Cytoscape.js GPU benchmarks

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
OS	Fedora Linux 43 (Workstation Edition) · 6.19.14-200.fc43.x86_64 · x64
Node	v24.18.0 (v8 13.6.233.17-node.50)
Commit	0d3561de merge: the round 69/70 algorithm matrix into v4
Machine id	5cf3f79c

Speedup overview v3 p50 ÷ gpu p50, log scale · right of the 1× line = gpu faster

sweep: edges() @ 2k

293822×

sweep: nodes() @ 2k

181229×

core: edges() @ 2k

91943×

core: nodes() @ 2k

51330×

sweep: elements() @ 2k

32290×

core: elements() @ 2k

21595×

core: graph scratch() set + get (v3 restyles the graph) @ 2k

16693×

core: removeScratch() @ 2k

9502×

core: graph data() set + get (v3 restyles the graph) @ 2k

8305×

compare: same() @ 2k

8235×

trav: successors() @ 2k

756×

trav: predecessors() @ 2k

732×

mut-bulk: hide + show @ 2k

276×

flags: show + hide + visible @ 2k

226×

parent drag (shift subtree + bb settle) @ 2k

186×

compound: commonAncestors @ 2k

127×

degree: nodes.totalDegree() @ 2k

111×

mut-bulk: shift @ 2k

105×

viewport: reset() @ 2k

104×

degree: nodes.maxDegree() @ 2k

97×

flags: selectify + unselectify + selectable @ 2k

97×

96×

flags: grabify + ungrabify + grabbable @ 2k

95×

flags: panify + unpanify + pannable @ 2k

94×

flags: lock + unlock + locked @ 2k

93×

core: options() @ 2k

85×

degree: node.degree() @ 2k

82×

reparent round-trip @ 2k

78×

degree: indegree + outdegree (2 calls) @ 2k

76×

collection: same + anySame + contains (3 calls) @ 2k

73×

algo: closenessCentrality (one root) @ 2k

71×

algo: pageRank (20 iters) @ 500

70×

algo: closenessCentrality (one root) @ 500

67×

dimensions: shift + silentShift (round-trip) @ 2k

62×

61×

algo: bfs (whole graph) @ 500

60×

sweep: nodes(":selected") vs nodes({selected}) @ 2k

59×

mut-bulk: positions(obj) @ 2k

59×

viewport: viewport({ zoom, pan }) @ 2k

59×

algo: dfs (whole graph) @ 2k

58×

algo: dfs (whole graph) @ 500

57×

algo: dijkstra + pathTo @ 2k

55×

algo: bfs (whole graph) @ 2k

55×

algo: closenessCentrality (weighted, one root) @ 500

54×

mut-bulk: lock + unlock @ 2k

53×

child drag (move + parent re-derive) @ 2k

52×

algo: dijkstra + pathTo @ 500

51×

dimensions: positions( fn ) over 100 @ 2k

50×

collection: component() @ 2k

44×

degree: minDegree + maxDegree (2 calls) @ 2k

42×

traverse: components() @ 2k

42×

38×

algo: degreeCentralityNormalized @ 500

37×

algo: degreeCentralityNormalized @ 2k

37×

algo: degreeCentrality (one root) @ 500

36×

sweep: selected nodes ($("node:selected") vs filter({group, selected})) @ 2k

35×

algo: degreeCentrality (one root) @ 2k

34×

algo: tarjan SCC @ 2k

34×

algo: tarjan SCC @ 500

34×

query: structural, parents (:parent vs { parent: true }) @ 2k

33×

bounds: getFitViewport() (compute without committing) @ 2k

33×

bounds: cy.fit() (whole-graph scan) @ 2k

32×

query: structural, orphans (:orphan vs { child: false }) @ 2k

32×

collection: silentPositions( fn ) over 100 @ 2k

31×

collection: difference @ 2k

29×

collection: intersection @ 2k

28×

scn: select-all + fit (listeners on) @ 2k

27×

emit: position write, no listeners @ 2k

27×

set: difference() @ 2k

27×

load: init from the definition form (construct + dispose) @ 2k

27×

set: intersection() @ 2k

26×

26×

algo: kruskal @ 2k

26×

algo: hopcroft-tarjan biconnected @ 2k

26×

algo: kruskal @ 500

25×

algo: hopcroft-tarjan biconnected @ 500

25×

degree: minIndegree + maxOutdegree (2 calls) @ 2k

24×

degree: maxIndegree + minOutdegree (2 calls) @ 2k

24×

core: id lookup ($("#id") vs $id()) @ 2k

24×

trav: band roots() @ 2k

24×

mut-bulk: positions(fn) @ 2k

23×

mut: position set @ 2k

23×

data: bulk write, object column (plain-array fallback) @ 2k

21×

data: bulk write, numeric column @ 2k

21×

algo: bellmanFord @ 500

21×

algo: bellmanFord @ 2k

20×

20×

viewport: center( eles ) @ 2k

19×

data: bulk write, one new dictionary entry per pass @ 2k

18×

core: add + remove @ 2k

18×

data: bulk write, dictionary string column (4 values) @ 2k

16×

scn: drag (select band of 100 + 8-step shift, position listener) @ 2k

16×

core: forceRender() (headless no-op) @ 2k

16×

flags: select + unselect + selected (band) @ 2k

16×

collection: absoluteComplement @ 2k

16×

dimensions: silentPosition @ 2k

15×

query: membership ([bucket in ...] vs { in }) @ 2k

15×

flags: active/activate/unactivate (3 calls) @ 2k

15×

viewport: minZoom/maxZoom get (2 calls) @ 2k

15×

core: selected ($(":selected") vs filter({selected})) @ 2k

14×

style: renderedStyle (one prop) @ 2k

14×

query: data equality ([bucket = 3] vs { data: { bucket: 3 } }) @ 2k

14×

dimensions: padding + paddedWidth + paddedHeight (3 calls) @ 2k

13×

sweep: selected ($(":selected") vs filter({selected})) @ 2k

13×

mut-bulk: data set @ 2k

13×

query: structural, childless (:childless vs { parent: false }) @ 2k

13×

traversal: roots + leaves (2 calls) @ 2k

13×

12×

query: two keys AND-ed @ 2k

12×

collection: toArray() @ 2k

12×

algo: fuzzyCMeans @ 500

12×

algo: betweennessCentrality (unweighted) @ 2k

11×

algo: betweennessCentrality (unweighted) @ 500

11×

dimensions: relativePosition get @ 2k

11×

11×

query: data comparison ([weight > 3] vs { gt }) @ 2k

11×

9.9×

sweep: all nodes ($("node") vs filter({group})) @ 2k

9.9×

collection: removeScratch() @ 2k

8.8×

core: removeAllListeners() @ 2k

8.7×

core: batchData() @ 2k

8.7×

algo: betweennessCentrality (weighted) @ 500

8.6×

query: predicate function (both sides) @ 2k

8.6×

algo: hierarchicalClustering (threshold) @ 500

8.3×

emit: 1 core listener @ 2k

8.3×

emit: delegated (v3 selector vs v4 predicate) @ 2k

8.2×

data: removeData (whole collection) @ 2k

7.9×

mut-bulk: select + unselect @ 2k

7.9×

scn: explore (expand 2-hop + select + fit) @ 2k

7.7×

emit: 1 element (ref-qualified) listener @ 2k

7.7×

element data: scratch get + set @ 2k

7.6×

scn: edit (remove + re-add 256 nodes + cascade, add/remove listeners) @ 2k

7.3×

core: collection() @ 2k

7.1×

compound: orphans + nonorphans (2 calls) @ 2k

7.0×

core: collection() @ 2k

7.0×

core: batch( fn ) (empty) @ 2k

6.4×

emit: 10 core listeners @ 2k

6.2×

mut-bulk: remove + re-add (256 nodes + cascade) @ 2k

6.2×

style: whole-object style() @ 2k

6.2×

core: startBatch + endBatch + batching @ 2k

6.0×

style: bypass set + removeStyle round-trip @ 2k

5.9×

collection: union @ 2k

5.9×

bounds: whole-graph boundingBox() @ 2k

5.8×

algo: kMedoids @ 500

5.7×

style: sheet swap (compile + applyAll) @ 2k

5.5×

traversal: allAreNeighbors @ 2k

5.4×

traverse: incomers() @ 2k

5.4×

traverse: outgoers() @ 2k

5.3×

set: union() @ 2k

5.2×

algo: hierholzer (directed) @ 500

5.1×

trav: outgoers() @ 2k

5.1×

algo: affinityPropagation @ 500

5.0×

trav: band neighborhood() @ 2k

4.9×

core: animated() @ 2k

4.9×

core: nodes ($("node") vs filter({group})) @ 2k

4.9×

trav: incomers() @ 2k

4.8×

compare: contains() @ 2k

4.8×

compound: children + descendants + siblings (3 calls) @ 2k

4.7×

algo: kMeans (fixed centroids) @ 500

4.6×

viewport: zoom() get @ 2k

4.6×

collection: length + empty + nonempty (3 calls) @ 2k

4.6×

mut: select + unselect @ 2k

4.2×

style: bypass set (ele.style name,value) @ 2k

4.1×

trav: neighborhood() @ 2k

3.9×

collection: symmetricDifference @ 2k

3.9×

viewport: width() + height() + size() (3 calls) @ 2k

3.8×

traverse: neighborhood() @ 2k

3.7×

compound: parent + parents + ancestors (3 calls) @ 2k

3.7×

traversal: parallelEdges + codirectedEdges (2 calls) @ 2k

3.6×

collection: layoutDimensions() @ 2k

3.6×

algo: aStar @ 500

3.5×

collection: cy() @ 2k

3.5×

traversal: connectedNodes @ 2k

3.4×

algo: aStar @ 2k

3.3×

collection: diff @ 2k

3.3×

traversal: sources + targets (2 calls) @ 2k

3.2×

dimensions: rendered dims (4 calls) @ 2k

3.1×

collection: byGroup @ 2k

3.1×

anim: start + stop, 512-node collection @ 2k

3.0×

traversal: openNeighborhood + closedNeighborhood (2 calls) @ 2k

3.0×

algo: hierholzer (directed) @ 2k

3.0×

collection: animated() @ 2k

2.9×

dimensions: width + height + outerWidth + outerHeight (4 calls) @ 2k

2.9×

style: numericStyle @ 2k

2.9×

anim: viewport animate + stop @ 2k

2.9×

iter: map() @ 2k

2.8×

flags: takesUpSpace + interactive + transparent (3 calls) @ 2k

2.8×

collection: delayAnimation() @ 2k

2.8×

core: window() @ 2k

2.8×

traverse: connectedEdges() @ 2k

2.8×

trav: band connectedNodes() @ 2k

2.8×

mut: data set @ 2k

2.8×

trav: closedNeighborhood() @ 2k

2.8×

load: cy.json() @ 2k

2.7×

core: container() @ 2k

2.6×

trav: connectedEdges() @ 2k

2.5×

dimensions: renderedPosition get @ 2k

2.5×

algo: closenessCentralityNormalized @ 500

2.4×

core: zoom set @ 2k

2.4×

collection: is (none match: the full walk) @ 2k

2.4×

core: onRender + offRender @ 2k

2.4×

traversal: isLoop + isSimple (2 calls) @ 2k

2.3×

core: resize() / invalidateSize() @ 2k

2.3×

data: read the whole object @ 2k

2.3×

core: destroyed() @ 2k

2.2×

style: read one prop (width) @ 2k

2.2×

core: isReady() @ 2k

2.2×

flags: grabbed + hidden + inactive (3 calls) @ 2k

2.2×

collection: slice() @ 2k

2.2×

emit: on() + off() round-trip @ 2k

2.1×

flags: effectiveOpacity @ 2k

2.1×

element data: json() one element @ 2k

2.0×

core: selectionType get @ 2k

2.00×

style: first apply on add (256-node band) @ 2k

1.96×

collection: min + max ( fn ) (2 calls) @ 2k

1.96×

element data: jsons() over 100 @ 2k

1.93×

json: element @ 2k

1.89×

collection: eq/first/last (3 calls) @ 2k

1.88×

core: gating setters round-trip (autolock/autoungrabify/autounselectify) @ 2k

1.86×

viewport: pan( p ) set @ 2k

1.85×

viewport: zoom( n ) set @ 2k

1.84×

dimensions: renderedBoundingBox @ 2k

1.83×

scn: refresh (bulk data write + mapped labels + filter + fit, data listener) @ 2k

1.83×

iter: filter(fn) @ 2k

1.83×

algo: floydWarshall @ 500

1.77×

style: read one prop (color) @ 2k

1.77×

collection: indexOfId() @ 2k

1.76×

sweep: filter(fn) @ 2k

1.76×

style( 'text-wrap' ) (was #73 in the switch) @ 2k

1.76×

collection: element accessors (id/group/isNode/isEdge, 4 calls) @ 2k

1.74×

style( 'text-max-width' ) (was #74 in the switch) @ 2k

1.74×

load: add a 256-node band (add + remove) @ 2k

1.65×

viewport: panBy() @ 2k

1.63×

bounds: one node boundingBox() @ 2k

1.61×

collection: some + every (2 calls) @ 2k

1.59×

anim: delay() (no channels touched) @ 2k

1.58×

viewport: renderedExtent() @ 2k

1.58×

collection: sort( fn ) @ 2k

1.57×

compound: isParent/isChild/isChildless/isOrphan (4 calls) @ 2k

1.56×

viewport: extent() @ 2k

1.50×

trav: band sources() @ 2k

1.50×

collection: allAre (all match: the full walk) @ 2k

1.49×

position: get @ 2k

1.46×

trav: band connectedEdges() @ 2k

1.45×

collection: reduce( fn ) @ 2k

1.40×

traversal: source() + target() (2 calls) @ 2k

1.36×

core: mutableElements() (v4 materializes; v3 is O(1)) @ 2k

1.36×

collection: indexOf @ 2k

1.35×

data: read one string key (dictionary decode) @ 2k

1.35×

data: read one numeric key @ 2k

1.35×

core: pan() get (x32) @ 2k

1.32×

style( 'target-distance-from-node' ) (was #150 in the switch) @ 2k

1.29×

anim: start + stop, one element @ 2k

1.28×

style( 'taxi-radius' ) (was #142 in the switch) @ 2k

1.28×

traversal: edgesWith @ 2k

1.27×

traversal: edgesTo @ 2k

1.26×

core: instanceString + headless + styleEnabled (3 calls) @ 2k

1.25×

collection: removed() + inside() (2 calls) @ 2k

1.16×

core: gating getters (5 calls) @ 2k

1.15×

style( 'border-width' ) (was #6 in the switch) @ 2k

1.15×

core: hasElementWithId + hasCompoundNodes (2 calls) @ 2k

1.12×

viewport: pan() get @ 2k

1.10×

algo: markovClustering @ 500

1.10×

core: extent() @ 2k

1.09×

algo: kargerStein @ 500

1.07×

iter: forEach() @ 2k

1.05×

data: get @ 2k

1.05×

core: filter(fn) @ 2k

1.01×

core: getElementById() @ 2k

1.00×

style( 'background-color' ) (was #4 in the switch) @ 2k

0.98×

0.1×1×10×100×1000×10000×100000×1000000×

Scaling speedup vs graph size

algorithms

group	× @ N=500	× @ N=2k
algo: bfs (whole graph)	60×	55×
algo: dfs (whole graph)	57×	58×
algo: dijkstra + pathTo	51×	55×
algo: aStar	3.5×	3.3×
algo: bellmanFord	21×	20×
algo: kruskal	25×	26×
algo: tarjan SCC	34×	34×
algo: hopcroft-tarjan biconnected	25×	26×
algo: hierholzer (directed)	5.1×	3.0×
algo: betweennessCentrality (unweighted)	11×	11×
algo: degreeCentralityNormalized	37×	37×
algo: closenessCentrality (one root)	67×	71×
algo: degreeCentrality (one root)	36×	34×
algo: pageRank (20 iters)	70×	–
algo: closenessCentrality (weighted, one root)	54×	–
algo: floydWarshall	1.77×	–
algo: closenessCentralityNormalized	2.4×	–
algo: markovClustering	1.10×	–
algo: hierarchicalClustering (threshold)	8.3×	–
algo: kMeans (fixed centroids)	4.6×	–
algo: kMedoids	5.7×	–
algo: fuzzyCMeans	12×	–
algo: affinityPropagation	5.0×	–
algo: kargerStein	1.07×	–
algo: betweennessCentrality (weighted)	8.6×	–

core+collection N=2,000 · 2.9 minv4 (gpu) v3

core: getElementById()

