Unify indexed machine execution

[SandroMaglione]

Summary

• run eligible flat, compound, and parallel machines on one indexed execution representation
• retain a specialized single-root dispatch loop with owner-local flat state slots instead of the previous Set/Map configuration executor
• preserve the general planner fallback and add explicit flat resume-through-terminal differential coverage

Changeset

• Added or updated for a library or package-metadata change
• Not required because this PR does not change src/ or package.json

Validation

• pnpm check
• Relevant example checks, when examples changed (not applicable)
• Reviewed the automated type-performance report, when the public TypeScript API or inference changed (no public type change; pnpm perf:types passes)
• Reviewed the automated runtime-performance report, when runtime behavior changed

Local stabilized five-process confirmation measured +8.2% flat burst, +4.3% observed burst, +4.8% child send, +3.9% compound burst, and +2.1% parallel burst. Full-size paired runs showed a larger flat signal but warmed significantly, so the hosted five-process report is the merge gate.

[github-actions]

Type performance

Measured with TypeScript 6.0.3 and skipLibCheck=true.

Scenario	Base	PR	Difference
Effect only	55	55	0 (0.0%)
Import effect-machine	55	55	0 (0.0%)
Machine.defineStates (3 states)	2,965	2,965	0 (0.0%)
Machine.make (3 states, 2 events)	8,715	8,715	0 (0.0%)
machine.handle (3 states, 2 transitions)	23,492	23,492	0 (0.0%)
machine.handle (depth 8)	117,075	117,075	0 (0.0%)
machine.handle (depth 12)	129,713	129,713	0 (0.0%)
machine.handle (depth 16)	145,055	145,055	0 (0.0%)
machine.handle (depth 24)	183,851	183,851	0 (0.0%)
machine.handle (wide depth 16)	214,189	214,189	0 (0.0%)
machine.handle (parallel/history/choice)	132,667	132,667	0 (0.0%)
machine.handle (4 successive calls)	128,674	128,674	0 (0.0%)
machine exact input/output/error/services	110,584	110,584	0 (0.0%)
execution adapter readiness	127,854	127,854	0 (0.0%)

Marginal instantiations are measured against the matching setup without that API call:

Scenario	Base	PR	Difference
Import effect-machine	0	0	0
Machine.defineStates (3 states)	2,910	2,910	0 (0.0%)
Machine.make (3 states, 2 events)	5,742	5,742	0 (0.0%)
machine.handle (3 states, 2 transitions)	14,777	14,777	0 (0.0%)
machine.handle (depth 8)	109,601	109,601	0 (0.0%)
machine.handle (depth 12)	121,039	121,039	0 (0.0%)
machine.handle (depth 16)	135,181	135,181	0 (0.0%)
machine.handle (depth 24)	171,577	171,577	0 (0.0%)
machine.handle (wide depth 16)	202,224	202,224	0 (0.0%)
machine.handle (parallel/history/choice)	114,331	114,331	0 (0.0%)
machine.handle (4 successive calls)	113,703	113,703	0 (0.0%)
machine exact input/output/error/services	100,670	100,670	0 (0.0%)
execution adapter readiness	102,092	102,092	0 (0.0%)

Check times (informational)

Scenario	Base	PR
Effect only	0.03s	0.03s
Import effect-machine	0.03s	0.03s
Machine.defineStates (3 states)	0.10s	0.10s
Machine.make (3 states, 2 events)	0.16s	0.16s
machine.handle (3 states, 2 transitions)	0.24s	0.24s
machine.handle (depth 8)	0.51s	0.52s
machine.handle (depth 12)	0.55s	0.56s
machine.handle (depth 16)	0.60s	0.60s
machine.handle (depth 24)	0.69s	0.69s
machine.handle (wide depth 16)	0.77s	0.76s
machine.handle (parallel/history/choice)	0.62s	0.63s
machine.handle (4 successive calls)	0.59s	0.58s
machine exact input/output/error/services	0.51s	0.52s
execution adapter readiness	0.57s	0.57s

Type instantiations are the comparison metric. Check time varies with runner load and is informational only.

[github-actions]

Runtime performance

Median of 5 independent benchmark processes on INTEL(R) XEON(R) PLATINUM 8573C with Node v24.18.0.

Pull request baseline

Scenario	Effect Machine	XState 5	XState 6 alpha
Plan counter transitions	170,843 transitions/s	185,955 transitions/s	151,432 transitions/s
Drain burst with terminal fence	399,132 increments/s	541,983 increments/s	141,334 increments/s
Drain burst with a change observer	250,093 increments/s	547,756 increments/s	143,017 increments/s
Lookup and send to one child	343,269 increments/s	540,523 increments/s	134,898 increments/s
Start and stop a machine	49,439 machines/s	250,752 machines/s	104,406 machines/s
Start and stop a parent with one child	19,918 families/s	96,283 families/s	42,479 families/s
Drain burst through a compound state	398,512 events/s	485,464 events/s	141,582 events/s
Drain burst through two parallel regions	389,795 events/s	358,331 events/s	117,892 events/s

