Skip to content

(WIP)perf(state-transition): use native PTC sampling - #9903

Draft
GrapeBaBa wants to merge 2 commits into
unstablefrom
gr/native-ptc-sampling
Draft

(WIP)perf(state-transition): use native PTC sampling#9903
GrapeBaBa wants to merge 2 commits into
unstablefrom
gr/native-ptc-sampling

Conversation

@GrapeBaBa

@GrapeBaBa GrapeBaBa commented Aug 24, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Motivation

Fixes #9771.

Gloas and Heze sanity/slots/pyspec_tests/historical_accumulator spend most of their time computing 256 epochs of PTC weighted selections. This PR moves the whole-epoch sampler to the Zig native shuffle binding from ChainSafe/lodestar-z#563.

Depends on ChainSafe/lodestar-z#563.

Changes

  • call computePtcIndicesForEpoch once per epoch with the existing flat shuffling
  • preserve the existing public helper signature and independent per-slot Uint32Array results
  • reuse retained ptcWindow SSZ child views instead of rebuilding all three epochs
  • batch post-Gloas BeaconState hashing with a reused HashComputationGroup
  • re-enable the Gloas and Heze mainnet historical accumulator cases
  • add a Lodestar-level differential test covering multi-committee slot offsets and the legacy committees-only call

Performance

Mainnet preset, consensus-specs v1.7.0-alpha.14, Node 25.2.1:

case unstable this PR
gloas historical_accumulator 26.39s 5.68s
heze historical_accumulator 26.30s 6.06s

Focused profiling reduced cumulative processPtcWindow time from about 12.27s to 1.04s over 256 epochs. The final state root is unchanged.

