpu	23.2 ms	33.9 ms	49.8 ms	21.4 ms	5
floydWarshall	gpu	6.40 ms	6.50 ms	6.50 ms	6.40 ms	5
floydWarshall	gpu first call	20.6 ms	20.6 ms	20.6 ms	20.6 ms	1
affinityPropagation	cpu	40.6 ms	47.4 ms	41.8 ms	37.9 ms	5
affinityPropagation	gpu	12.8 ms	12.8 ms	13.0 ms	12.1 ms	5
affinityPropagation	gpu first call	25.2 ms	25.2 ms	25.2 ms	25.2 ms	1
closenessCentralityNormalized	cpu	23.2 ms	23.2 ms	23.3 ms	23.0 ms	5
closenessCentralityNormalized	gpu	3.60 ms	3.62 ms	3.70 ms	3.50 ms	5
closenessCentralityNormalized	gpu first call	7.30 ms	7.30 ms	7.30 ms	7.30 ms	1
effectiveResistance	cpu	52.4 ms	53.1 ms	52.9 ms	52.0 ms	5
effectiveResistance	gpu	7.10 ms	6.90 ms	7.40 ms	6.20 ms	5
effectiveResistance	gpu first call	14.0 ms	14.0 ms	14.0 ms	14.0 ms	1
simRank	cpu	147 ms	207 ms	149 ms	146 ms	5
simRank	gpu	4.40 ms	4.58 ms	4.90 ms	4.20 ms	5
simRank	gpu first call	13.0 ms	13.0 ms	13.0 ms	13.0 ms	1
rwrProximity	cpu	167 ms	168 ms	168 ms	166 ms	5
rwrProximity	gpu	4.80 ms	5.58 ms	4.90 ms	4.60 ms	5
rwrProximity	gpu first call	14.8 ms	14.8 ms	14.8 ms	14.8 ms	1
heatKernel	cpu	305 ms	306 ms	309 ms	303 ms	5
heatKernel	gpu	4.00 ms	4.06 ms	4.30 ms	3.80 ms	5
heatKernel	gpu first call	13.2 ms	13.2 ms	13.2 ms	13.2 ms	1

algorithms-gpu N=512 · 1.4 minv4 (gpu) v4 (cpu)