1.00×

core: id lookup ($("#id") vs $id())

24×

core: nodes()

51330×

core: edges()

91943×

core: elements()

21595×

core: nodes ($("node") vs filter({group}))

4.9×

core: selected ($(":selected") vs filter({selected}))

14×

core: filter(fn)

1.01×

core: collection()

7.1×

core: pan() get (x32)

1.32×

core: extent()

1.09×

core: add + remove

18×

core: zoom set

2.4×

traverse: neighborhood()

3.7×

traverse: connectedEdges()

2.8×

traverse: outgoers()

5.3×

traverse: incomers()

5.4×

traverse: components()

42×

degree: node.degree()

82×

degree: nodes.totalDegree()

111×

degree: nodes.maxDegree()

97×

data: get

1.05×

position: get

1.46×

json: element

1.89×

set: union()

5.2×

set: intersection()

26×

set: difference()

27×

iter: map()

2.8×

iter: forEach()

1.05×

iter: filter(fn)

1.83×

compare: same()

8235×

compare: contains()

4.8×

mut: data set

2.8×

mut: position set

23×

mut: select + unselect

4.2×

1 ns10 ns100 ns1 µs10 µs100 µs1 ms10 ms100 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
core: getElementById()	v3	90.8 ns	93.7 ns	128 ns	87.4 ns	1830
core: getElementById()	gpu	91.2 ns	95.0 ns	157 ns	88.6 ns	1802
core: id lookup ($("#id") vs $id())	v3	2.03 µs	2.38 µs	3.63 µs	1.89 µs	67
core: id lookup ($("#id") vs $id())	gpu	84.5 ns	87.5 ns	161 ns	82.4 ns	1954
core: nodes()	v3	421 µs	436 µs	701 µs	415 µs	1604
core: nodes()	gpu	8.21 ns	8.72 ns	18.8 ns	8.12 ns	19429
core: edges()	v3	727 µs	751 µs	1.21 ms	694 µs	931
core: edges()	gpu	7.91 ns	8.36 ns	15.5 ns	7.72 ns	20283
core: elements()	v3	149 µs	159 µs	411 µs	145 µs	4375
core: elements()	gpu	6.89 ns	7.29 ns	17.5 ns	6.68 ns	23328
core: nodes ($("node") vs filter({group}))	v3	550 µs	576 µs	1.27 ms	539 µs	1212
core: nodes ($("node") vs filter({group}))	gpu	113 µs	121 µs	323 µs	82.7 µs	5706
core: selected ($(":selected") vs filter({selected}))	v3	362 µs	379 µs	726 µs	355 µs	1836
core: selected ($(":selected") vs filter({selected}))	gpu	25.0 µs	25.5 µs	25.3 µs	24.8 µs	12
core: filter(fn)	v3	424 µs	433 µs	583 µs	415 µs	1613
core: filter(fn)	gpu	419 µs	450 µs	872 µs	403 µs	1532
core: collection()	v3	458 ns	554 ns	1.57 µs	447 ns	302
core: collection()	gpu	64.8 ns	68.4 ns	133 ns	61.7 ns	2423
core: pan() get (x32)	v3	44.5 ns	46.9 ns	60.9 ns	42.7 ns	3633
core: pan() get (x32)	gpu	33.8 ns	35.4 ns	50.8 ns	32.8 ns	4829
core: extent()	v3	22.6 ns	40.6 ns	146 ns	21.5 ns	4101
core: extent()	gpu	20.6 ns	29.7 ns	116 ns	19.1 ns	5590
core: add + remove	v3	120 µs	125 µs	262 µs	111 µs	5565
core: add + remove	gpu	6.77 µs	7.97 µs	25.7 µs	6.38 µs	80653
core: zoom set	v3	1.37 µs	1.38 µs	1.53 µs	1.24 µs	117
core: zoom set	gpu	568 ns	575 ns	697 ns	546 ns	282
traverse: neighborhood()	v3	3.64 µs	3.82 µs	4.59 µs	3.42 µs	41
traverse: neighborhood()	gpu	972 ns	976 ns	1.17 µs	909 ns	162
traverse: connectedEdges()	v3	1.06 µs	1.16 µs	1.94 µs	1.02 µs	142
traverse: connectedEdges()	gpu	383 ns	409 ns	1.34 µs	373 ns	408
traverse: outgoers()	v3	2.39 µs	2.59 µs	3.56 µs	2.30 µs	62
traverse: outgoers()	gpu	451 ns	561 ns	2.03 µs	441 ns	289
traverse: incomers()	v3	2.41 µs	2.60 µs	3.46 µs	2.28 µs	62
traverse: incomers()	gpu	451 ns	532 ns	2.02 µs	441 ns	302
traverse: components()	v3	81.1 ms	83.5 ms	87.0 ms	75.9 ms	12
traverse: components()	gpu	1.94 ms	2.05 ms	3.33 ms	1.87 ms	326
degree: node.degree()	v3	3.41 µs	3.74 µs	4.98 µs	3.17 µs	42
degree: node.degree()	gpu	41.6 ns	42.9 ns	73.7 ns	39.4 ns	4001
degree: nodes.totalDegree()	v3	7.12 ms	7.83 ms	12.1 ms	6.79 ms	84
degree: nodes.totalDegree()	gpu	64.2 µs	66.8 µs	130 µs	61.0 µs	10313
degree: nodes.maxDegree()	v3	7.19 ms	8.03 ms	12.6 ms	6.74 ms	83
degree: nodes.maxDegree()	gpu	74.0 µs	75.1 µs	114 µs	69.9 µs	9335
data: get	v3	30.5 ns	31.5 ns	59.2 ns	29.5 ns	5430
data: get	gpu	29.1 ns	30.3 ns	44.5 ns	28.1 ns	5646
position: get	v3	31.9 ns	33.4 ns	72.0 ns	30.9 ns	5123
position: get	gpu	21.8 ns	24.7 ns	48.8 ns	21.0 ns	6745
json: element	v3	302 ns	396 ns	3.53 µs	295 ns	423
json: element	gpu	159 ns	165 ns	239 ns	155 ns	988
set: union()	v3	93.0 µs	103 µs	176 µs	90.0 µs	6732
set: union()	gpu	17.8 µs	18.2 µs	18.9 µs	17.3 µs	12
set: intersection()	v3	92.0 µs	108 µs	194 µs	89.3 µs	6398
set: intersection()	gpu	3.51 µs	3.55 µs	3.93 µs	3.40 µs	43
set: difference()	v3	91.7 µs	104 µs	179 µs	89.5 µs	6534
set: difference()	gpu	3.43 µs	3.49 µs	3.97 µs	3.35 µs	44
iter: map()	v3	29.0 µs	30.8 µs	77.5 µs	28.0 µs	22410
iter: map()	gpu	10.3 µs	10.3 µs	10.4 µs	10.2 µs	13
iter: forEach()	v3	3.49 µs	3.83 µs	7.92 µs	3.43 µs	35
iter: forEach()	gpu	3.32 µs	3.36 µs	3.85 µs	3.27 µs	45
iter: filter(fn)	v3	242 µs	249 µs	336 µs	240 µs	2812
iter: filter(fn)	gpu	133 µs	137 µs	179 µs	127 µs	5093
compare: same()	v3	124 µs	127 µs	199 µs	121 µs	5518
compare: same()	gpu	15.0 ns	15.9 ns	33.6 ns	14.3 ns	10737
compare: contains()	v3	6.12 µs	6.12 µs	6.25 µs	6.01 µs	25
compare: contains()	gpu	1.28 µs	1.29 µs	1.40 µs	1.25 µs	127
mut: data set	v3	642 ns	659 ns	783 ns	630 ns	254
mut: data set	gpu	233 ns	239 ns	364 ns	223 ns	700
mut: position set	v3	1.23 µs	1.25 µs	1.44 µs	1.20 µs	133
mut: position set	gpu	53.7 ns	57.0 ns	105 ns	52.7 ns	2915
mut: select + unselect	v3	2.61 µs	2.81 µs	3.60 µs	2.49 µs	56
mut: select + unselect	gpu	625 ns	634 ns	743 ns	609 ns	255

materializers N=2,000 · 50.0 sv4 (gpu) v3

sweep: nodes()

181229×

sweep: edges()

293822×

sweep: elements()

32290×

sweep: all nodes ($("node") vs filter({group}))

9.9×

sweep: selected ($(":selected") vs filter({selected}))

13×

sweep: selected nodes ($("node:selected") vs filter({group, selected}))

35×

sweep: nodes(":selected") vs nodes({selected})

59×

sweep: filter(fn)

1.76×

1 ns10 ns100 ns1 µs10 µs100 µs1 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
sweep: nodes()	v3	437 µs	454 µs	798 µs	421 µs	1542
sweep: nodes()	gpu	2.41 ns	2.55 ns	5.38 ns	2.30 ns	64642
sweep: edges()	v3	739 µs	757 µs	1.13 ms	719 µs	923
sweep: edges()	gpu	2.51 ns	2.62 ns	5.11 ns	2.46 ns	63345
sweep: elements()	v3	157 µs	163 µs	273 µs	153 µs	4282
sweep: elements()	gpu	4.85 ns	5.24 ns	12.2 ns	4.70 ns	32031
sweep: all nodes ($("node") vs filter({group}))	v3	574 µs	609 µs	924 µs	554 µs	1147
sweep: all nodes ($("node") vs filter({group}))	gpu	58.0 µs	72.7 µs	166 µs	53.9 µs	9534
sweep: selected ($(":selected") vs filter({selected}))	v3	351 µs	362 µs	568 µs	344 µs	1932
sweep: selected ($(":selected") vs filter({selected}))	gpu	26.4 µs	26.8 µs	27.1 µs	26.2 µs	12
sweep: selected nodes ($("node:selected") vs filter({group, selected}))	v3	338 µs	359 µs	609 µs	327 µs	1933
sweep: selected nodes ($("node:selected") vs filter({group, selected}))	gpu	9.52 µs	9.61 µs	9.86 µs	9.41 µs	14
sweep: nodes(":selected") vs nodes({selected})	v3	556 µs	572 µs	997 µs	541 µs	1218
sweep: nodes(":selected") vs nodes({selected})	gpu	9.37 µs	9.37 µs	9.67 µs	9.05 µs	15
sweep: filter(fn)	v3	432 µs	452 µs	824 µs	427 µs	1548
sweep: filter(fn)	gpu	246 µs	269 µs	561 µs	238 µs	2588

mutators N=2,000 · 56.3 sv4 (gpu) v3

mut-bulk: select + unselect

7.9×

mut-bulk: hide + show

276×

mut-bulk: lock + unlock

53×

mut-bulk: positions(obj)

59×

mut-bulk: positions(fn)

23×

mut-bulk: shift

105×

mut-bulk: data set

13×

mut-bulk: remove + re-add (256 nodes + cascade)

6.2×

10 µs100 µs1 ms10 ms100 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
mut-bulk: select + unselect	v3	2.19 ms	2.27 ms	3.76 ms	2.11 ms	288
mut-bulk: select + unselect	gpu	278 µs	290 µs	439 µs	273 µs	2417
mut-bulk: hide + show	v3	59.6 ms	64.1 ms	67.9 ms	57.6 ms	12
mut-bulk: hide + show	gpu	216 µs	230 µs	717 µs	210 µs	3019
mut-bulk: lock + unlock	v3	2.04 ms	2.13 ms	3.64 ms	1.96 ms	324
mut-bulk: lock + unlock	gpu	38.6 µs	39.0 µs	39.6 µs	38.2 µs	12
mut-bulk: positions(obj)	v3	2.21 ms	2.28 ms	3.49 ms	2.14 ms	302
mut-bulk: positions(obj)	gpu	37.5 µs	39.3 µs	71.0 µs	36.4 µs	17561
mut-bulk: positions(fn)	v3	3.40 ms	3.52 ms	5.01 ms	3.24 ms	192
mut-bulk: positions(fn)	gpu	146 µs	152 µs	247 µs	142 µs	4504
mut-bulk: shift	v3	3.41 ms	3.56 ms	4.59 ms	3.31 ms	187
mut-bulk: shift	gpu	32.3 µs	33.3 µs	57.6 µs	30.9 µs	20244
mut-bulk: data set	v3	867 µs	898 µs	1.52 ms	834 µs	775
mut-bulk: data set	gpu	65.6 µs	68.7 µs	134 µs	64.3 µs	10051
mut-bulk: remove + re-add (256 nodes + cascade)	v3	40.4 ms	44.2 ms	54.6 ms	38.6 ms	12
mut-bulk: remove + re-add (256 nodes + cascade)	gpu	6.47 ms	6.57 ms	8.37 ms	5.71 ms	100

scenarios N=2,000 · 37.9 sv4 (gpu) v3

scn: explore (expand 2-hop + select + fit)

7.7×

scn: select-all + fit (listeners on)

27×

scn: drag (select band of 100 + 8-step shift, position listener)

16×

scn: edit (remove + re-add 256 nodes + cascade, add/remove listeners)

7.3×

scn: refresh (bulk data write + mapped labels + filter + fit, data listener)

1.83×

10 µs100 µs1 ms10 ms100 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
scn: explore (expand 2-hop + select + fit)	v3	290 µs	339 µs	927 µs	275 µs	2048
scn: explore (expand 2-hop + select + fit)	gpu	37.5 µs	42.8 µs	177 µs	34.1 µs	15952
scn: select-all + fit (listeners on)	v3	25.1 ms	25.0 ms	26.4 ms	23.3 ms	23
scn: select-all + fit (listeners on)	gpu	922 µs	976 µs	1.51 ms	859 µs	691
scn: drag (select band of 100 + 8-step shift, position listener)	v3	1.64 ms	1.69 ms	2.23 ms	1.60 ms	397
scn: drag (select band of 100 + 8-step shift, position listener)	gpu	102 µs	105 µs	156 µs	98.0 µs	6293
scn: edit (remove + re-add 256 nodes + cascade, add/remove listeners)	v3	47.4 ms	47.3 ms	49.9 ms	44.6 ms	11
scn: edit (remove + re-add 256 nodes + cascade, add/remove listeners)	gpu	6.51 ms	6.60 ms	8.26 ms	6.00 ms	99
scn: refresh (bulk data write + mapped labels + filter + fit, data listener)	v3	7.74 ms	8.06 ms	10.4 ms	7.12 ms	80
scn: refresh (bulk data write + mapped labels + filter + fit, data listener)	gpu	4.22 ms	4.38 ms	6.98 ms	3.77 ms	148

traversal N=2,000 · 2.0 minv4 (gpu) v3

trav: neighborhood()

3.9×

trav: closedNeighborhood()

2.8×

trav: connectedEdges()

2.5×

trav: outgoers()

5.1×

trav: incomers()

4.8×

trav: successors()

756×

trav: predecessors()

732×

trav: band neighborhood()

4.9×

trav: band connectedEdges()

1.45×

trav: band roots()

24×

trav: band sources()

1.50×

trav: band connectedNodes()