Effect runtime reference points

Scenario	Effect Machine	Effect runtime primitives
Start and stop a raw compiled process	54,413 processes/s	—
Start and interrupt a suspended fiber	—	232,937 fibers/s
Start and stop a queue worker	—	170,039 workers/s
Start and stop an actor shell	—	153,610 actors/s
Start and stop two actor shells	—	95,293 families/s
Update an owner-only mutable snapshot	—	350,631,136 updates/s
Update a synchronized snapshot	—	2,142,162 updates/s
Create, resolve, and await a terminal latch	—	2,185,735 latches/s

Memory profile	Effect Machine	XState 5	XState 6 alpha	Effect runtime primitives
Idle machine	2.0 KiB	3.8 KiB	6.1 KiB	—
Raw generic managed process	12.3 KiB	—	—	—
Raw compiled process	2.7 KiB	—	—	—
Two independent idle machines	3.5 KiB	7.3 KiB	11.8 KiB	—
Idle parent with one child	5.8 KiB	5.5 KiB	8.8 KiB	—
Parent with observed child registry	10.0 KiB	—	—	—
Parent with observed invoked child snapshots	5.9 KiB	—	—	—
Suspended Effect fiber	—	—	—	0.7 KiB
Effect queue with waiting fiber	—	—	—	2.8 KiB
Effect mailbox actor shell	—	—	—	3.2 KiB
Two Effect actor shells	—	—	—	6.9 KiB

Effect Machine change from base

Metric	Base	Base variability	PR	PR variability	Difference
Plan counter transitions	174,281 transitions/s	1.4% MAD	170,843 transitions/s	0.6% MAD	-2.0%
Drain burst with terminal fence	347,225 increments/s	2.3% MAD	399,132 increments/s	3.0% MAD	+14.9%
Drain burst with a change observer	233,315 increments/s	0.3% MAD	250,093 increments/s	0.5% MAD	+7.2%
Lookup and send to one child	312,105 increments/s	2.0% MAD	343,269 increments/s	2.8% MAD	+10.0%
Start and stop a machine	49,618 machines/s	2.0% MAD	49,439 machines/s	1.4% MAD	-0.4%
Start and stop a parent with one child	20,168 families/s	2.4% MAD	19,918 families/s	1.5% MAD	-1.2%
Drain burst through a compound state	409,586 events/s	0.9% MAD	398,512 events/s	3.6% MAD	-2.7%
Drain burst through two parallel regions	384,713 events/s	3.4% MAD	389,795 events/s	1.4% MAD	+1.3%
Idle machine heap per unit	2.0 KiB	0.0% MAD	2.0 KiB	0.0% MAD	+0.0%
Raw generic managed process heap per unit	12.3 KiB	0.0% MAD	12.3 KiB	0.0% MAD	-0.0%
Raw compiled process heap per unit	2.7 KiB	0.2% MAD	2.7 KiB	0.0% MAD	-0.4%
Two independent idle machines heap per unit	3.5 KiB	0.0% MAD	3.5 KiB	0.0% MAD	+0.0%
Idle parent with one child heap per unit	5.8 KiB	0.0% MAD	5.8 KiB	0.1% MAD	-0.0%
Parent with observed child registry heap per unit	10.0 KiB	0.0% MAD	10.0 KiB	0.0% MAD	-0.0%
Parent with observed invoked child snapshots heap per unit	5.9 KiB	0.0% MAD	5.9 KiB	0.0% MAD	-0.0%

Effect runtime reference change from base

Metric	Base	Base variability	PR	PR variability	Difference
Start and stop a raw compiled process	54,315 processes/s	0.6% MAD	54,413 processes/s	0.8% MAD	+0.2%
Start and interrupt a suspended fiber	236,407 fibers/s	1.8% MAD	232,937 fibers/s	2.5% MAD	-1.5%
Start and stop a queue worker	171,380 workers/s	0.9% MAD	170,039 workers/s	1.4% MAD	-0.8%
Start and stop an actor shell	157,134 actors/s	0.9% MAD	153,610 actors/s	0.2% MAD	-2.2%
Start and stop two actor shells	89,087 families/s	3.4% MAD	95,293 families/s	1.8% MAD	+7.0%
Update an owner-only mutable snapshot	350,754,122 updates/s	0.0% MAD	350,631,136 updates/s	0.0% MAD	-0.0%
Update a synchronized snapshot	2,189,535 updates/s	1.0% MAD	2,142,162 updates/s	2.2% MAD	-2.2%
Create, resolve, and await a terminal latch	2,160,831 latches/s	1.3% MAD	2,185,735 latches/s	1.9% MAD	+1.2%

Versions and interpretation

• Effect Machine: 0.3.0
• XState 5: 5.32.5
• XState 6 alpha: 6.0.0-alpha.27
• Effect runtime primitives: 4.0.0-beta.105

Higher throughput is better; lower heap is better. Variability is the median absolute deviation across independent processes, relative to their median. Runtime measurements on shared GitHub-hosted hardware remain informational, so small differences should be confirmed across multiple workflow runs.