markovClustering

301×

pageRank

0.06×

pageRankDense

0.15×

floydWarshall

16×

betweennessCentrality

3.9×

affinityPropagation

3.7×

closenessCentralityNormalized

25×

triangleCount

2.3×

neighborhoodSimilarity

2.00×

katzCentrality

0.05×

effectiveResistance

21×

simRank

90×

rwrProximity

124×

heatKernel

674×

motifCensus

2.0×

100 µs1 ms10 ms100 ms1 s10 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
markovClustering	cpu	4.75 s	4.75 s	4.76 s	4.73 s	5
markovClustering	gpu	15.8 ms	15.7 ms	15.9 ms	14.9 ms	5
markovClustering	gpu first call	21.9 ms	21.9 ms	21.9 ms	21.9 ms	1
pageRank	cpu	400 µs	420 µs	400 µs	200 µs	5
pageRank	gpu	6.60 ms	6.72 ms	6.80 ms	6.40 ms	5
pageRank	gpu first call	13.4 ms	13.4 ms	13.4 ms	13.4 ms	1
pageRankDense	cpu	1.00 ms	960 µs	1.00 ms	900 µs	5
pageRankDense	gpu	6.70 ms	7.90 ms	8.00 ms	6.50 ms	5
pageRankDense	gpu first call	30.0 ms	30.0 ms	30.0 ms	30.0 ms	1
floydWarshall	cpu	182 ms	181 ms	183 ms	176 ms	5
floydWarshall	gpu	11.6 ms	11.9 ms	11.6 ms	11.4 ms	5
floydWarshall	gpu first call	13.7 ms	13.7 ms	13.7 ms	13.7 ms	1
betweennessCentrality	cpu	34.8 ms	35.0 ms	35.0 ms	34.6 ms	5
betweennessCentrality	gpu	8.90 ms	9.00 ms	9.00 ms	8.50 ms	5
betweennessCentrality	gpu first call	19.2 ms	19.2 ms	19.2 ms	19.2 ms	1
affinityPropagation	cpu	204 ms	311 ms	470 ms	198 ms	5
affinityPropagation	gpu	54.6 ms	54.2 ms	54.7 ms	53.4 ms	5
affinityPropagation	gpu first call	59.1 ms	59.1 ms	59.1 ms	59.1 ms	1
closenessCentralityNormalized	cpu	180 ms	174 ms	182 ms	162 ms	5
closenessCentralityNormalized	gpu	7.20 ms	7.46 ms	7.50 ms	7.00 ms	5
closenessCentralityNormalized	gpu first call	8.10 ms	8.10 ms	8.10 ms	8.10 ms	1
triangleCount	cpu	15.8 ms	16.0 ms	15.9 ms	15.3 ms	5
triangleCount	gpu	6.90 ms	6.92 ms	7.10 ms	6.60 ms	5
triangleCount	gpu first call	29.1 ms	29.1 ms	29.1 ms	29.1 ms	1
neighborhoodSimilarity	cpu	14.6 ms	14.5 ms	14.6 ms	14.0 ms	5
neighborhoodSimilarity	gpu	7.30 ms	7.34 ms	7.50 ms	7.00 ms	5
neighborhoodSimilarity	gpu first call	36.4 ms	36.4 ms	36.4 ms	36.4 ms	1
katzCentrality	cpu	300 µs	300 µs	300 µs	200 µs	5
katzCentrality	gpu	6.00 ms	6.02 ms	6.30 ms	5.60 ms	5
katzCentrality	gpu first call	10.8 ms	10.8 ms	10.8 ms	10.8 ms	1
effectiveResistance	cpu	411 ms	601 ms	892 ms	402 ms	5
effectiveResistance	gpu	20.0 ms	19.9 ms	20.4 ms	18.4 ms	5
effectiveResistance	gpu first call	19.3 ms	19.3 ms	19.3 ms	19.3 ms	1
simRank	cpu	1.07 s	1.51 s	1.08 s	1.05 s	5
simRank	gpu	11.9 ms	11.9 ms	12.0 ms	11.8 ms	5
simRank	gpu first call	29.7 ms	29.7 ms	29.7 ms	29.7 ms	1
rwrProximity	cpu	1.29 s	1.30 s	1.30 s	1.29 s	5
rwrProximity	gpu	10.4 ms	11.2 ms	10.5 ms	10.3 ms	5
rwrProximity	gpu first call	31.1 ms	31.1 ms	31.1 ms	31.1 ms	1
heatKernel	cpu	4.72 s	4.72 s	4.72 s	4.70 s	5
heatKernel	gpu	7.00 ms	7.80 ms	7.80 ms	6.80 ms	5
heatKernel	gpu first call	28.4 ms	28.4 ms	28.4 ms	28.4 ms	1
motifCensus	cpu	42.5 ms	51.4 ms	66.6 ms	39.6 ms	5
motifCensus	gpu	21.2 ms	19.9 ms	22.4 ms	12.8 ms	5
motifCensus	gpu first call	47.0 ms	47.0 ms	47.0 ms	47.0 ms	1

algorithms-gpu N=1,024 · 13.3 minv4 (gpu) v4 (cpu)