2.8×

100 ns1 µs10 µs100 µs1 ms10 ms100 ms1 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
trav: neighborhood()	v3	3.99 µs	4.09 µs	5.26 µs	3.40 µs	37
trav: neighborhood()	gpu	1.02 µs	1.07 µs	2.02 µs	847 ns	154
trav: closedNeighborhood()	v3	5.87 µs	5.83 µs	6.26 µs	5.38 µs	26
trav: closedNeighborhood()	gpu	2.13 µs	2.31 µs	2.90 µs	2.08 µs	68
trav: connectedEdges()	v3	980 ns	1.06 µs	1.80 µs	943 ns	158
trav: connectedEdges()	gpu	390 ns	410 ns	928 ns	386 ns	411
trav: outgoers()	v3	2.33 µs	2.47 µs	3.05 µs	2.26 µs	65
trav: outgoers()	gpu	456 ns	536 ns	2.00 µs	442 ns	309
trav: incomers()	v3	2.33 µs	2.47 µs	3.04 µs	2.20 µs	65
trav: incomers()	gpu	482 ns	567 ns	2.12 µs	453 ns	292
trav: successors()	v3	439 ms	439 ms	445 ms	424 ms	12
trav: successors()	gpu	581 µs	626 µs	1.86 ms	559 µs	1105
trav: predecessors()	v3	437 ms	437 ms	448 ms	424 ms	12
trav: predecessors()	gpu	597 µs	632 µs	1.88 ms	583 µs	1073
trav: band neighborhood()	v3	186 µs	198 µs	267 µs	177 µs	3519
trav: band neighborhood()	gpu	37.8 µs	38.5 µs	40.9 µs	37.0 µs	12
trav: band connectedEdges()	v3	24.4 µs	25.0 µs	26.6 µs	24.2 µs	12
trav: band connectedEdges()	gpu	16.8 µs	17.4 µs	20.3 µs	16.6 µs	12
trav: band roots()	v3	108 µs	118 µs	174 µs	102 µs	5906
trav: band roots()	gpu	4.53 µs	4.55 µs	4.65 µs	4.47 µs	33
trav: band sources()	v3	6.67 µs	6.69 µs	6.99 µs	6.43 µs	22
trav: band sources()	gpu	4.45 µs	4.52 µs	5.07 µs	4.36 µs	33
trav: band connectedNodes()	v3	14.1 µs	14.5 µs	14.4 µs	13.9 µs	12
trav: band connectedNodes()	gpu	5.10 µs	5.14 µs	5.42 µs	4.99 µs	29

algorithms N=2,000 · 4.7 minv4 (gpu) v3

algo: bfs (whole graph)

55×

algo: dfs (whole graph)

58×

algo: dijkstra + pathTo

55×

algo: aStar

3.3×

algo: bellmanFord

20×

algo: kruskal

26×

algo: tarjan SCC

34×

algo: hopcroft-tarjan biconnected

26×

algo: hierholzer (directed)

3.0×

algo: betweennessCentrality (unweighted)

11×

algo: degreeCentralityNormalized

37×

algo: closenessCentrality (one root)

71×

algo: degreeCentrality (one root)

34×

100 ns1 µs10 µs100 µs1 ms10 ms100 ms1 s10 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
algo: bfs (whole graph)	v3	52.3 ms	53.4 ms	56.5 ms	50.3 ms	9
algo: bfs (whole graph)	gpu	948 µs	970 µs	1.43 ms	890 µs	713
algo: dfs (whole graph)	v3	50.3 ms	50.7 ms	52.2 ms	49.5 ms	10
algo: dfs (whole graph)	gpu	868 µs	899 µs	1.44 ms	847 µs	764
algo: dijkstra + pathTo	v3	53.2 ms	53.5 ms	54.3 ms	52.6 ms	10
algo: dijkstra + pathTo	gpu	961 µs	975 µs	1.22 ms	934 µs	705
algo: aStar	v3	1.70 ms	1.75 ms	2.44 ms	1.67 ms	396
algo: aStar	gpu	507 µs	529 µs	898 µs	492 µs	1319
algo: bellmanFord	v3	236 ms	235 ms	254 ms	215 ms	12
algo: bellmanFord	gpu	11.8 ms	11.9 ms	12.5 ms	11.6 ms	55
algo: kruskal	v3	13.3 ms	14.2 ms	17.3 ms	12.8 ms	45
algo: kruskal	gpu	510 µs	523 µs	789 µs	467 µs	1325
algo: tarjan SCC	v3	23.1 ms	24.3 ms	28.0 ms	21.6 ms	25
algo: tarjan SCC	gpu	672 µs	848 µs	5.88 ms	633 µs	798
algo: hopcroft-tarjan biconnected	v3	88.3 ms	89.4 ms	98.3 ms	82.7 ms	12
algo: hopcroft-tarjan biconnected	gpu	3.44 ms	3.56 ms	4.67 ms	3.32 ms	189
algo: hierholzer (directed)	v3	23.1 ms	23.6 ms	25.9 ms	21.7 ms	26
algo: hierholzer (directed)	gpu	7.79 ms	7.87 ms	8.71 ms	7.66 ms	76
algo: betweennessCentrality (unweighted)	v3	4.08 s	4.09 s	4.12 s	4.01 s	12
algo: betweennessCentrality (unweighted)	gpu	358 ms	376 ms	443 ms	349 ms	12
algo: degreeCentralityNormalized	v3	14.6 ms	15.4 ms	18.3 ms	13.7 ms	41
algo: degreeCentralityNormalized	gpu	392 µs	408 µs	750 µs	379 µs	1711
algo: closenessCentrality (one root)	v3	60.4 ms	64.0 ms	77.3 ms	59.8 ms	12
algo: closenessCentrality (one root)	gpu	848 µs	871 µs	1.64 ms	823 µs	796
algo: degreeCentrality (one root)	v3	7.43 µs	7.27 µs	7.92 µs	6.71 µs	19
algo: degreeCentrality (one root)	gpu	216 ns	220 ns	256 ns	214 ns	751

algorithms N=500 · 10.3 minv4 (gpu) v3

algo: bfs (whole graph)

60×

algo: dfs (whole graph)

57×

algo: dijkstra + pathTo

51×

algo: aStar

3.5×

algo: bellmanFord

21×

algo: kruskal

25×

algo: tarjan SCC

34×

algo: hopcroft-tarjan biconnected

25×

algo: hierholzer (directed)

5.1×

algo: betweennessCentrality (unweighted)

11×

algo: degreeCentralityNormalized

37×

algo: closenessCentrality (one root)

67×

algo: degreeCentrality (one root)

36×

algo: pageRank (20 iters)

70×

algo: closenessCentrality (weighted, one root)

54×

algo: floydWarshall

1.77×

algo: closenessCentralityNormalized

2.4×

algo: markovClustering

1.10×

algo: hierarchicalClustering (threshold)

8.3×

algo: kMeans (fixed centroids)

4.6×

algo: kMedoids

5.7×

algo: fuzzyCMeans

12×

algo: affinityPropagation

5.0×

algo: kargerStein

1.07×

algo: betweennessCentrality (weighted)

8.6×

100 ns1 µs10 µs100 µs1 ms10 ms100 ms1 s10 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
algo: bfs (whole graph)	v3	12.7 ms	13.8 ms	18.5 ms	11.5 ms	47
algo: bfs (whole graph)	gpu	209 µs	216 µs	323 µs	203 µs	3234
algo: dfs (whole graph)	v3	11.5 ms	12.4 ms	15.9 ms	10.9 ms	52
algo: dfs (whole graph)	gpu	203 µs	210 µs	282 µs	198 µs	3322
algo: dijkstra + pathTo	v3	12.2 ms	12.9 ms	17.4 ms	11.5 ms	50
algo: dijkstra + pathTo	gpu	241 µs	248 µs	332 µs	229 µs	2820
algo: aStar	v3	443 µs	456 µs	621 µs	423 µs	1535
algo: aStar	gpu	125 µs	131 µs	199 µs	119 µs	5353
algo: bellmanFord	v3	17.1 ms	17.3 ms	18.1 ms	16.9 ms	37
algo: bellmanFord	gpu	814 µs	841 µs	2.04 ms	785 µs	826
algo: kruskal	v3	3.24 ms	3.54 ms	6.90 ms	3.16 ms	194
algo: kruskal	gpu	128 µs	135 µs	237 µs	124 µs	5152
algo: tarjan SCC	v3	5.39 ms	5.92 ms	9.42 ms	5.29 ms	113
algo: tarjan SCC	gpu	158 µs	174 µs	344 µs	151 µs	3916
algo: hopcroft-tarjan biconnected	v3	20.3 ms	21.3 ms	24.5 ms	19.0 ms	29
algo: hopcroft-tarjan biconnected	gpu	817 µs	845 µs	1.24 ms	800 µs	825
algo: hierholzer (directed)	v3	4.21 ms	4.50 ms	7.55 ms	4.13 ms	145
algo: hierholzer (directed)	gpu	823 µs	850 µs	1.22 ms	796 µs	810
algo: betweennessCentrality (unweighted)	v3	258 ms	260 ms	264 ms	256 ms	12
algo: betweennessCentrality (unweighted)	gpu	24.0 ms	24.2 ms	25.7 ms	23.2 ms	25
algo: degreeCentralityNormalized	v3	3.56 ms	3.90 ms	7.95 ms	3.42 ms	175
algo: degreeCentralityNormalized	gpu	95.3 µs	98.6 µs	178 µs	93.7 µs	7087
algo: closenessCentrality (one root)	v3	13.6 ms	14.5 ms	18.0 ms	13.1 ms	43
algo: closenessCentrality (one root)	gpu	204 µs	212 µs	382 µs	196 µs	3302
algo: degreeCentrality (one root)	v3	7.64 µs	7.54 µs	8.45 µs	6.77 µs	19
algo: degreeCentrality (one root)	gpu	213 ns	218 ns	271 ns	205 ns	780
algo: pageRank (20 iters)	v3	2.74 ms	2.84 ms	6.20 ms	2.65 ms	236
algo: pageRank (20 iters)	gpu	39.2 µs	44.4 µs	171 µs	38.4 µs	15613
algo: closenessCentrality (weighted, one root)	v3	13.9 ms	14.7 ms	18.1 ms	13.1 ms	44
algo: closenessCentrality (weighted, one root)	gpu	257 µs	262 µs	332 µs	252 µs	2667
algo: floydWarshall	v3	248 ms	272 ms	384 ms	245 ms	12
algo: floydWarshall	gpu	140 ms	141 ms	143 ms	138 ms	12
algo: closenessCentralityNormalized	v3	272 ms	274 ms	277 ms	269 ms	12
algo: closenessCentralityNormalized	gpu	111 ms	119 ms	144 ms	111 ms	12
algo: markovClustering	v3	4.13 s	4.38 s	4.65 s	4.08 s	12
algo: markovClustering	gpu	3.77 s	3.87 s	3.78 s	3.73 s	12
algo: hierarchicalClustering (threshold)	v3	34.7 ms	34.9 ms	36.2 ms	34.4 ms	16
algo: hierarchicalClustering (threshold)	gpu	4.18 ms	4.34 ms	5.65 ms	4.12 ms	157
algo: kMeans (fixed centroids)	v3	9.00 ms	9.07 ms	9.83 ms	8.72 ms	73
algo: kMeans (fixed centroids)	gpu	1.94 ms	2.01 ms	3.11 ms	1.89 ms	345
algo: kMedoids	v3	232 ms	298 ms	436 ms	72.4 ms	12
algo: kMedoids	gpu	40.5 ms	36.8 ms	44.8 ms	16.0 ms	16
algo: fuzzyCMeans	v3	78.6 ms	79.7 ms	81.5 ms	78.3 ms	12
algo: fuzzyCMeans	gpu	6.79 ms	6.86 ms	7.85 ms	6.58 ms	99
algo: affinityPropagation	v3	209 ms	213 ms	219 ms	201 ms	12
algo: affinityPropagation	gpu	41.5 ms	48.4 ms	44.7 ms	40.1 ms	9
algo: kargerStein	v3	321 ms	327 ms	356 ms	307 ms	12
algo: kargerStein	gpu	301 ms	307 ms	328 ms	295 ms	12
algo: betweennessCentrality (weighted)	v3	712 ms	714 ms	723 ms	703 ms	12
algo: betweennessCentrality (weighted)	gpu	83.2 ms	84.7 ms	86.6 ms	82.0 ms	12

mappers N=2,000 · 23.9 s

mapper: bulk write, color — cpu eval

1.31 ms

mapper: bulk write, color — gpu path

175 µs

mapper: bulk write, color + size — cpu eval

1.59 ms

mapper: bulk write, color + size — gpu color / cpu size

1.53 ms

mapper: bulk write, mapped labels (round-5 path) — label refresh

4.10 ms

100 µs1 ms10 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
mapper: bulk write, color	cpu eval	1.31 ms	1.35 ms	2.00 ms	1.29 ms	511
mapper: bulk write, color	gpu path	175 µs	183 µs	314 µs	172 µs	3820
mapper: bulk write, color + size	cpu eval	1.59 ms	1.61 ms	1.85 ms	1.56 ms	432
mapper: bulk write, color + size	gpu color / cpu size	1.53 ms	1.57 ms	2.41 ms	1.49 ms	443
mapper: bulk write, mapped labels (round-5 path)	label refresh	4.10 ms	4.31 ms	6.96 ms	3.50 ms	157

layouts N=2,000 · 1.5 minv4 (gpu) v3

96×

20×

61×

12×

26×

38×

11×

113 ms

layout: force seed — spectral vs scatter (20 iterations) — spectral (landmark MDS, the default)

113 ms

56.8 ms

370 µs

31.5 µs

528 ns

layout: contract fixed cost (empty impl vs bulk placement) — wrapper + bulk write

39.4 µs

9.9×

100 ns1 µs10 µs100 µs1 ms10 ms100 ms1 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
layout: grid	v3	12.3 ms	12.7 ms	15.2 ms	10.5 ms	50
layout: grid	gpu	128 µs	140 µs	514 µs	79.1 µs	4720
layout: preset	v3	5.80 ms	6.09 ms	8.01 ms	5.45 ms	105
layout: preset	gpu	293 µs	321 µs	732 µs	287 µs	2159
layout: preset (fn form)	v3	5.97 ms	6.30 ms	9.30 ms	5.43 ms	106
layout: preset (fn form)	gpu	98.4 µs	108 µs	375 µs	95.6 µs	6372
layout: circle	v3	9.84 ms	10.3 ms	15.2 ms	9.20 ms	63
layout: circle	gpu	795 µs	899 µs	2.80 ms	771 µs	753
layout: concentric	v3	31.9 ms	31.6 ms	33.5 ms	29.0 ms	17
layout: concentric	gpu	1.22 ms	1.45 ms	4.75 ms	1.17 ms	466
layout: breadthfirst	v3	414 ms	422 ms	455 ms	398 ms	12
layout: breadthfirst	gpu	11.0 ms	11.5 ms	17.0 ms	10.0 ms	54
layout: random	v3	5.74 ms	6.18 ms	10.1 ms	5.42 ms	108
layout: random	gpu	533 µs	587 µs	1.53 ms	515 µs	1133
layout: force (CPU executor, 20 iterations)	gpu	113 ms	121 ms	140 ms	102 ms	12
layout: force seed — spectral vs scatter (20 iterations)	spectral (landmark MDS, the default)	113 ms	112 ms	114 ms	103 ms	12
layout: force seed — spectral vs scatter (20 iterations)	scatter	56.8 ms	59.8 ms	58.0 ms	56.0 ms	12
layout: plumbing (layoutPositions vs setPositions)	finisher	370 µs	406 µs	947 µs	360 µs	1672
layout: plumbing (layoutPositions vs setPositions)	bulk	31.5 µs	42.0 µs	255 µs	29.9 µs	15839
layout: contract fixed cost (empty impl vs bulk placement)	wrapper only	528 ns	972 ns	4.08 µs	520 ns	161
layout: contract fixed cost (empty impl vs bulk placement)	wrapper + bulk write	39.4 µs	41.4 µs	51.3 µs	37.5 µs	12
layout: scope (eles.layout, 10% of nodes)	v3	998 µs	1.08 ms	2.20 ms	965 µs	633
layout: scope (eles.layout, 10% of nodes)	gpu	101 µs	119 µs	272 µs	97.2 µs	5809

style N=2,000 · 1.7 minv4 (gpu) v3

style: sheet swap (compile + applyAll)

5.5×

style: compile vs compile + apply — compile only (in batch)

25.1 µs

style: compile vs compile + apply — compile + apply

2.30 ms

style: first apply on add (256-node band)

1.96×

style: applyAll, flat vs compound (100 parents — the 14.6 partition) — flat

2.30 ms

style: applyAll, flat vs compound (100 parents — the 14.6 partition) — compound

2.73 ms

style: applyAll — constant vs state-case (the 57.1d partition) — constant sheet

2.26 ms

style: applyAll — constant vs state-case (the 57.1d partition) — state-case sheet

2.28 ms

style: applyAll — state-case vs data-case (what the partition avoids) — state-case sheet

2.27 ms

style: applyAll — state-case vs data-case (what the partition avoids) — data-case sheet

10.0 ms

style: select+unselect 256-band — constant vs state-case sheet — constant sheet (skips restyle)

6.79 µs

style: select+unselect 256-band — constant vs state-case sheet — paint-only state sheet (round-61 diff path)