@github-actions

github-actions Bot commented Aug 24, 2026 •
edited
Loading

Copy link
Copy Markdown
Contributor

Performance Report

鉁旓笍 no performance regression detected

Full benchmark results
Benchmark suite Current: 79df103 Previous: 62b30e7 Ratio
getPubkeys - native cache - req 1000 vs - 250000 vc 460.20 us/op 443.50 us/op 1.04
getPubkeys - validatorsArr - req 1000 vs - 250000 vc 38.199 us/op 41.017 us/op 0.93
BLS verify - blst 855.28 us/op 909.00 us/op 0.94
BLS verifyMultipleSignatures 3 - blst 1.2934 ms/op 1.3756 ms/op 0.94
BLS verifyMultipleSignatures 8 - blst 2.1925 ms/op 2.2786 ms/op 0.96
BLS verifyMultipleSignatures 32 - blst 6.9676 ms/op 7.5322 ms/op 0.93
BLS verifyMultipleSignatures 64 - blst 13.377 ms/op 14.478 ms/op 0.92
BLS verifyMultipleSignatures 128 - blst 25.062 ms/op 28.102 ms/op 0.89
BLS deserializing 10000 signatures 619.38 ms/op 650.53 ms/op 0.95
BLS deserializing 100000 signatures 6.4097 s/op 6.5578 s/op 0.98
BLS verifyMultipleSignatures - same message - 3 - blst 949.28 us/op 966.55 us/op 0.98
BLS verifyMultipleSignatures - same message - 8 - blst 1.0657 ms/op 1.0971 ms/op 0.97
BLS verifyMultipleSignatures - same message - 32 - blst 1.6908 ms/op 1.7123 ms/op 0.99
BLS verifyMultipleSignatures - same message - 64 - blst 2.5242 ms/op 2.5534 ms/op 0.99
BLS verifyMultipleSignatures - same message - 128 - blst 4.1894 ms/op 4.2318 ms/op 0.99
BLS aggregatePubkeys 32 - blst 19.774 us/op 19.501 us/op 1.01
BLS aggregatePubkeys 128 - blst 71.518 us/op 70.387 us/op 1.02
getSlashingsAndExits - default max 45.195 us/op 40.312 us/op 1.12
getSlashingsAndExits - 2k 543.80 us/op 369.44 us/op 1.47
proposeBlockBody type=full, size=empty 748.05 us/op 853.92 us/op 0.88
isKnown best case - 1 super set check 172.00 ns/op 171.00 ns/op 1.01
isKnown normal case - 2 super set checks 178.00 ns/op 170.00 ns/op 1.05
isKnown worse case - 16 super set checks 171.00 ns/op 168.00 ns/op 1.02
validate api signedAggregateAndProof - struct 1.5031 ms/op 1.5440 ms/op 0.97
validate gossip signedAggregateAndProof - struct 1.4992 ms/op 1.5431 ms/op 0.97
batch validate gossip attestation - vc 640000 - chunk 32 128.48 us/op 129.07 us/op 1.00
batch validate gossip attestation - vc 640000 - chunk 64 108.25 us/op 109.11 us/op 0.99
batch validate gossip attestation - vc 640000 - chunk 128 100.59 us/op 98.489 us/op 1.02
batch validate gossip attestation - vc 640000 - chunk 256 107.95 us/op 99.240 us/op 1.09
bytes32 toHexString 312.00 ns/op 296.00 ns/op 1.05
bytes32 Buffer.toString(hex) 176.00 ns/op 181.00 ns/op 0.97
bytes32 Buffer.toString(hex) from Uint8Array 247.00 ns/op 250.00 ns/op 0.99
bytes32 Buffer.toString(hex) + 0x 175.00 ns/op 182.00 ns/op 0.96
Return object 10000 times 0.21580 ns/op 0.21800 ns/op 0.99
Throw Error 10000 times 3.5438 us/op 3.3667 us/op 1.05
toHex 94.154 ns/op 104.75 ns/op 0.90
Buffer.from 86.357 ns/op 92.948 ns/op 0.93
shared Buffer 58.090 ns/op 61.184 ns/op 0.95
fastMsgIdFn sha256 / 200 bytes 1.4930 us/op 1.5060 us/op 0.99
fastMsgIdFn h32 xxhash / 200 bytes 158.00 ns/op 164.00 ns/op 0.96
fastMsgIdFn h64 xxhash / 200 bytes 214.00 ns/op 215.00 ns/op 1.00
fastMsgIdFn sha256 / 1000 bytes 4.7940 us/op 4.8470 us/op 0.99
fastMsgIdFn h32 xxhash / 1000 bytes 252.00 ns/op 257.00 ns/op 0.98
fastMsgIdFn h64 xxhash / 1000 bytes 259.00 ns/op 265.00 ns/op 0.98
fastMsgIdFn sha256 / 10000 bytes 42.167 us/op 42.496 us/op 0.99
fastMsgIdFn h32 xxhash / 10000 bytes 1.2870 us/op 1.2860 us/op 1.00
fastMsgIdFn h64 xxhash / 10000 bytes 841.00 ns/op 827.00 ns/op 1.02
send data - 1000 256B messages 4.4702 ms/op 4.3512 ms/op 1.03
send data - 1000 512B messages 5.2897 ms/op 5.3499 ms/op 0.99
send data - 1000 1024B messages 6.1778 ms/op 5.7252 ms/op 1.08
send data - 1000 1200B messages 12.313 ms/op 6.5103 ms/op 1.89
send data - 1000 2048B messages 16.930 ms/op 15.136 ms/op 1.12
send data - 1000 4096B messages 156.02 ms/op 110.02 ms/op 1.42
send data - 1000 16384B messages 692.42 ms/op 434.02 ms/op 1.60