markovClustering

650×

pageRank

0.02×

pageRankDense

0.19×

floydWarshall

37×

betweennessCentrality

10×

affinityPropagation

3.7×

kMedoids

23×

hierarchicalClustering

1.01×

closenessCentralityNormalized

49×

triangleCount

4.7×

neighborhoodSimilarity

3.9×

katzCentrality

0.02×

effectiveResistance

43×

simRank

156×

rwrProximity

277×

heatKernel

2222×

motifCensus

7.9×

100 µs1 ms10 ms100 ms1 s10 s100 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
markovClustering	cpu	30.0 s	30.0 s	30.0 s	30.0 s	5
markovClustering	gpu	46.2 ms	47.9 ms	46.6 ms	46.0 ms	5
markovClustering	gpu first call	53.3 ms	53.3 ms	53.3 ms	53.3 ms	1
pageRank	cpu	400 µs	500 µs	600 µs	300 µs	5
pageRank	gpu	18.2 ms	18.4 ms	18.3 ms	17.8 ms	5
pageRank	gpu first call	20.2 ms	20.2 ms	20.2 ms	20.2 ms	1
pageRankDense	cpu	3.70 ms	3.70 ms	3.80 ms	3.60 ms	5
pageRankDense	gpu	19.3 ms	19.6 ms	19.3 ms	19.1 ms	5
pageRankDense	gpu first call	107 ms	107 ms	107 ms	107 ms	1
floydWarshall	cpu	1.40 s	1.41 s	1.40 s	1.40 s	5
floydWarshall	gpu	37.9 ms	42.9 ms	47.9 ms	35.5 ms	5
floydWarshall	gpu first call	40.0 ms	40.0 ms	40.0 ms	40.0 ms	1
betweennessCentrality	cpu	152 ms	155 ms	152 ms	151 ms	5
betweennessCentrality	gpu	15.2 ms	15.2 ms	15.3 ms	15.1 ms	5
betweennessCentrality	gpu first call	19.1 ms	19.1 ms	19.1 ms	19.1 ms	1
affinityPropagation	cpu	1.07 s	1.40 s	1.20 s	1.04 s	5
affinityPropagation	gpu	293 ms	293 ms	294 ms	289 ms	5
affinityPropagation	gpu first call	291 ms	291 ms	291 ms	291 ms	1
kMedoids	cpu	74.1 ms	73.9 ms	97.8 ms	38.3 ms	5
kMedoids	gpu	3.20 ms	3.26 ms	3.30 ms	3.10 ms	5
kMedoids	gpu first call	12.0 ms	12.0 ms	12.0 ms	12.0 ms	1
hierarchicalClustering	cpu	27.3 ms	29.5 ms	35.5 ms	24.1 ms	5
hierarchicalClustering	gpu	26.9 ms	26.9 ms	27.0 ms	26.7 ms	5
hierarchicalClustering	gpu first call	32.9 ms	32.9 ms	32.9 ms	32.9 ms	1
closenessCentralityNormalized	cpu	1.29 s	1.29 s	1.30 s	1.28 s	5
closenessCentralityNormalized	gpu	26.4 ms	27.3 ms	26.7 ms	26.2 ms	5
closenessCentralityNormalized	gpu first call	27.8 ms	27.8 ms	27.8 ms	27.8 ms	1
triangleCount	cpu	104 ms	154 ms	107 ms	104 ms	5
triangleCount	gpu	22.2 ms	22.6 ms	24.4 ms	20.4 ms	5
triangleCount	gpu first call	103 ms	103 ms	103 ms	103 ms	1
neighborhoodSimilarity	cpu	90.9 ms	137 ms	207 ms	88.8 ms	5
neighborhoodSimilarity	gpu	23.5 ms	23.3 ms	23.5 ms	23.0 ms	5
neighborhoodSimilarity	gpu first call	90.9 ms	90.9 ms	90.9 ms	90.9 ms	1
katzCentrality	cpu	200 µs	160 µs	200 µs	0.00 ns	5
katzCentrality	gpu	13.0 ms	13.2 ms	13.2 ms	12.9 ms	5
katzCentrality	gpu first call	14.0 ms	14.0 ms	14.0 ms	14.0 ms	1
effectiveResistance	cpu	3.39 s	3.41 s	3.42 s	3.39 s	5
effectiveResistance	gpu	79.7 ms	81.9 ms	84.3 ms	78.9 ms	5
effectiveResistance	gpu first call	81.9 ms	81.9 ms	81.9 ms	81.9 ms	1
simRank	cpu	8.37 s	8.37 s	8.37 s	8.34 s	5
simRank	gpu	53.7 ms	54.1 ms	54.3 ms	53.0 ms	5
simRank	gpu first call	129 ms	129 ms	129 ms	129 ms	1
rwrProximity	cpu	10.3 s	10.3 s	10.3 s	10.3 s	5
rwrProximity	gpu	37.2 ms	37.9 ms	38.1 ms	36.8 ms	5
rwrProximity	gpu first call	109 ms	109 ms	109 ms	109 ms	1
heatKernel	cpu	74.9 s	74.9 s	74.9 s	74.7 s	5
heatKernel	gpu	33.7 ms	34.2 ms	34.1 ms	33.2 ms	5
heatKernel	gpu first call	116 ms	116 ms	116 ms	116 ms	1
motifCensus	cpu	329 ms	408 ms	540 ms	314 ms	5
motifCensus	gpu	41.5 ms	41.7 ms	41.9 ms	41.3 ms	5
motifCensus	gpu first call	111 ms	111 ms	111 ms	111 ms	1