32.3 µs

style: select+unselect 256-band — constant vs state-case sheet — state-case sheet (all channels; full write)

376 µs

style: read one prop (color)

1.77×

style: read one prop (width)

2.2×

style: numericStyle

2.9×

style: renderedStyle (one prop)

14×

style: whole-object style()

6.2×

style: bypass set (ele.style name,value)

4.1×

style: bypass set + removeStyle round-trip

5.9×

style: select+unselect 256-band under the default sheet — bypass punch-out — no bypasses (diff path; the O(0) gate row)

33.4 µs

style: select+unselect 256-band under the default sheet — bypass punch-out — 128 of 256 bypassed (punched out to merged writes)

233 µs

style: sheet swap (applyAll) — 0 vs 128 bypasses — no bypasses

2.47 ms

style: sheet swap (applyAll) — 0 vs 128 bypasses — 128 bypassed

8.91 ms

10 ns100 ns1 µs10 µs100 µs1 ms10 ms100 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
style: sheet swap (compile + applyAll)	v3	12.7 ms	13.0 ms	17.4 ms	12.1 ms	46
style: sheet swap (compile + applyAll)	gpu	2.32 ms	2.36 ms	3.00 ms	2.23 ms	282
style: compile vs compile + apply	compile only (in batch)	25.1 µs	27.7 µs	26.6 µs	24.7 µs	12
style: compile vs compile + apply	compile + apply	2.30 ms	2.36 ms	3.29 ms	2.21 ms	283
style: first apply on add (256-node band)	v3	1.41 ms	1.66 ms	7.22 ms	1.17 ms	400
style: first apply on add (256-node band)	gpu	717 µs	836 µs	2.59 ms	680 µs	825
style: applyAll, flat vs compound (100 parents — the 14.6 partition)	flat	2.30 ms	2.34 ms	2.85 ms	2.22 ms	297
style: applyAll, flat vs compound (100 parents — the 14.6 partition)	compound	2.73 ms	2.83 ms	4.76 ms	2.67 ms	243
style: applyAll — constant vs state-case (the 57.1d partition)	constant sheet	2.26 ms	2.30 ms	2.70 ms	2.21 ms	301
style: applyAll — constant vs state-case (the 57.1d partition)	state-case sheet	2.28 ms	2.35 ms	3.13 ms	2.24 ms	294
style: applyAll — state-case vs data-case (what the partition avoids)	state-case sheet	2.27 ms	2.31 ms	2.72 ms	2.22 ms	297
style: applyAll — state-case vs data-case (what the partition avoids)	data-case sheet	10.0 ms	10.2 ms	11.0 ms	9.78 ms	62
style: select+unselect 256-band — constant vs state-case sheet	constant sheet (skips restyle)	6.79 µs	6.79 µs	6.89 µs	6.67 µs	20
style: select+unselect 256-band — constant vs state-case sheet	paint-only state sheet (round-61 diff path)	32.3 µs	33.1 µs	48.1 µs	31.3 µs	19777
style: select+unselect 256-band — constant vs state-case sheet	state-case sheet (all channels; full write)	376 µs	384 µs	486 µs	371 µs	1749
style: read one prop (color)	v3	96.0 ns	108 ns	639 ns	88.0 ns	2000298
style: read one prop (color)	gpu	54.2 ns	59.3 ns	99.2 ns	53.0 ns	2868
style: read one prop (width)	v3	115 ns	124 ns	424 ns	105 ns	1894145
style: read one prop (width)	gpu	51.6 ns	57.4 ns	111 ns	48.7 ns	2967
style: numericStyle	v3	138 ns	129 ns	356 ns	82.0 ns	1794711
style: numericStyle	gpu	47.8 ns	50.9 ns	112 ns	46.4 ns	3375
style: renderedStyle (one prop)	v3	910 ns	979 ns	2.00 µs	870 ns	538373
style: renderedStyle (one prop)	gpu	63.6 ns	67.9 ns	175 ns	61.7 ns	2519
style: whole-object style()	v3	85.8 µs	92.3 µs	196 µs	84.0 µs	7546
style: whole-object style()	gpu	14.0 µs	15.4 µs	50.3 µs	13.3 µs	43802
style: bypass set (ele.style name,value)	v3	10.2 µs	10.3 µs	10.8 µs	9.90 µs	11
style: bypass set (ele.style name,value)	gpu	2.52 µs	2.80 µs	4.04 µs	2.40 µs	56
style: bypass set + removeStyle round-trip	v3	21.1 µs	22.6 µs	23.8 µs	20.1 µs	12
style: bypass set + removeStyle round-trip	gpu	3.56 µs	3.90 µs	4.65 µs	3.39 µs	39
style: select+unselect 256-band under the default sheet — bypass punch-out	no bypasses (diff path; the O(0) gate row)	33.4 µs	35.7 µs	80.1 µs	32.2 µs	19159
style: select+unselect 256-band under the default sheet — bypass punch-out	128 of 256 bypassed (punched out to merged writes)	233 µs	250 µs	498 µs	229 µs	2788
style: sheet swap (applyAll) — 0 vs 128 bypasses	no bypasses	2.47 ms	2.51 ms	3.07 ms	2.39 ms	263
style: sheet swap (applyAll) — 0 vs 128 bypasses	128 bypassed	8.91 ms	9.17 ms	10.6 ms	8.64 ms	71

style-bundle N=2,000 · 58.5 sv4 (gpu) v3

style( 'background-color' ) (was #4 in the switch)

0.98×

style( 'border-width' ) (was #6 in the switch)

1.15×

style( 'text-wrap' ) (was #73 in the switch)

1.76×

style( 'text-max-width' ) (was #74 in the switch)

1.74×

style( 'taxi-radius' ) (was #142 in the switch)

1.28×

style( 'target-distance-from-node' ) (was #150 in the switch)

1.29×

whole-object style() (gpu-only: v3 returns a different shape) — gpu node

13.1 µs

whole-object style() (gpu-only: v3 returns a different shape) — gpu edge

14.5 µs

colour-valued vs numeric reads (the string build) — gpu background-color (colour)

78.3 ns

colour-valued vs numeric reads (the string build) — gpu border-color (colour)

70.5 ns

colour-valued vs numeric reads (the string build) — gpu border-width (number)

65.4 ns

colour-valued vs numeric reads (the string build) — gpu width (number)

62.2 ns

the getters that ride readProp — gpu numericStyle( width )

61.5 ns

the getters that ride readProp — gpu effectiveOpacity()

22.7 ns

the getters that ride readProp — gpu renderedStyle( width )

79.9 ns

10 ns100 ns1 µs10 µs100 µs

All stats (table view)

group	bench	p50	avg	p99	min	samples
style( 'background-color' ) (was #4 in the switch)	v3	59.2 ns	61.1 ns	92.7 ns	57.2 ns	2806
style( 'background-color' ) (was #4 in the switch)	gpu	60.4 ns	62.0 ns	75.3 ns	59.8 ns	2763
style( 'border-width' ) (was #6 in the switch)	v3	61.7 ns	62.5 ns	74.6 ns	58.5 ns	2748
style( 'border-width' ) (was #6 in the switch)	gpu	53.7 ns	54.7 ns	65.5 ns	51.5 ns	3124
style( 'text-wrap' ) (was #73 in the switch)	v3	61.7 ns	62.7 ns	75.5 ns	58.4 ns	2729
style( 'text-wrap' ) (was #73 in the switch)	gpu	35.1 ns	35.7 ns	46.1 ns	34.0 ns	4787
style( 'text-max-width' ) (was #74 in the switch)	v3	59.4 ns	60.9 ns	76.7 ns	56.8 ns	2823
style( 'text-max-width' ) (was #74 in the switch)	gpu	34.2 ns	34.9 ns	46.2 ns	33.0 ns	4902
style( 'taxi-radius' ) (was #142 in the switch)	v3	61.8 ns	62.2 ns	73.4 ns	58.7 ns	2764
style( 'taxi-radius' ) (was #142 in the switch)	gpu	48.3 ns	52.3 ns	95.3 ns	46.4 ns	3257
style( 'target-distance-from-node' ) (was #150 in the switch)	v3	61.3 ns	63.0 ns	79.6 ns	59.7 ns	2726
style( 'target-distance-from-node' ) (was #150 in the switch)	gpu	47.6 ns	49.0 ns	66.3 ns	46.9 ns	3493
whole-object style() (gpu-only: v3 returns a different shape)	gpu node	13.1 µs	13.7 µs	27.8 µs	12.3 µs	48880
whole-object style() (gpu-only: v3 returns a different shape)	gpu edge	14.5 µs	15.3 µs	37.6 µs	13.8 µs	43989
colour-valued vs numeric reads (the string build)	gpu background-color (colour)	78.3 ns	80.8 ns	130 ns	76.0 ns	2118
colour-valued vs numeric reads (the string build)	gpu border-color (colour)	70.5 ns	73.5 ns	107 ns	68.3 ns	2326
colour-valued vs numeric reads (the string build)	gpu border-width (number)	65.4 ns	67.1 ns	88.7 ns	63.1 ns	2549
colour-valued vs numeric reads (the string build)	gpu width (number)	62.2 ns	64.4 ns	108 ns	60.5 ns	2657
the getters that ride readProp	gpu numericStyle( width )	61.5 ns	62.3 ns	74.5 ns	59.5 ns	2746
the getters that ride readProp	gpu effectiveOpacity()	22.7 ns	23.9 ns	36.1 ns	21.9 ns	7160
the getters that ride readProp	gpu renderedStyle( width )	79.9 ns	82.1 ns	110 ns	78.9 ns	2083

load N=2,000 · 58.5 sv4 (gpu) v3

load: init from the definition form (construct + dispose)

27×

load: v4 ingest forms (construct + dispose) — definition form

4.50 ms

load: v4 ingest forms (construct + dispose) — columnar form

3.34 ms

load: v4 ingest forms (construct + dispose) — wire buffer

2.89 ms

load: v4 ingest forms (construct + dispose) — def clone alone (control)

311 µs

load: conversion — toColumnarElements

708 µs

load: conversion — serializeElements (from columnar)

180 µs

load: conversion — serializeElements (from defs)

886 µs

load: conversion — deserializeElements

4.60 µs

load: export from a live graph — cy.serialize() (wire)

502 µs

load: export from a live graph — cy.json() (v4)

716 µs

load: cy.json()

2.7×

load: add a 256-node band (add + remove)

1.65×

load: add a 256-node band, v4 forms — definition form

2.29 ms

load: add a 256-node band, v4 forms — columnar form

1.94 ms

load: add a 256-node band, v4 forms — wire buffer

1.92 ms

1 µs10 µs100 µs1 ms10 ms100 ms1 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
load: init from the definition form (construct + dispose)	v3	134 ms	147 ms	188 ms	127 ms	12
load: init from the definition form (construct + dispose)	gpu	5.03 ms	6.62 ms	20.3 ms	4.22 ms	97
load: v4 ingest forms (construct + dispose)	definition form	4.50 ms	6.50 ms	16.3 ms	4.18 ms	102
load: v4 ingest forms (construct + dispose)	columnar form	3.34 ms	4.46 ms	13.5 ms	3.23 ms	150
load: v4 ingest forms (construct + dispose)	wire buffer	2.89 ms	4.36 ms	13.0 ms	2.79 ms	156
load: v4 ingest forms (construct + dispose)	def clone alone (control)	311 µs	619 µs	16.6 ms	213 µs	1061
load: conversion	toColumnarElements	708 µs	753 µs	1.15 ms	689 µs	921
load: conversion	serializeElements (from columnar)	180 µs	369 µs	993 µs	173 µs	1886
load: conversion	serializeElements (from defs)	886 µs	903 µs	1.31 ms	852 µs	770
load: conversion	deserializeElements	4.60 µs	4.34 µs	5.72 µs	3.33 µs	30
load: export from a live graph	cy.serialize() (wire)	502 µs	577 µs	1.98 ms	486 µs	1192
load: export from a live graph	cy.json() (v4)	716 µs	1.05 ms	6.79 ms	687 µs	636
load: cy.json()	v3	2.00 ms	2.34 ms	6.88 ms	1.94 ms	290
load: cy.json()	gpu	739 µs	839 µs	2.63 ms	723 µs	828
load: add a 256-node band (add + remove)	v3	3.65 ms	4.36 ms	11.0 ms	3.26 ms	151
load: add a 256-node band (add + remove)	gpu	2.22 ms	2.49 ms	7.48 ms	1.48 ms	271
load: add a 256-node band, v4 forms	definition form	2.29 ms	2.30 ms	5.09 ms	1.51 ms	297
load: add a 256-node band, v4 forms	columnar form	1.94 ms	2.00 ms	3.58 ms	1.31 ms	335
load: add a 256-node band, v4 forms	wire buffer	1.92 ms	1.96 ms	3.54 ms	1.28 ms	350

spatial N=2,000 · 1.9 minv4 (gpu) v3

pick: hit (the walk; one inside-test) — zoom 1

20.8 µs

pick: hit (the walk; one inside-test) — zoom 2

21.1 µs

pick: miss walk — exact vs 24 px touch halo (57.9) — exact

46.6 µs

pick: miss walk — exact vs 24 px touch halo (57.9) — padPx 24

44.8 µs

pick: shape inside-test, 2000 coincident misses — ellipse

88.3 µs

pick: shape inside-test, 2000 coincident misses — star

235 µs

pick: shape inside-test, 2000 coincident misses — round-hexagon

794 µs

pick: shape inside-test, 2000 coincident misses — barrel

686 µs

pick: shape inside-test, 2000 coincident misses — cut-rectangle

100 µs

pick: shape inside-test, 2000 coincident misses — polygon

211 µs

pick: round-hexagon inside-test by zoom (not affine-invariant) — zoom 1

790 µs

pick: round-hexagon inside-test by zoom (not affine-invariant) — zoom 2

799 µs

pick: background miss (full highWater walk) — gpu

45.8 µs

pick: label box in the scan (text-events) — text-events: no

45.3 µs

pick: label box in the scan (text-events) — text-events: yes

126 µs

box: elementsInBox by covered fraction — 10%

148 µs

box: elementsInBox by covered fraction — 50%

193 µs

box: elementsInBox by covered fraction — 100%

321 µs

box: label containment (boxSelectionIncludesLabels) — off

228 µs

box: label containment (boxSelectionIncludesLabels) — on

263 µs

box: overlap vs contain (straight edges) — contain

235 µs

box: overlap vs contain (straight edges) — overlap

508 µs

box: overlap vs contain (curved edges — the segment walk) — contain

1.28 ms

box: overlap vs contain (curved edges — the segment walk) — overlap

2.20 ms

bounds: whole-graph boundingBox()

5.8×

bounds: boundingBox, labels in vs out (v4) — includeLabels: true (default)

1.36 ms

bounds: boundingBox, labels in vs out (v4) — includeLabels: false

811 µs

bounds: cy.fit() (whole-graph scan)

32×

bounds: getFitViewport() (compute without committing)

33×

bounds: boundingBoxAt (v4 only) — gpu

2.65 ms

bounds: one node boundingBox()

1.61×

bounds: labelBoundingBox (v4 only) — gpu

427 ns

bounds: compound + taxi conservative scan (round 54, v4 only) — warm (curve bbs memoized)

75.4 µs

bounds: compound + taxi conservative scan (round 54, v4 only) — cold (geometry write per call)

149 µs

100 ns1 µs10 µs100 µs1 ms10 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
pick: hit (the walk; one inside-test)	zoom 1	20.8 µs	21.6 µs	42.6 µs	20.3 µs	31826
pick: hit (the walk; one inside-test)	zoom 2	21.1 µs	21.6 µs	21.9 µs	20.9 µs	12
pick: miss walk — exact vs 24 px touch halo (57.9)	exact	46.6 µs	47.0 µs	47.5 µs	46.0 µs	12
pick: miss walk — exact vs 24 px touch halo (57.9)	padPx 24	44.8 µs	46.2 µs	87.6 µs	43.5 µs	15031
pick: shape inside-test, 2000 coincident misses	ellipse	88.3 µs	90.5 µs	133 µs	84.7 µs	7707
pick: shape inside-test, 2000 coincident misses	star	235 µs	248 µs	414 µs	229 µs	2812
pick: shape inside-test, 2000 coincident misses	round-hexagon	794 µs	811 µs	1.07 ms	768 µs	863
pick: shape inside-test, 2000 coincident misses	barrel	686 µs	706 µs	1.01 ms	673 µs	988
pick: shape inside-test, 2000 coincident misses	cut-rectangle	100 µs	103 µs	161 µs	97.3 µs	6795
pick: shape inside-test, 2000 coincident misses	polygon	211 µs	217 µs	312 µs	203 µs	3225
pick: round-hexagon inside-test by zoom (not affine-invariant)	zoom 1	790 µs	804 µs	953 µs	769 µs	870
pick: round-hexagon inside-test by zoom (not affine-invariant)	zoom 2	799 µs	813 µs	1.27 ms	777 µs	861
pick: background miss (full highWater walk)	gpu	45.8 µs	46.2 µs	47.3 µs	44.8 µs	12
pick: label box in the scan (text-events)	text-events: no	45.3 µs	45.7 µs	46.2 µs	44.7 µs	12
pick: label box in the scan (text-events)	text-events: yes	126 µs	130 µs	215 µs	122 µs	5325
box: elementsInBox by covered fraction	10%	148 µs	153 µs	241 µs	147 µs	4566
box: elementsInBox by covered fraction	50%	193 µs	201 µs	308 µs	190 µs	3479
box: elementsInBox by covered fraction	100%	321 µs	337 µs	793 µs	314 µs	2073
box: label containment (boxSelectionIncludesLabels)	off	228 µs	238 µs	402 µs	225 µs	2941
box: label containment (boxSelectionIncludesLabels)	on	263 µs	274 µs	635 µs	259 µs	2552
box: overlap vs contain (straight edges)	contain	235 µs	245 µs	362 µs	231 µs	2855
box: overlap vs contain (straight edges)	overlap	508 µs	514 µs	896 µs	447 µs	1356
box: overlap vs contain (curved edges — the segment walk)	contain	1.28 ms	1.28 ms	1.70 ms	1.24 ms	544
box: overlap vs contain (curved edges — the segment walk)	overlap	2.20 ms	2.25 ms	2.94 ms	2.18 ms	309
bounds: whole-graph boundingBox()	v3	7.96 ms	8.11 ms	9.60 ms	7.58 ms	80
bounds: whole-graph boundingBox()	gpu	1.38 ms	1.41 ms	1.85 ms	1.35 ms	471
bounds: boundingBox, labels in vs out (v4)	includeLabels: true (default)	1.36 ms	1.39 ms	2.00 ms	1.35 ms	498
bounds: boundingBox, labels in vs out (v4)	includeLabels: false	811 µs	834 µs	1.26 ms	789 µs	837
bounds: cy.fit() (whole-graph scan)	v3	7.72 ms	7.88 ms	9.43 ms	7.32 ms	85
bounds: cy.fit() (whole-graph scan)	gpu	238 µs	248 µs	402 µs	230 µs	2814
bounds: getFitViewport() (compute without committing)	v3	7.76 ms	7.88 ms	9.07 ms	7.39 ms	85
bounds: getFitViewport() (compute without committing)	gpu	236 µs	245 µs	310 µs	232 µs	2849
bounds: boundingBoxAt (v4 only)	gpu	2.65 ms	2.73 ms	3.83 ms	2.60 ms	249
bounds: one node boundingBox()	v3	1.64 µs	1.66 µs	1.91 µs	1.60 µs	98
bounds: one node boundingBox()	gpu	1.01 µs	1.03 µs	1.26 µs	973 ns	158
bounds: labelBoundingBox (v4 only)	gpu	427 ns	453 ns	976 ns	416 ns	371
bounds: compound + taxi conservative scan (round 54, v4 only)	warm (curve bbs memoized)	75.4 µs	78.3 µs	170 µs	73.3 µs	8896
bounds: compound + taxi conservative scan (round 54, v4 only)	cold (geometry write per call)	149 µs	154 µs	240 µs	144 µs	4552

data N=2,000 · 1.4 minv4 (gpu) v3

data: bulk write, numeric column

21×

data: bulk write, dictionary string column (4 values)

16×

data: bulk write, one new dictionary entry per pass

18×

data: bulk write, object column (plain-array fallback)

21×

data: removeData (whole collection)

7.9×

data: read one numeric key

1.35×

data: read one string key (dictionary decode)

1.35×

data: read the whole object

2.3×

query: data equality ([bucket = 3] vs { data: { bucket: 3 } })

14×

query: data comparison ([weight > 3] vs { gt })

11×

query: two keys AND-ed

12×

query: membership ([bucket in ...] vs { in })

15×

query: predicate function (both sides)

8.6×

query: structural, parents (:parent vs { parent: true })

33×

query: structural, childless (:childless vs { parent: false })

13×

query: structural, orphans (:orphan vs { child: false })

32×

10 ns100 ns1 µs10 µs100 µs1 ms10 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
data: bulk write, numeric column	v3	1.78 ms	1.82 ms	3.15 ms	1.70 ms	366
data: bulk write, numeric column	gpu	84.2 µs	87.7 µs	130 µs	81.2 µs	7977
data: bulk write, dictionary string column (4 values)	v3	1.77 ms	1.80 ms	2.32 ms	1.72 ms	372
data: bulk write, dictionary string column (4 values)	gpu	108 µs	113 µs	311 µs	95.9 µs	6204
data: bulk write, one new dictionary entry per pass	v3	1.79 ms	1.98 ms	3.33 ms	1.77 ms	347
data: bulk write, one new dictionary entry per pass	gpu	101 µs	105 µs	156 µs	95.7 µs	6650
data: bulk write, object column (plain-array fallback)	v3	1.78 ms	1.84 ms	3.12 ms	1.75 ms	377
data: bulk write, object column (plain-array fallback)	gpu	83.3 µs	85.4 µs	123 µs	81.1 µs	8192
data: removeData (whole collection)	v3	2.62 ms	2.68 ms	3.95 ms	2.51 ms	256
data: removeData (whole collection)	gpu	330 µs	344 µs	658 µs	315 µs	1959
data: read one numeric key	v3	38.5 ns	40.1 ns	90.1 ns	37.2 ns	4280
data: read one numeric key	gpu	28.6 ns	30.1 ns	47.9 ns	27.5 ns	5682
data: read one string key (dictionary decode)	v3	42.0 ns	43.6 ns	88.1 ns	40.2 ns	3937
data: read one string key (dictionary decode)	gpu	31.2 ns	31.5 ns	47.6 ns	27.0 ns	5433
data: read the whole object	v3	47.7 ns	49.3 ns	77.9 ns	45.5 ns	3486
data: read the whole object	gpu	21.0 ns	22.4 ns	40.4 ns	20.1 ns	7606
query: data equality ([bucket = 3] vs { data: { bucket: 3 } })	v3	656 µs	681 µs	1.14 ms	638 µs	1021
query: data equality ([bucket = 3] vs { data: { bucket: 3 } })	gpu	47.3 µs	51.2 µs	125 µs	41.8 µs	13491
query: data comparison ([weight > 3] vs { gt })	v3	751 µs	777 µs	1.21 ms	719 µs	898
query: data comparison ([weight > 3] vs { gt })	gpu	71.5 µs	75.6 µs	181 µs	58.1 µs	9190
query: two keys AND-ed	v3	824 µs	876 µs	1.46 ms	777 µs	795
query: two keys AND-ed	gpu	66.7 µs	70.7 µs	137 µs	64.3 µs	9828
query: membership ([bucket in ...] vs { in })	v3	1.00 ms	1.03 ms	1.46 ms	961 µs	677
query: membership ([bucket in ...] vs { in })	gpu	65.1 µs	70.9 µs	154 µs	62.2 µs	9770
query: predicate function (both sides)	v3	651 µs	673 µs	937 µs	638 µs	1036
query: predicate function (both sides)	gpu	76.1 µs	79.3 µs	142 µs	73.8 µs	8774
query: structural, parents (:parent vs { parent: true })	v3	515 µs	543 µs	1.07 ms	508 µs	1288
query: structural, parents (:parent vs { parent: true })	gpu	15.4 µs	16.3 µs	19.1 µs	15.1 µs	12
query: structural, childless (:childless vs { parent: false })	v3	796 µs	814 µs	981 µs	784 µs	857
query: structural, childless (:childless vs { parent: false })	gpu	61.4 µs	78.4 µs	207 µs	55.9 µs	8698
query: structural, orphans (:orphan vs { child: false })	v3	488 µs	499 µs	619 µs	482 µs	1397
query: structural, orphans (:orphan vs { child: false })	gpu	15.4 µs	16.0 µs	16.3 µs	15.2 µs	12

events N=2,000 · 1.3 minv4 (gpu) v3

emit: position write, no listeners

27×

emit: 1 core listener

8.3×

emit: 1 element (ref-qualified) listener

7.7×

emit: delegated (v3 selector vs v4 predicate)

8.2×

emit: 10 core listeners

6.2×

emit: on() + off() round-trip

2.1×

emit: flat vs phased (compound bubbling, 14.5) — orphan (flat emit)

305 ns

emit: flat vs phased (compound bubbling, 14.5) — child of 2 ancestors (phased)

850 ns

emit: flat vs phased, no listeners — orphan (flat)

83.5 ns

emit: flat vs phased, no listeners — child (phased)

506 ns

anim: start + stop, one element

1.28×

anim: start against a running animation (round-21 eviction) — disjoint channels (composes)

1.26 µs

anim: start against a running animation (round-21 eviction) — overlapping channels (evicts)

813 ns

anim: start + stop, 512-node collection

3.0×

anim: delay() (no channels touched)

1.58×

anim: controls (v4 only) — pause + resume

456 ns

anim: controls (v4 only) — reverse

873 ns

anim: viewport animate + stop

2.9×

10 ns100 ns1 µs10 µs100 µs1 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
emit: position write, no listeners	v3	1.31 µs	1.34 µs	1.64 µs	1.27 µs	111
emit: position write, no listeners	gpu	48.9 ns	52.3 ns	99.5 ns	46.9 ns	3009
emit: 1 core listener	v3	1.40 µs	1.43 µs	1.93 µs	1.36 µs	116
emit: 1 core listener	gpu	169 ns	179 ns	306 ns	165 ns	893
emit: 1 element (ref-qualified) listener	v3	1.41 µs	1.52 µs	2.27 µs	1.37 µs	102
emit: 1 element (ref-qualified) listener	gpu	183 ns	202 ns	1.08 µs	178 ns	783
emit: delegated (v3 selector vs v4 predicate)	v3	1.50 µs	1.57 µs	2.75 µs	1.46 µs	103
emit: delegated (v3 selector vs v4 predicate)	gpu	183 ns	197 ns	437 ns	181 ns	803
emit: 10 core listeners	v3	1.97 µs	2.03 µs	2.22 µs	1.92 µs	77
emit: 10 core listeners	gpu	316 ns	323 ns	462 ns	308 ns	497
emit: on() + off() round-trip	v3	460 ns	561 ns	1.45 µs	422 ns	289
emit: on() + off() round-trip	gpu	216 ns	260 ns	913 ns	183 ns	631
emit: flat vs phased (compound bubbling, 14.5)	orphan (flat emit)	305 ns	356 ns	1.97 µs	280 ns	975657
emit: flat vs phased (compound bubbling, 14.5)	child of 2 ancestors (phased)	850 ns	923 ns	1.96 µs	818 ns	171
emit: flat vs phased, no listeners	orphan (flat)	83.5 ns	93.8 ns	679 ns	80.3 ns	1800
emit: flat vs phased, no listeners	child (phased)	506 ns	579 ns	1.99 µs	483 ns	285
anim: start + stop, one element	v3	824 ns	1.25 µs	3.44 µs	752 ns	421778
anim: start + stop, one element	gpu	642 ns	1.06 µs	4.63 µs	588 ns	459980
anim: start against a running animation (round-21 eviction)	disjoint channels (composes)	1.26 µs	1.94 µs	4.23 µs	1.07 µs	81
anim: start against a running animation (round-21 eviction)	overlapping channels (evicts)	813 ns	1.41 µs	3.50 µs	793 ns	106
anim: start + stop, 512-node collection	v3	299 µs	434 µs	2.49 ms	287 µs	1472
anim: start + stop, 512-node collection	gpu	100 µs	112 µs	181 µs	96.5 µs	5739
anim: delay() (no channels touched)	v3	651 ns	959 ns	3.40 µs	631 ns	167
anim: delay() (no channels touched)	gpu	411 ns	723 ns	2.66 µs	407 ns	224
anim: controls (v4 only)	pause + resume	456 ns	738 ns	2.55 µs	452 ns	208
anim: controls (v4 only)	reverse	873 ns	1.50 µs	3.41 µs	849 ns	99
anim: viewport animate + stop	v3	572 ns	804 ns	2.92 µs	558 ns	192
anim: viewport animate + stop	gpu	198 ns	219 ns	575 ns	194 ns	726

store N=2,000 · 1.0 min

id index: set x 2000 (build) — gpu

269 µs

id index: single-key operations — get (probe)

66.0 ns

id index: single-key operations — has

51.2 ns

id index: single-key operations — idAt (memoized hit)

8.97 ns

id index: single-key operations — hashAt

6.98 ns

id index: remove + re-set (churn, drives the round-11 reclaim) — gpu

231 ns

adjacency: rebuild (4000 edges, two counting passes) — gpu

39.1 µs

adjacency: reads — outDegree (O(1))

6.16 ns

adjacency: reads — outEdges

44.7 ns

adjacency: reads — connectedEdges

118 ns

adjacency: addEdge + removeEdge (the overlay path) — gpu

143 ns

blob: write x 2000 (build) — gpu

105 µs

blob: single-record operations — rewrite in place (same length)

30.1 ns

blob: single-record operations — offsetOf

4.37 ns

blob: single-record operations — free + rewrite

155 ns

dirty: mark (per-write span coalescing) — contiguous marks

12.9 ns

dirty: mark (per-write span coalescing) — scattered marks

17.8 ns

dirty: take (a frame's drain) — gpu

598 ns

images: registry bookkeeping — acquire (existing url)

203 ns

images: registry bookkeeping — acquire (new url) + release

533 ns

charts: sheet apply, chart vs no chart — no chart

2.28 ms

charts: sheet apply, chart vs no chart — pie from data (blob record per node)

2.72 ms

charts: data write refreshing chart-values — gpu

1.49 ms

1 ns10 ns100 ns1 µs10 µs100 µs1 ms10 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
id index: set x 2000 (build)	gpu	269 µs	278 µs	412 µs	254 µs	2526
id index: single-key operations	get (probe)	66.0 ns	65.4 ns	112 ns	53.5 ns	2609
id index: single-key operations	has	51.2 ns	52.6 ns	78.0 ns	49.5 ns	3248
id index: single-key operations	idAt (memoized hit)	8.97 ns	9.27 ns	17.3 ns	8.71 ns	18307
id index: single-key operations	hashAt	6.98 ns	7.21 ns	14.1 ns	6.89 ns	23495
id index: remove + re-set (churn, drives the round-11 reclaim)	gpu	231 ns	237 ns	301 ns	225 ns	708
adjacency: rebuild (4000 edges, two counting passes)	gpu	39.1 µs	47.1 µs	232 µs	36.8 µs	14698
adjacency: reads	outDegree (O(1))	6.16 ns	6.36 ns	12.6 ns	5.78 ns	26564
adjacency: reads	outEdges	44.7 ns	45.8 ns	60.1 ns	43.8 ns	3725
adjacency: reads	connectedEdges	118 ns	181 ns	2.11 µs	114 ns	923
adjacency: addEdge + removeEdge (the overlay path)	gpu	143 ns	147 ns	207 ns	141 ns	1137
blob: write x 2000 (build)	gpu	105 µs	130 µs	851 µs	101 µs	5344
blob: single-record operations	rewrite in place (same length)	30.1 ns	31.0 ns	42.6 ns	29.8 ns	5511
blob: single-record operations	offsetOf	4.37 ns	4.48 ns	7.54 ns	4.18 ns	37472
blob: single-record operations	free + rewrite	155 ns	181 ns	587 ns	147 ns	939
dirty: mark (per-write span coalescing)	contiguous marks	12.9 ns	13.3 ns	23.3 ns	12.5 ns	12753
dirty: mark (per-write span coalescing)	scattered marks	17.8 ns	18.2 ns	28.1 ns	17.2 ns	9347
dirty: take (a frame's drain)	gpu	598 ns	702 ns	1.20 µs	583 ns	233
images: registry bookkeeping	acquire (existing url)	203 ns	204 ns	244 ns	198 ns	834
images: registry bookkeeping	acquire (new url) + release	533 ns	597 ns	2.46 µs	519 ns	277
charts: sheet apply, chart vs no chart	no chart	2.28 ms	2.37 ms	4.19 ms	2.21 ms	284
charts: sheet apply, chart vs no chart	pie from data (blob record per node)	2.72 ms	2.77 ms	3.31 ms	2.63 ms	246
charts: data write refreshing chart-values	gpu	1.49 ms	1.53 ms	1.91 ms	1.44 ms	454

surface N=2,000 · 8.8 minv4 (gpu) v3

viewport: zoom() get

4.6×

viewport: zoom( n ) set

1.84×

viewport: pan() get

1.10×

viewport: pan( p ) set

1.85×

viewport: panBy()