algorithms-gpu N=2,048 · 45.2 sv4 (gpu) v4 (cpu)

pageRank

0.01×

pageRankDense

0.19×

betweennessCentrality

18×

kMedoids

109×

hierarchicalClustering

0.94×

triangleCount

9.0×

neighborhoodSimilarity

6.6×

katzCentrality

0.01×

motifCensus

24×

100 µs1 ms10 ms100 ms1 s10 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
pageRank	cpu	600 µs	640 µs	600 µs	500 µs	5
pageRank	gpu	69.0 ms	69.9 ms	71.2 ms	67.9 ms	5
pageRank	gpu first call	83.1 ms	83.1 ms	83.1 ms	83.1 ms	1
pageRankDense	cpu	15.4 ms	15.5 ms	15.5 ms	15.3 ms	5
pageRankDense	gpu	80.5 ms	81.2 ms	82.3 ms	79.7 ms	5
pageRankDense	gpu first call	397 ms	397 ms	397 ms	397 ms	1
betweennessCentrality	cpu	635 ms	655 ms	706 ms	610 ms	5
betweennessCentrality	gpu	35.1 ms	35.4 ms	35.7 ms	34.7 ms	5
betweennessCentrality	gpu first call	39.1 ms	39.1 ms	39.1 ms	39.1 ms	1
kMedoids	cpu	381 ms	366 ms	437 ms	264 ms	5
kMedoids	gpu	3.50 ms	3.76 ms	3.60 ms	3.40 ms	5
kMedoids	gpu first call	4.60 ms	4.60 ms	4.60 ms	4.60 ms	1
hierarchicalClustering	cpu	98.2 ms	98.3 ms	98.8 ms	97.5 ms	5
hierarchicalClustering	gpu	105 ms	106 ms	109 ms	103 ms	5
hierarchicalClustering	gpu first call	131 ms	131 ms	131 ms	131 ms	1
triangleCount	cpu	765 ms	1.54 s	2.70 s	761 ms	5
triangleCount	gpu	85.1 ms	86.8 ms	85.1 ms	83.3 ms	5
triangleCount	gpu first call	398 ms	398 ms	398 ms	398 ms	1
neighborhoodSimilarity	cpu	602 ms	601 ms	602 ms	599 ms	5
neighborhoodSimilarity	gpu	91.7 ms	95.3 ms	95.4 ms	88.7 ms	5
neighborhoodSimilarity	gpu first call	406 ms	406 ms	406 ms	406 ms	1
katzCentrality	cpu	300 µs	280 µs	300 µs	200 µs	5
katzCentrality	gpu	34.9 ms	35.5 ms	36.5 ms	34.2 ms	5
katzCentrality	gpu first call	37.2 ms	37.2 ms	37.2 ms	37.2 ms	1
motifCensus	cpu	2.88 s	3.45 s	4.46 s	2.58 s	5
motifCensus	gpu	121 ms	121 ms	124 ms	109 ms	5
motifCensus	gpu first call	460 ms	460 ms	460 ms	460 ms	1