1.63×

viewport: extent()

1.50×

viewport: renderedExtent()

1.58×

viewport: width() + height() + size() (3 calls)

3.8×

viewport: minZoom/maxZoom get (2 calls)

15×

viewport: center( eles )

19×

viewport: reset()

104×

viewport: viewport({ zoom, pan })

59×

viewport: getCenterPan( eles ) — gpu

6.62 µs

core: instanceString + headless + styleEnabled (3 calls)

1.25×

core: hasElementWithId + hasCompoundNodes (2 calls)

1.12×

core: options()

85×

core: mutableElements() (v4 materializes; v3 is O(1))

1.36×

core: gating getters (5 calls)

1.15×

core: gating setters round-trip (autolock/autoungrabify/autounselectify)

1.86×

core: interaction tuning getters (5 calls) — gpu

7.25 ns

core: graph data() set + get (v3 restyles the graph)

8305×

core: graph scratch() set + get (v3 restyles the graph)

16693×

core: batch( fn ) (empty)

6.4×

core: startBatch + endBatch + batching

6.0×

core: collection()

7.0×

core: compact() (no dead slots — the early-out) — gpu

16.7 µs

collection: element accessors (id/group/isNode/isEdge, 4 calls)

1.74×

collection: length + empty + nonempty (3 calls)

4.6×

collection: eq/first/last (3 calls)

1.88×

collection: slice()

2.2×

collection: toArray()

12×

collection: sort( fn )

1.57×

collection: reduce( fn )

1.40×

collection: min + max ( fn ) (2 calls)

1.96×

collection: some + every (2 calls)

1.59×

collection: same + anySame + contains (3 calls)

73×

collection: allAre (all match: the full walk)

1.49×

collection: is (none match: the full walk)

2.4×

collection: union

5.9×

collection: intersection

28×

collection: difference

29×

collection: symmetricDifference

3.9×

collection: absoluteComplement

16×

collection: byGroup

3.1×

collection: diff

3.3×

collection: indexOf

1.35×

traversal: openNeighborhood + closedNeighborhood (2 calls)

3.0×

traversal: connectedNodes

3.4×

traversal: sources + targets (2 calls)

3.2×

traversal: source() + target() (2 calls)

1.36×

traversal: edgesWith

1.27×

traversal: edgesTo

1.26×

traversal: parallelEdges + codirectedEdges (2 calls)

3.6×

traversal: roots + leaves (2 calls)

13×

traversal: allAreNeighbors

5.4×

traversal: isLoop + isSimple (2 calls)

2.3×

degree: indegree + outdegree (2 calls)

76×

degree: minDegree + maxDegree (2 calls)

42×

degree: minIndegree + maxOutdegree (2 calls)

24×

dimensions: renderedPosition get

2.5×

dimensions: relativePosition get

11×

dimensions: width + height + outerWidth + outerHeight (4 calls)

2.9×

dimensions: rendered dims (4 calls)

3.1×

dimensions: renderedBoundingBox

1.83×

dimensions: midpoint + renderedMidpoint (2 calls) — gpu

1.17 µs

dimensions: source/targetEndpoint (2 calls) — gpu

377 ns

dimensions: rendered source/targetEndpoint (2 calls) — gpu

399 ns

dimensions: shift + silentShift (round-trip)

62×

dimensions: silentPosition

15×

dimensions: positions( fn ) over 100

50×

dimensions: padding + paddedWidth + paddedHeight (3 calls)

13×

curves: isBundledBezier — gpu

33.5 ns

curves: controlPoints + renderedControlPoints (2 calls) — gpu

534 ns

curves: segmentPoints (undefined for bezier) — gpu

41.6 ns

compound: parent + parents + ancestors (3 calls)

3.7×

compound: children + descendants + siblings (3 calls)

4.7×

compound: orphans + nonorphans (2 calls)

7.0×

compound: commonAncestors

127×

compound: isParent/isChild/isChildless/isOrphan (4 calls)

1.56×

flags: select + unselect + selected (band)

16×

flags: selectify + unselectify + selectable

97×

flags: lock + unlock + locked

93×

flags: grabify + ungrabify + grabbable

95×

flags: show + hide + visible

226×

flags: active/activate/unactivate (3 calls)

15×

flags: panify + unpanify + pannable

94×

flags: takesUpSpace + interactive + transparent (3 calls)

2.8×

flags: effectiveOpacity

2.1×

element data: scratch get + set

7.6×

element data: json() one element

2.0×

element data: jsons() over 100

1.93×

element data: label() — gpu

15.1 ns

core: batchData()

8.7×

core: animated()

4.9×

core: zoomRange get — gpu

2.16 ns

core: forceRender() (headless no-op)

16×

core: resize() / invalidateSize()

2.3×

core: onRender + offRender

2.4×

core: removeScratch()

9502×

core: selectionType get

2.00×

core: isReady()

2.2×

core: window()

2.8×

core: container()

2.6×

core: destroyed()

2.2×

core: removeAllListeners()

8.7×

collection: cy()

3.5×

collection: indexOfId()

1.76×

collection: removed() + inside() (2 calls)

1.16×

collection: silentPositions( fn ) over 100

31×

collection: animated()

2.9×

collection: delayAnimation()

2.8×

collection: removeScratch()

8.8×

collection: component()

44×

collection: layoutDimensions()

3.6×

degree: maxIndegree + minOutdegree (2 calls)

24×

flags: grabbed + hidden + inactive (3 calls)

2.2×

curves: renderedSegmentPoints — gpu

36.2 ns

layout contract: edgeSlots + endpoints + degreeOf — gpu

34.7 µs

formats: isColumnarElements + isSerializedElements + isPackedIds — gpu

4.61 ns

0.100 ns1 ns10 ns100 ns1 µs10 µs100 µs1 ms10 ms100 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
viewport: zoom() get	v3	5.59 ns	5.73 ns	10.9 ns	5.40 ns	29387
viewport: zoom() get	gpu	1.21 ns	1.23 ns	1.29 ns	1.21 ns	129473
viewport: zoom( n ) set	v3	602 ns	613 ns	811 ns	565 ns	270
viewport: zoom( n ) set	gpu	326 ns	331 ns	392 ns	315 ns	507
viewport: pan() get	v3	5.39 ns	5.53 ns	6.54 ns	5.28 ns	30478
viewport: pan() get	gpu	4.91 ns	5.03 ns	9.08 ns	4.80 ns	33540
viewport: pan( p ) set	v3	610 ns	613 ns	679 ns	583 ns	273
viewport: pan( p ) set	gpu	331 ns	337 ns	397 ns	322 ns	501
viewport: panBy()	v3	561 ns	569 ns	676 ns	546 ns	296
viewport: panBy()	gpu	343 ns	348 ns	508 ns	308 ns	483
viewport: extent()	v3	29.3 ns	44.1 ns	206 ns	28.2 ns	3837
viewport: extent()	gpu	19.5 ns	26.0 ns	106 ns	18.4 ns	6460
viewport: renderedExtent()	v3	26.8 ns	28.2 ns	51.7 ns	25.7 ns	6007
viewport: renderedExtent()	gpu	17.0 ns	23.8 ns	106 ns	15.7 ns	7066
viewport: width() + height() + size() (3 calls)	v3	20.9 ns	21.9 ns	44.6 ns	20.0 ns	7763
viewport: width() + height() + size() (3 calls)	gpu	5.55 ns	5.84 ns	11.3 ns	5.38 ns	28838
viewport: minZoom/maxZoom get (2 calls)	v3	14.8 ns	15.4 ns	25.9 ns	14.1 ns	11063
viewport: minZoom/maxZoom get (2 calls)	gpu	0.999 ns	1.05 ns	1.84 ns	0.977 ns	149337
viewport: center( eles )	v3	122 µs	128 µs	227 µs	117 µs	5432
viewport: center( eles )	gpu	6.51 µs	6.57 µs	6.88 µs	6.43 µs	22
viewport: reset()	v3	927 ns	1.02 µs	1.56 µs	894 ns	162
viewport: reset()	gpu	8.95 ns	9.31 ns	17.8 ns	8.51 ns	18156
viewport: viewport({ zoom, pan })	v3	929 ns	1.02 µs	1.60 µs	897 ns	164
viewport: viewport({ zoom, pan })	gpu	15.9 ns	16.6 ns	38.7 ns	15.1 ns	10246
viewport: getCenterPan( eles )	gpu	6.62 µs	6.60 µs	6.74 µs	6.46 µs	22
core: instanceString + headless + styleEnabled (3 calls)	v3	32.3 ns	33.9 ns	62.5 ns	31.0 ns	5028
core: instanceString + headless + styleEnabled (3 calls)	gpu	25.8 ns	26.9 ns	41.4 ns	25.1 ns	6334
core: hasElementWithId + hasCompoundNodes (2 calls)	v3	35.8 ns	36.6 ns	47.1 ns	34.5 ns	4690
core: hasElementWithId + hasCompoundNodes (2 calls)	gpu	32.0 ns	33.6 ns	59.5 ns	31.1 ns	5073
core: options()	v3	90.6 ns	94.2 ns	171 ns	89.1 ns	1810
core: options()	gpu	1.07 ns	1.10 ns	1.53 ns	1.03 ns	144392
core: mutableElements() (v4 materializes; v3 is O(1))	v3	3.19 ns	3.35 ns	6.56 ns	3.14 ns	49737
core: mutableElements() (v4 materializes; v3 is O(1))	gpu	2.35 ns	2.47 ns	6.14 ns	2.20 ns	66894
core: gating getters (5 calls)	v3	15.5 ns	16.2 ns	28.7 ns	14.8 ns	10566
core: gating getters (5 calls)	gpu	13.5 ns	14.0 ns	25.8 ns	12.7 ns	12199
core: gating setters round-trip (autolock/autoungrabify/autounselectify)	v3	11.2 ns	11.6 ns	19.8 ns	10.8 ns	14707
core: gating setters round-trip (autolock/autoungrabify/autounselectify)	gpu	6.00 ns	6.18 ns	12.8 ns	5.74 ns	27354
core: interaction tuning getters (5 calls)	gpu	7.25 ns	7.51 ns	13.7 ns	7.00 ns	22619
core: graph data() set + get (v3 restyles the graph)	v3	2.74 ms	2.78 ms	3.29 ms	2.63 ms	247
core: graph data() set + get (v3 restyles the graph)	gpu	330 ns	338 ns	428 ns	317 ns	501
core: graph scratch() set + get (v3 restyles the graph)	v3	2.77 ms	2.91 ms	5.41 ms	2.65 ms	236
core: graph scratch() set + get (v3 restyles the graph)	gpu	166 ns	172 ns	276 ns	161 ns	992
core: batch( fn ) (empty)	v3	533 ns	638 ns	1.56 µs	499 ns	264
core: batch( fn ) (empty)	gpu	83.7 ns	88.1 ns	156 ns	82.1 ns	1927
core: startBatch + endBatch + batching	v3	540 ns	653 ns	1.62 µs	516 ns	258
core: startBatch + endBatch + batching	gpu	90.4 ns	162 ns	2.85 µs	87.5 ns	1022
core: collection()	v3	451 ns	527 ns	1.53 µs	440 ns	318
core: collection()	gpu	64.6 ns	67.8 ns	112 ns	61.7 ns	2504
core: compact() (no dead slots — the early-out)	gpu	16.7 µs	21.1 µs	96.1 µs	16.4 µs	32389
collection: element accessors (id/group/isNode/isEdge, 4 calls)	v3	60.6 ns	63.4 ns	95.8 ns	58.1 ns	2678
collection: element accessors (id/group/isNode/isEdge, 4 calls)	gpu	34.8 ns	37.5 ns	135 ns	33.3 ns	4526
collection: length + empty + nonempty (3 calls)	v3	6.68 ns	6.88 ns	11.8 ns	6.47 ns	24680
collection: length + empty + nonempty (3 calls)	gpu	1.47 ns	1.52 ns	2.20 ns	1.47 ns	106622
collection: eq/first/last (3 calls)	v3	15.2 ns	15.7 ns	26.5 ns	14.7 ns	10858
collection: eq/first/last (3 calls)	gpu	8.09 ns	8.40 ns	15.7 ns	7.93 ns	20168
collection: slice()	v3	3.71 µs	3.80 µs	4.22 µs	3.61 µs	40
collection: slice()	gpu	1.70 µs	1.71 µs	1.80 µs	1.64 µs	95
collection: toArray()	v3	8.80 µs	8.97 µs	9.89 µs	8.63 µs	15
collection: toArray()	gpu	754 ns	770 ns	944 ns	733 ns	212
collection: sort( fn )	v3	14.2 µs	14.8 µs	14.9 µs	14.0 µs	12
collection: sort( fn )	gpu	9.06 µs	9.10 µs	9.25 µs	8.98 µs	15
collection: reduce( fn )	v3	157 ns	160 ns	184 ns	120 ns	1064
collection: reduce( fn )	gpu	112 ns	114 ns	131 ns	110 ns	1504
collection: min + max ( fn ) (2 calls)	v3	11.0 µs	11.0 µs	11.4 µs	10.8 µs	12
collection: min + max ( fn ) (2 calls)	gpu	5.60 µs	5.66 µs	6.00 µs	5.52 µs	25
collection: some + every (2 calls)	v3	894 ns	912 ns	1.23 µs	874 ns	182
collection: some + every (2 calls)	gpu	562 ns	567 ns	623 ns	555 ns	299
collection: same + anySame + contains (3 calls)	v3	7.33 µs	7.35 µs	7.50 µs	7.21 µs	20
collection: same + anySame + contains (3 calls)	gpu	100 ns	103 ns	142 ns	97.1 ns	1663
collection: allAre (all match: the full walk)	v3	3.66 µs	3.65 µs	3.96 µs	3.39 µs	42
collection: allAre (all match: the full walk)	gpu	2.45 µs	2.51 µs	2.89 µs	2.41 µs	63
collection: is (none match: the full walk)	v3	5.66 µs	5.67 µs	6.04 µs	5.36 µs	26
collection: is (none match: the full walk)	gpu	2.38 µs	2.42 µs	2.94 µs	2.33 µs	65
collection: union	v3	95.3 µs	105 µs	152 µs	93.2 µs	6587
collection: union	gpu	16.2 µs	16.4 µs	16.5 µs	15.7 µs	12
collection: intersection	v3	90.4 µs	100 µs	149 µs	87.0 µs	6932
collection: intersection	gpu	3.20 µs	3.24 µs	3.52 µs	3.14 µs	48
collection: difference	v3	92.3 µs	105 µs	203 µs	89.4 µs	6614
collection: difference	gpu	3.16 µs	3.18 µs	3.45 µs	3.08 µs	50
collection: symmetricDifference	v3	25.7 µs	26.2 µs	27.1 µs	25.2 µs	12
collection: symmetricDifference	gpu	6.67 µs	6.70 µs	7.00 µs	6.56 µs	21
collection: absoluteComplement	v3	6.37 ms	6.93 ms	10.7 ms	6.18 ms	97
collection: absoluteComplement	gpu	409 µs	419 µs	538 µs	398 µs	1670
collection: byGroup	v3	1.00 ms	1.02 ms	1.17 ms	984 µs	685
collection: byGroup	gpu	326 µs	336 µs	549 µs	321 µs	2080
collection: diff	v3	37.7 µs	38.2 µs	39.9 µs	36.3 µs	12
collection: diff	gpu	11.4 µs	11.4 µs	11.8 µs	9.83 µs	12
collection: indexOf	v3	31.6 ns	33.0 ns	65.9 ns	31.0 ns	5199
collection: indexOf	gpu	23.4 ns	24.2 ns	36.6 ns	22.6 ns	7062
traversal: openNeighborhood + closedNeighborhood (2 calls)	v3	9.45 µs	9.58 µs	10.1 µs	9.31 µs	14
traversal: openNeighborhood + closedNeighborhood (2 calls)	gpu	3.17 µs	3.39 µs	3.98 µs	3.04 µs	46
traversal: connectedNodes	v3	1.01 µs	1.10 µs	1.85 µs	986 ns	151
traversal: connectedNodes	gpu	301 ns	307 ns	371 ns	291 ns	552
traversal: sources + targets (2 calls)	v3	1.69 µs	1.88 µs	2.84 µs	1.64 µs	87
traversal: sources + targets (2 calls)	gpu	520 ns	740 ns	2.50 µs	508 ns	226
traversal: source() + target() (2 calls)	v3	77.3 ns	80.0 ns	132 ns	62.2 ns	2136
traversal: source() + target() (2 calls)	gpu	56.8 ns	58.8 ns	88.9 ns	54.7 ns	2906
traversal: edgesWith	v3	67.7 µs	69.1 µs	104 µs	65.6 µs	10079
traversal: edgesWith	gpu	53.2 µs	53.6 µs	55.2 µs	51.8 µs	12
traversal: edgesTo	v3	67.6 µs	69.7 µs	113 µs	65.4 µs	9990
traversal: edgesTo	gpu	53.7 µs	54.6 µs	55.8 µs	52.6 µs	12
traversal: parallelEdges + codirectedEdges (2 calls)	v3	3.59 µs	3.83 µs	4.61 µs	3.36 µs	41
traversal: parallelEdges + codirectedEdges (2 calls)	gpu	988 ns	1.20 µs	2.00 µs	978 ns	139
traversal: roots + leaves (2 calls)	v3	5.04 ms	5.43 ms	8.86 ms	4.89 ms	125
traversal: roots + leaves (2 calls)	gpu	392 µs	402 µs	607 µs	382 µs	1740
traversal: allAreNeighbors	v3	202 µs	218 µs	318 µs	197 µs	3174
traversal: allAreNeighbors	gpu	37.3 µs	37.7 µs	38.3 µs	36.9 µs	12
traversal: isLoop + isSimple (2 calls)	v3	135 ns	139 ns	211 ns	131 ns	1233
traversal: isLoop + isSimple (2 calls)	gpu	58.7 ns	60.7 ns	96.3 ns	57.5 ns	2818
degree: indegree + outdegree (2 calls)	v3	3.51 µs	3.84 µs	5.40 µs	3.31 µs	40
degree: indegree + outdegree (2 calls)	gpu	46.4 ns	48.2 ns	80.0 ns	45.0 ns	3536
degree: minDegree + maxDegree (2 calls)	v3	16.4 ms	17.4 ms	20.5 ms	15.2 ms	36
degree: minDegree + maxDegree (2 calls)	gpu	388 µs	398 µs	591 µs	376 µs	1760
degree: minIndegree + maxOutdegree (2 calls)	v3	9.17 ms	10.0 ms	13.6 ms	8.78 ms	66
degree: minIndegree + maxOutdegree (2 calls)	gpu	376 µs	386 µs	516 µs	370 µs	1823
dimensions: renderedPosition get	v3	74.5 ns	79.7 ns	184 ns	73.0 ns	2127
dimensions: renderedPosition get	gpu	29.8 ns	33.4 ns	105 ns	29.0 ns	5050
dimensions: relativePosition get	v3	838 ns	847 ns	956 ns	817 ns	196
dimensions: relativePosition get	gpu	77.9 ns	82.1 ns	159 ns	74.5 ns	2061
dimensions: width + height + outerWidth + outerHeight (4 calls)	v3	611 ns	613 ns	662 ns	609 ns	277
dimensions: width + height + outerWidth + outerHeight (4 calls)	gpu	212 ns	216 ns	242 ns	211 ns	785
dimensions: rendered dims (4 calls)	v3	707 ns	716 ns	943 ns	692 ns	235
dimensions: rendered dims (4 calls)	gpu	228 ns	233 ns	274 ns	227 ns	729
dimensions: renderedBoundingBox	v3	1.71 µs	1.75 µs	2.11 µs	1.66 µs	92
dimensions: renderedBoundingBox	gpu	935 ns	935 ns	993 ns	895 ns	177
dimensions: midpoint + renderedMidpoint (2 calls)	gpu	1.17 µs	1.18 µs	1.24 µs	1.14 µs	136
dimensions: source/targetEndpoint (2 calls)	gpu	377 ns	385 ns	444 ns	375 ns	440
dimensions: rendered source/targetEndpoint (2 calls)	gpu	399 ns	403 ns	467 ns	391 ns	421
dimensions: shift + silentShift (round-trip)	v3	299 µs	309 µs	611 µs	294 µs	2259
dimensions: shift + silentShift (round-trip)	gpu	4.82 µs	4.84 µs	4.98 µs	4.77 µs	31
dimensions: silentPosition	v3	874 ns	884 ns	990 ns	843 ns	190
dimensions: silentPosition	gpu	56.6 ns	59.3 ns	96.7 ns	54.4 ns	2874
dimensions: positions( fn ) over 100	v3	178 µs	183 µs	269 µs	175 µs	3820
dimensions: positions( fn ) over 100	gpu	3.53 µs	3.54 µs	3.75 µs	3.45 µs	44
dimensions: padding + paddedWidth + paddedHeight (3 calls)	v3	2.04 µs	2.06 µs	2.26 µs	1.98 µs	80
dimensions: padding + paddedWidth + paddedHeight (3 calls)	gpu	153 ns	156 ns	248 ns	148 ns	1097
curves: isBundledBezier	gpu	33.5 ns	35.0 ns	62.8 ns	20.9 ns	4873
curves: controlPoints + renderedControlPoints (2 calls)	gpu	534 ns	543 ns	719 ns	513 ns	302
curves: segmentPoints (undefined for bezier)	gpu	41.6 ns	42.8 ns	54.4 ns	36.7 ns	4007
compound: parent + parents + ancestors (3 calls)	v3	2.91 µs	3.31 µs	4.45 µs	2.81 µs	48
compound: parent + parents + ancestors (3 calls)	gpu	792 ns	1.11 µs	2.44 µs	769 ns	145
compound: children + descendants + siblings (3 calls)	v3	8.60 µs	8.53 µs	8.81 µs	7.71 µs	16
compound: children + descendants + siblings (3 calls)	gpu	1.81 µs	2.33 µs	3.74 µs	1.73 µs	68
compound: orphans + nonorphans (2 calls)	v3	3.67 µs	3.94 µs	4.72 µs	3.54 µs	39
compound: orphans + nonorphans (2 calls)	gpu	523 ns	769 ns	3.18 µs	521 ns	210
compound: commonAncestors	v3	9.61 µs	9.66 µs	9.90 µs	9.32 µs	14
compound: commonAncestors	gpu	75.6 ns	78.9 ns	134 ns	74.1 ns	2159
compound: isParent/isChild/isChildless/isOrphan (4 calls)	v3	70.6 ns	72.2 ns	83.9 ns	61.3 ns	2368
compound: isParent/isChild/isChildless/isOrphan (4 calls)	gpu	45.1 ns	45.9 ns	57.8 ns	44.2 ns	3715
flags: select + unselect + selected (band)	v3	196 µs	203 µs	330 µs	189 µs	3448
flags: select + unselect + selected (band)	gpu	12.5 µs	12.5 µs	12.6 µs	12.5 µs	10
flags: selectify + unselectify + selectable	v3	196 µs	207 µs	398 µs	190 µs	3381
flags: selectify + unselectify + selectable	gpu	2.03 µs	2.04 µs	2.15 µs	1.99 µs	79
flags: lock + unlock + locked	v3	190 µs	198 µs	317 µs	186 µs	3530
flags: lock + unlock + locked	gpu	2.04 µs	2.05 µs	2.17 µs	2.01 µs	80
flags: grabify + ungrabify + grabbable	v3	192 µs	200 µs	392 µs	188 µs	3492
flags: grabify + ungrabify + grabbable	gpu	2.03 µs	2.04 µs	2.20 µs	2.00 µs	81
flags: show + hide + visible	v3	2.57 ms	2.74 ms	4.60 ms	2.47 ms	251
flags: show + hide + visible	gpu	11.4 µs	11.4 µs	11.8 µs	11.2 µs	11
flags: active/activate/unactivate (3 calls)	v3	196 µs	201 µs	408 µs	194 µs	3475
flags: active/activate/unactivate (3 calls)	gpu	13.1 µs	13.1 µs	13.2 µs	13.0 µs	9
flags: panify + unpanify + pannable	v3	190 µs	195 µs	289 µs	187 µs	3567
flags: panify + unpanify + pannable	gpu	2.02 µs	2.03 µs	2.22 µs	1.97 µs	81
flags: takesUpSpace + interactive + transparent (3 calls)	v3	195 ns	201 ns	246 ns	195 ns	853
flags: takesUpSpace + interactive + transparent (3 calls)	gpu	69.7 ns	72.0 ns	93.0 ns	68.1 ns	2376
flags: effectiveOpacity	v3	51.6 ns	53.2 ns	98.0 ns	47.7 ns	3225
flags: effectiveOpacity	gpu	25.1 ns	25.7 ns	38.6 ns	24.1 ns	6677
element data: scratch get + set	v3	1.31 µs	1.33 µs	1.46 µs	1.28 µs	125
element data: scratch get + set	gpu	171 ns	175 ns	237 ns	170 ns	978
element data: json() one element	v3	301 ns	387 ns	3.56 µs	293 ns	422
element data: json() one element	gpu	150 ns	194 ns	538 ns	147 ns	825
element data: jsons() over 100	v3	39.8 µs	40.3 µs	42.9 µs	36.3 µs	12
element data: jsons() over 100	gpu	20.6 µs	21.3 µs	24.4 µs	16.0 µs	12
element data: label()	gpu	15.1 ns	15.9 ns	27.6 ns	14.6 ns	10734
core: batchData()	v3	2.47 µs	2.57 µs	3.18 µs	2.35 µs	61
core: batchData()	gpu	285 ns	376 ns	2.27 µs	281 ns	437
core: animated()	v3	7.53 ns	7.89 ns	14.4 ns	7.22 ns	21457
core: animated()	gpu	1.53 ns	1.61 ns	2.57 ns	1.50 ns	102339
core: zoomRange get	gpu	2.16 ns	2.29 ns	6.15 ns	2.01 ns	72839
core: forceRender() (headless no-op)	v3	23.3 ns	23.9 ns	36.8 ns	22.5 ns	7151
core: forceRender() (headless no-op)	gpu	1.48 ns	1.55 ns	2.10 ns	1.45 ns	106429
core: resize() / invalidateSize()	v3	286 ns	289 ns	357 ns	280 ns	586
core: resize() / invalidateSize()	gpu	125 ns	130 ns	197 ns	124 ns	1311
core: onRender + offRender	v3	454 ns	474 ns	827 ns	441 ns	354
core: onRender + offRender	gpu	192 ns	234 ns	685 ns	187 ns	716
core: removeScratch()	v3	2.64 ms	2.70 ms	3.17 ms	2.62 ms	253
core: removeScratch()	gpu	278 ns	284 ns	356 ns	272 ns	599
core: selectionType get	v3	4.51 ns	4.82 ns	9.05 ns	4.35 ns	35091
core: selectionType get	gpu	2.26 ns	2.40 ns	6.24 ns	2.10 ns	69540
core: isReady()	v3	3.26 ns	3.40 ns	6.69 ns	3.18 ns	49427
core: isReady()	gpu	1.48 ns	1.55 ns	2.49 ns	1.48 ns	105505
core: window()	v3	4.10 ns	4.34 ns	8.04 ns	4.10 ns	38817
core: window()	gpu	1.48 ns	1.54 ns	2.23 ns	1.48 ns	107089
core: container()	v3	3.93 ns	4.06 ns	7.55 ns	3.82 ns	41546
core: container()	gpu	1.51 ns	1.59 ns	3.02 ns	1.48 ns	103395
core: destroyed()	v3	3.31 ns	3.42 ns	7.11 ns	3.19 ns	49128
core: destroyed()	gpu	1.48 ns	1.53 ns	2.42 ns	1.48 ns	107443
core: removeAllListeners()	v3	13.8 ns	14.6 ns	30.2 ns	13.3 ns	11659
core: removeAllListeners()	gpu	1.58 ns	1.67 ns	3.03 ns	1.58 ns	99000
collection: cy()	v3	7.58 ns	8.14 ns	15.3 ns	7.32 ns	20893
collection: cy()	gpu	2.16 ns	2.40 ns	6.09 ns	2.10 ns	69440
collection: indexOfId()	v3	66.0 ns	69.0 ns	122 ns	64.5 ns	2482
collection: indexOfId()	gpu	37.5 ns	38.5 ns	54.7 ns	36.2 ns	4444
collection: removed() + inside() (2 calls)	v3	17.7 ns	18.6 ns	31.5 ns	17.0 ns	9199
collection: removed() + inside() (2 calls)	gpu	15.2 ns	15.8 ns	26.7 ns	14.6 ns	10840
collection: silentPositions( fn ) over 100	v3	111 µs	116 µs	233 µs	109 µs	6035
collection: silentPositions( fn ) over 100	gpu	3.58 µs	3.61 µs	3.80 µs	3.51 µs	43
collection: animated()	v3	12.2 ns	12.6 ns	22.8 ns	11.9 ns	13639
collection: animated()	gpu	4.19 ns	4.34 ns	7.96 ns	4.08 ns	39006
collection: delayAnimation()	v3	456 ns	592 ns	2.10 µs	453 ns	282
collection: delayAnimation()	gpu	164 ns	171 ns	285 ns	158 ns	983
collection: removeScratch()	v3	1.81 µs	1.84 µs	2.11 µs	1.76 µs	87
collection: removeScratch()	gpu	205 ns	209 ns	299 ns	197 ns	815
collection: component()	v3	80.4 ms	82.8 ms	85.7 ms	76.0 ms	12
collection: component()	gpu	1.83 ms	2.07 ms	5.36 ms	1.78 ms	334
collection: layoutDimensions()	v3	624 ns	638 ns	864 ns	609 ns	265
collection: layoutDimensions()	gpu	174 ns	181 ns	316 ns	170 ns	939
degree: maxIndegree + minOutdegree (2 calls)	v3	9.23 ms	10.1 ms	14.3 ms	8.78 ms	66
degree: maxIndegree + minOutdegree (2 calls)	gpu	381 µs	390 µs	583 µs	371 µs	1789
flags: grabbed + hidden + inactive (3 calls)	v3	85.1 ns	88.2 ns	177 ns	80.5 ns	1939
flags: grabbed + hidden + inactive (3 calls)	gpu	38.8 ns	40.2 ns	63.8 ns	36.2 ns	4264
curves: renderedSegmentPoints	gpu	36.2 ns	37.4 ns	69.1 ns	34.6 ns	4578
layout contract: edgeSlots + endpoints + degreeOf	gpu	34.7 µs	36.9 µs	64.8 µs	29.7 µs	18648
formats: isColumnarElements + isSerializedElements + isPackedIds	gpu	4.61 ns	4.80 ns	8.96 ns	4.51 ns	35261

compaction N=20,000 · 54.3 s

cpu pick: background miss (full highWater walk) — peak

222 µs

cpu pick: background miss (full highWater walk) — compacted

50.3 µs

cpu pick: node hit — peak

222 µs

cpu pick: node hit — compacted

49.7 µs

isCurrent: current ref (forwards present vs absent) — no forwards (peak)

5.95 ns

isCurrent: current ref (forwards present vs absent) — forwards present (compacted)

5.93 ns

isCurrent: stale-ref repair (chase + rewrite, alloc included) — current (baseline alloc)