algorithms-gpu N=4,096 · 17.3 sv4 (gpu) v4 (cpu)

kMeans

7.7×

fuzzyCMeans

30×

kMedoids

159×

hierarchicalClustering

0.96×

1 ms10 ms100 ms1 s10 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
kMeans	cpu	30.8 ms	33.4 ms	33.9 ms	29.5 ms	5
kMeans	gpu	4.00 ms	4.18 ms	4.10 ms	3.60 ms	5
kMeans	gpu first call	12.6 ms	12.6 ms	12.6 ms	12.6 ms	1
fuzzyCMeans	cpu	153 ms	153 ms	154 ms	149 ms	5
fuzzyCMeans	gpu	5.10 ms	5.22 ms	5.30 ms	4.90 ms	5
fuzzyCMeans	gpu first call	17.8 ms	17.8 ms	17.8 ms	17.8 ms	1
kMedoids	cpu	1.75 s	1.82 s	1.97 s	1.60 s	5
kMedoids	gpu	11.0 ms	11.2 ms	11.6 ms	10.8 ms	5
kMedoids	gpu first call	12.0 ms	12.0 ms	12.0 ms	12.0 ms	1
hierarchicalClustering	cpu	422 ms	423 ms	427 ms	417 ms	5
hierarchicalClustering	gpu	439 ms	443 ms	439 ms	435 ms	5
hierarchicalClustering	gpu first call	469 ms	469 ms	469 ms	469 ms	1

Cytoscape.js GPU benchmarks

Speedup overview cpu p50 ÷ gpu p50, log scale · right of the 1× line = gpu faster

Scaling speedup vs graph size

algorithms-gpu

algorithms-gpu N=256 · 9.3 sv4 (gpu) v4 (cpu)

algorithms-gpu N=512 · 1.4 minv4 (gpu) v4 (cpu)

algorithms-gpu N=1,024 · 13.3 minv4 (gpu) v4 (cpu)

algorithms-gpu N=2,048 · 45.2 sv4 (gpu) v4 (cpu)

algorithms-gpu N=4,096 · 17.3 sv4 (gpu) v4 (cpu)

algorithms-gpu N=16,384 · 4.6 sv4 (gpu) v4 (cpu)

algorithms-gpu N=65,536 · 19.8 sv4 (gpu) v4 (cpu)

group	bench	p50	avg	p99	min	samples
kMeans	cpu	126 ms	126 ms	127 ms	124 ms	5
kMeans	gpu	7.50 ms	8.64 ms	9.70 ms	7.20 ms	5
kMeans	gpu first call	18.5 ms	18.5 ms	18.5 ms	18.5 ms	1
fuzzyCMeans	cpu	601 ms	605 ms	607 ms	595 ms	5
fuzzyCMeans	gpu	11.7 ms	11.6 ms	11.7 ms	10.1 ms	5
fuzzyCMeans	gpu first call	21.5 ms	21.5 ms	21.5 ms	21.5 ms	1

group	bench	p50	avg	p99	min	samples
kMeans	cpu	608 ms	615 ms	613 ms	597 ms	5
kMeans	gpu	27.1 ms	27.2 ms	27.3 ms	26.7 ms	5
kMeans	gpu first call	69.7 ms	69.7 ms	69.7 ms	69.7 ms	1
fuzzyCMeans	cpu	2.57 s	2.57 s	2.59 s	2.51 s	5
fuzzyCMeans	gpu	37.2 ms	37.6 ms	38.9 ms	35.0 ms	5
fuzzyCMeans	gpu first call	91.4 ms	91.4 ms	91.4 ms	91.4 ms	1