8.89 ns

isCurrent: stale-ref repair (chase + rewrite, alloc included) — stale → repair

48.3 ns

control: nodes({ selected }) scan (order-list, expected parity) — peak

23.6 µs

control: nodes({ selected }) scan (order-list, expected parity) — compacted

22.1 µs

control: whole-graph boundingBox (expected parity) — peak

111 µs

control: whole-graph boundingBox (expected parity) — compacted

110 µs

1 ns10 ns100 ns1 µs10 µs100 µs1 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
cpu pick: background miss (full highWater walk)	peak	222 µs	225 µs	269 µs	216 µs	3119
cpu pick: background miss (full highWater walk)	compacted	50.3 µs	50.7 µs	51.5 µs	49.7 µs	12
cpu pick: node hit	peak	222 µs	228 µs	276 µs	221 µs	3082
cpu pick: node hit	compacted	49.7 µs	50.6 µs	55.1 µs	49.1 µs	12
isCurrent: current ref (forwards present vs absent)	no forwards (peak)	5.95 ns	6.10 ns	10.6 ns	5.74 ns	27746
isCurrent: current ref (forwards present vs absent)	forwards present (compacted)	5.93 ns	6.10 ns	10.4 ns	5.69 ns	27759
isCurrent: stale-ref repair (chase + rewrite, alloc included)	current (baseline alloc)	8.89 ns	9.85 ns	23.7 ns	8.43 ns	17110
isCurrent: stale-ref repair (chase + rewrite, alloc included)	stale → repair	48.3 ns	50.1 ns	78.5 ns	47.6 ns	3412
control: nodes({ selected }) scan (order-list, expected parity)	peak	23.6 µs	26.8 µs	34.6 µs	21.2 µs	12
control: nodes({ selected }) scan (order-list, expected parity)	compacted	22.1 µs	22.9 µs	24.7 µs	21.2 µs	12
control: whole-graph boundingBox (expected parity)	peak	111 µs	117 µs	232 µs	110 µs	5982
control: whole-graph boundingBox (expected parity)	compacted	110 µs	115 µs	207 µs	106 µs	6094

compound N=2,000 · 27.8 sv4 (gpu) v3

parent drag (shift subtree + bb settle)

186×

child drag (move + parent re-derive)

52×

reparent round-trip

78×

1 µs10 µs100 µs1 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
parent drag (shift subtree + bb settle)	v3	256 µs	289 µs	719 µs	237 µs	2408
parent drag (shift subtree + bb settle)	gpu	1.37 µs	1.44 µs	2.21 µs	1.33 µs	110
child drag (move + parent re-derive)	v3	81.4 µs	85.6 µs	134 µs	77.5 µs	8036
child drag (move + parent re-derive)	gpu	1.57 µs	1.59 µs	1.91 µs	1.53 µs	101
reparent round-trip	v3	84.4 µs	91.4 µs	163 µs	81.9 µs	7534
reparent round-trip	gpu	1.08 µs	1.09 µs	1.24 µs	1.05 µs	147

curves N=2,000 · 1.7 s

curve premium: build (style apply + derivation) — curved

11.1 ms

curve premium: build (style apply + derivation) — straight

8.03 ms

curve premium: node.position x 1000 (drag) — curved

388 µs

curve premium: node.position x 1000 (drag) — straight

299 µs

curve premium: nodes().positions() (bulk move) — curved

313 µs

curve premium: nodes().positions() (bulk move) — straight

296 µs

curve premium: first read after a bulk move (cold) — curved

5.95 ms

curve premium: first read after a bulk move (cold) — straight

2.88 ms

curve premium: same read again (warm) — curved

2.19 ms

curve premium: same read again (warm) — straight

1.39 ms

curve premium: edge.midpoint x 4000 — curved

2.16 ms

curve premium: edge.midpoint x 4000 — straight

4.16 ms

curve premium: edge.controlPoints x 4000 — curved

2.20 ms

curve premium: edge.controlPoints x 4000 — straight

1.39 ms

curve premium: cy.elements().boundingBox() x 20 — curved

35.6 ms

curve premium: cy.elements().boundingBox() x 20 — straight

21.7 ms

curve premium: cy.fit() x 20 (store scan) — curved

3.88 ms

curve premium: cy.fit() x 20 (store scan) — straight

3.61 ms

curve premium: elementsInBox x 20 (half the graph) — curved

24.6 ms

curve premium: elementsInBox x 20 (half the graph) — straight

5.91 ms

curve premium: edge.hide/show x 1000 (bundle re-fan) — curved

3.81 ms

curve premium: edge.hide/show x 1000 (bundle re-fan) — straight

2.24 ms

100 µs1 ms10 ms100 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
curve premium: build (style apply + derivation)	curved	11.1 ms	11.1 ms	11.1 ms	11.1 ms	1
curve premium: build (style apply + derivation)	straight	8.03 ms	8.03 ms	8.03 ms	8.03 ms	1
curve premium: node.position x 1000 (drag)	curved	388 µs	388 µs	388 µs	388 µs	1
curve premium: node.position x 1000 (drag)	straight	299 µs	299 µs	299 µs	299 µs	1
curve premium: nodes().positions() (bulk move)	curved	313 µs	313 µs	313 µs	313 µs	1
curve premium: nodes().positions() (bulk move)	straight	296 µs	296 µs	296 µs	296 µs	1
curve premium: first read after a bulk move (cold)	curved	5.95 ms	5.95 ms	5.95 ms	5.95 ms	1
curve premium: first read after a bulk move (cold)	straight	2.88 ms	2.88 ms	2.88 ms	2.88 ms	1
curve premium: same read again (warm)	curved	2.19 ms	2.19 ms	2.19 ms	2.19 ms	1
curve premium: same read again (warm)	straight	1.39 ms	1.39 ms	1.39 ms	1.39 ms	1
curve premium: edge.midpoint x 4000	curved	2.16 ms	2.16 ms	2.16 ms	2.16 ms	1
curve premium: edge.midpoint x 4000	straight	4.16 ms	4.16 ms	4.16 ms	4.16 ms	1
curve premium: edge.controlPoints x 4000	curved	2.20 ms	2.20 ms	2.20 ms	2.20 ms	1
curve premium: edge.controlPoints x 4000	straight	1.39 ms	1.39 ms	1.39 ms	1.39 ms	1
curve premium: cy.elements().boundingBox() x 20	curved	35.6 ms	35.6 ms	35.6 ms	35.6 ms	1
curve premium: cy.elements().boundingBox() x 20	straight	21.7 ms	21.7 ms	21.7 ms	21.7 ms	1
curve premium: cy.fit() x 20 (store scan)	curved	3.88 ms	3.88 ms	3.88 ms	3.88 ms	1
curve premium: cy.fit() x 20 (store scan)	straight	3.61 ms	3.61 ms	3.61 ms	3.61 ms	1
curve premium: elementsInBox x 20 (half the graph)	curved	24.6 ms	24.6 ms	24.6 ms	24.6 ms	1
curve premium: elementsInBox x 20 (half the graph)	straight	5.91 ms	5.91 ms	5.91 ms	5.91 ms	1
curve premium: edge.hide/show x 1000 (bundle re-fan)	curved	3.81 ms	3.81 ms	3.81 ms	3.81 ms	1
curve premium: edge.hide/show x 1000 (bundle re-fan)	straight	2.24 ms	2.24 ms	2.24 ms	2.24 ms	1

arrows N=2,000 · 8.7 s

arrow premium: build (style apply + arrow fold) — arrowed

5.77 ms

arrow premium: build (style apply + arrow fold) — none

6.48 ms

arrow premium: cy.style() re-apply x 4 — arrowed

8.14 ms

arrow premium: cy.style() re-apply x 4 — none

7.82 ms

arrow premium: node.position x 200000 (drag) — arrowed

21.6 ms

arrow premium: node.position x 200000 (drag) — none

21.6 ms

arrow premium: edge.sourceEndpoint x 200000 — arrowed

60.5 ms

arrow premium: edge.sourceEndpoint x 200000 — none

58.5 ms

arrow premium: edge.targetEndpoint x 200000 — arrowed

60.4 ms

arrow premium: edge.targetEndpoint x 200000 — none

59.3 ms

arrow premium: edge.midpoint x 200000 — arrowed

144 ms

arrow premium: edge.midpoint x 200000 — none

138 ms

arrow premium: edge.style('target-arrow-shape') x 200000 — arrowed

33.7 ms

arrow premium: edge.style('target-arrow-shape') x 200000 — none

28.3 ms

arrow premium: cy.elements().boundingBox() x 20 — arrowed

21.0 ms

arrow premium: cy.elements().boundingBox() x 20 — none

21.0 ms

arrow premium: elementsInBox x 20 (half the graph) — arrowed

5.84 ms

arrow premium: elementsInBox x 20 (half the graph) — none

5.76 ms

arrow premium: control: sourceEndpoint, bezier vs straight — arrowed

75.2 ms

arrow premium: control: sourceEndpoint, bezier vs straight — none

59.3 ms

arrow premium: edge column bytes — B/edge

156 ms

1 ms10 ms100 ms1 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
arrow premium: build (style apply + arrow fold)	arrowed	5.77 ms	5.77 ms	5.77 ms	5.77 ms	1
arrow premium: build (style apply + arrow fold)	none	6.48 ms	6.48 ms	6.48 ms	6.48 ms	1
arrow premium: cy.style() re-apply x 4	arrowed	8.14 ms	8.14 ms	8.14 ms	8.14 ms	1
arrow premium: cy.style() re-apply x 4	none	7.82 ms	7.82 ms	7.82 ms	7.82 ms	1
arrow premium: node.position x 200000 (drag)	arrowed	21.6 ms	21.6 ms	21.6 ms	21.6 ms	1
arrow premium: node.position x 200000 (drag)	none	21.6 ms	21.6 ms	21.6 ms	21.6 ms	1
arrow premium: edge.sourceEndpoint x 200000	arrowed	60.5 ms	60.5 ms	60.5 ms	60.5 ms	1
arrow premium: edge.sourceEndpoint x 200000	none	58.5 ms	58.5 ms	58.5 ms	58.5 ms	1
arrow premium: edge.targetEndpoint x 200000	arrowed	60.4 ms	60.4 ms	60.4 ms	60.4 ms	1
arrow premium: edge.targetEndpoint x 200000	none	59.3 ms	59.3 ms	59.3 ms	59.3 ms	1
arrow premium: edge.midpoint x 200000	arrowed	144 ms	144 ms	144 ms	144 ms	1
arrow premium: edge.midpoint x 200000	none	138 ms	138 ms	138 ms	138 ms	1
arrow premium: edge.style('target-arrow-shape') x 200000	arrowed	33.7 ms	33.7 ms	33.7 ms	33.7 ms	1
arrow premium: edge.style('target-arrow-shape') x 200000	none	28.3 ms	28.3 ms	28.3 ms	28.3 ms	1
arrow premium: cy.elements().boundingBox() x 20	arrowed	21.0 ms	21.0 ms	21.0 ms	21.0 ms	1
arrow premium: cy.elements().boundingBox() x 20	none	21.0 ms	21.0 ms	21.0 ms	21.0 ms	1
arrow premium: elementsInBox x 20 (half the graph)	arrowed	5.84 ms	5.84 ms	5.84 ms	5.84 ms	1
arrow premium: elementsInBox x 20 (half the graph)	none	5.76 ms	5.76 ms	5.76 ms	5.76 ms	1
arrow premium: control: sourceEndpoint, bezier vs straight	arrowed	75.2 ms	75.2 ms	75.2 ms	75.2 ms	1
arrow premium: control: sourceEndpoint, bezier vs straight	none	59.3 ms	59.3 ms	59.3 ms	59.3 ms	1
arrow premium: edge column bytes	B/edge	156 ms	156 ms	156 ms	156 ms	1

labels N=20,000 · 1.6 s

labels: breakLines x 20000 — gpu

84.9 ms

labels: estimateBlock x 20000 — gpu

94.3 ms

labels: addNode x 20000 — gpu

29.4 ms

labels: setLabel (build) x 20000 — gpu

98.3 ms

labels: setLabel (rewrite) x 20000 — gpu

101 ms

labels: boundingBox with 20000 label terms — gpu

3.01 ms

1 ms10 ms100 ms1 s

All stats (table view)

group	bench	p50	avg	p99	min	samples
labels: breakLines x 20000	gpu	84.9 ms	84.9 ms	84.9 ms	84.9 ms	1
labels: estimateBlock x 20000	gpu	94.3 ms	94.3 ms	94.3 ms	94.3 ms	1
labels: addNode x 20000	gpu	29.4 ms	29.4 ms	29.4 ms	29.4 ms	1
labels: setLabel (build) x 20000	gpu	98.3 ms	98.3 ms	98.3 ms	98.3 ms	1
labels: setLabel (rewrite) x 20000	gpu	101 ms	101 ms	101 ms	101 ms	1
labels: boundingBox with 20000 label terms	gpu	3.01 ms	3.01 ms	3.01 ms	3.01 ms	1

transitions N=2,000 · 28.8 s

transition: auto-extent shift, whole-channel re-derive (2000 slots) — transitions off

1.16 ms

transition: auto-extent shift, whole-channel re-derive (2000 slots) — transitions on

3.66 ms

transition: bulk tween tick (2000 slots, one manager tick) — tick

130 µs

transition: explicit-domain write (O(changed) — one element) — transitions off

15.1 µs

transition: explicit-domain write (O(changed) — one element) — transitions on

18.5 µs

transition: sheet swap, whole-graph apply (2000 slots diff + tween) — transitions off

2.12 ms

transition: sheet swap, whole-graph apply (2000 slots diff + tween) — transitions on

4.91 ms

10 µs100 µs1 ms10 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
transition: auto-extent shift, whole-channel re-derive (2000 slots)	transitions off	1.16 ms	1.20 ms	2.18 ms	1.13 ms	574
transition: auto-extent shift, whole-channel re-derive (2000 slots)	transitions on	3.66 ms	3.85 ms	5.53 ms	3.43 ms	175
transition: bulk tween tick (2000 slots, one manager tick)	tick	130 µs	134 µs	240 µs	127 µs	5194
transition: explicit-domain write (O(changed) — one element)	transitions off	15.1 µs	16.0 µs	20.2 µs	14.5 µs	12
transition: explicit-domain write (O(changed) — one element)	transitions on	18.5 µs	18.7 µs	19.4 µs	17.8 µs	12
transition: sheet swap, whole-graph apply (2000 slots diff + tween)	transitions off	2.12 ms	2.14 ms	2.49 ms	2.06 ms	323
transition: sheet swap, whole-graph apply (2000 slots diff + tween)	transitions on	4.91 ms	4.51 ms	6.70 ms	3.50 ms	149

geometry-tween N=2,000 · 32.7 s

size tick: unlabelled (2000 slots) — tick

544 µs

size tick: labelled, center/center (re-anchor early-out) (2000 slots) — tick

589 µs

size tick: labelled, hanging (sidecar rewrite per tick) (2000 slots) — tick

1.22 ms

edge width tick (4000 slots) — bare

290 µs

edge width tick (4000 slots) — with rides (casing + match-line arrows)

753 µs

padding tick + auto-bounds flush (250 parents × 8 children) — tick + flush

173 µs

font-size tick (2000 labels) — wrap none (dims scale-patch)

359 µs

font-size tick (2000 labels) — wrapped (re-break per tick)

5.75 ms

baseline: paint (opacity) tick (2000 slots) — tick

137 µs

100 µs1 ms10 ms

All stats (table view)

group	bench	p50	avg	p99	min	samples
size tick: unlabelled (2000 slots)	tick	544 µs	570 µs	884 µs	519 µs	1207
size tick: labelled, center/center (re-anchor early-out) (2000 slots)	tick	589 µs	612 µs	940 µs	562 µs	1145
size tick: labelled, hanging (sidecar rewrite per tick) (2000 slots)	tick	1.22 ms	1.31 ms	2.20 ms	1.10 ms	515
edge width tick (4000 slots)	bare	290 µs	303 µs	553 µs	279 µs	2298
edge width tick (4000 slots)	with rides (casing + match-line arrows)	753 µs	768 µs	1.04 ms	723 µs	904
padding tick + auto-bounds flush (250 parents × 8 children)	tick + flush	173 µs	182 µs	376 µs	168 µs	3840
font-size tick (2000 labels)	wrap none (dims scale-patch)	359 µs	399 µs	903 µs	345 µs	1737
font-size tick (2000 labels)	wrapped (re-break per tick)	5.75 ms	5.83 ms	6.52 ms	5.47 ms	115
baseline: paint (opacity) tick (2000 slots)	tick	137 µs	142 µs	241 µs	134 µs	4909