send data - 1000 65536B messages 2.6727 s/op 1.8759 s/op 1.42
enrSubnets - fastDeserialize 64 bits 796.00 ns/op 769.00 ns/op 1.04
enrSubnets - ssz BitVector 64 bits 277.00 ns/op 274.00 ns/op 1.01
enrSubnets - fastDeserialize 4 bits 102.00 ns/op 108.00 ns/op 0.94
enrSubnets - ssz BitVector 4 bits 280.00 ns/op 283.00 ns/op 0.99
prioritizePeers score -10:0 att 32-0.1 sync 2-0 245.03 us/op 209.64 us/op 1.17
prioritizePeers score 0:0 att 32-0.25 sync 2-0.25 251.59 us/op 237.48 us/op 1.06
prioritizePeers score 0:0 att 32-0.5 sync 2-0.5 361.50 us/op 344.38 us/op 1.05
prioritizePeers score 0:0 att 64-0.75 sync 4-0.75 627.16 us/op 610.43 us/op 1.03
prioritizePeers score 0:0 att 64-1 sync 4-1 727.31 us/op 719.91 us/op 1.01
array of 16000 items push then shift 1.2872 us/op 1.3325 us/op 0.97
LinkedList of 16000 items push then shift 8.3970 ns/op 7.8060 ns/op 1.08
array of 16000 items push then pop 88.860 ns/op 83.962 ns/op 1.06
LinkedList of 16000 items push then pop 6.6280 ns/op 6.2200 ns/op 1.07
array of 24000 items push then shift 1.9137 us/op 1.9753 us/op 0.97
LinkedList of 24000 items push then shift 8.0800 ns/op 7.2460 ns/op 1.12
array of 24000 items push then pop 125.63 ns/op 120.11 ns/op 1.05
LinkedList of 24000 items push then pop 7.1460 ns/op 6.2140 ns/op 1.15
intersect bitArray bitLen 8 3.8850 ns/op 3.9220 ns/op 0.99
intersect array and set length 8 30.514 ns/op 29.988 ns/op 1.02
intersect bitArray bitLen 128 23.421 ns/op 23.754 ns/op 0.99
intersect array and set length 128 518.09 ns/op 506.74 ns/op 1.02
bitArray.getTrueBitIndexes() bitLen 128 1.0980 us/op 955.00 ns/op 1.15
bitArray.getTrueBitIndexes() bitLen 248 1.8630 us/op 1.7180 us/op 1.08
bitArray.getTrueBitIndexes() bitLen 512 3.8850 us/op 3.6720 us/op 1.06
Full columns - reconstruct all 6 blobs 208.52 us/op 171.31 us/op 1.22
Full columns - reconstruct half of the blobs out of 6 99.645 us/op 77.335 us/op 1.29
Full columns - reconstruct single blob out of 6 39.597 us/op 38.550 us/op 1.03
Half columns - reconstruct all 6 blobs 410.55 ms/op 401.72 ms/op 1.02
Half columns - reconstruct half of the blobs out of 6 202.54 ms/op 199.98 ms/op 1.01
Half columns - reconstruct single blob out of 6 72.351 ms/op 71.689 ms/op 1.01
Set add up to 64 items then delete first 2.8379 us/op 1.6723 us/op 1.70
OrderedSet add up to 64 items then delete first 4.0656 us/op 2.5408 us/op 1.60
Set add up to 64 items then delete last 2.6366 us/op 1.8735 us/op 1.41
OrderedSet add up to 64 items then delete last 3.5801 us/op 2.7856 us/op 1.29
Set add up to 64 items then delete middle 2.8313 us/op 1.8737 us/op 1.51
OrderedSet add up to 64 items then delete middle 7.0566 us/op 4.2886 us/op 1.65
Set add up to 128 items then delete first 5.8647 us/op 3.7222 us/op 1.58
OrderedSet add up to 128 items then delete first 8.4653 us/op 5.7567 us/op 1.47
Set add up to 128 items then delete last 4.5193 us/op 3.5562 us/op 1.27
OrderedSet add up to 128 items then delete last 7.1752 us/op 5.4019 us/op 1.33
Set add up to 128 items then delete middle 5.1885 us/op 3.5824 us/op 1.45
OrderedSet add up to 128 items then delete middle 14.244 us/op 11.318 us/op 1.26
Set add up to 256 items then delete first 12.108 us/op 7.6690 us/op 1.58
OrderedSet add up to 256 items then delete first 17.093 us/op 12.318 us/op 1.39
Set add up to 256 items then delete last 9.4174 us/op 7.0721 us/op 1.33
OrderedSet add up to 256 items then delete last 17.584 us/op 11.576 us/op 1.52
Set add up to 256 items then delete middle 10.644 us/op 7.1115 us/op 1.50
OrderedSet add up to 256 items then delete middle 40.435 us/op 35.875 us/op 1.13
runFastConfirmationRules vc:100000 bc:96 eq:0 6.1192 ms/op 4.7644 ms/op 1.28
runFastConfirmationRules vc:600000 bc:96 eq:0 38.953 ms/op 36.610 ms/op 1.06
runFastConfirmationRules vc:1000000 bc:96 eq:0 60.957 ms/op 60.684 ms/op 1.00
runFastConfirmationRules vc:600000 bc:320 eq:0 36.209 ms/op 38.419 ms/op 0.94
runFastConfirmationRules vc:100000 bc:96 eq:1000 1.2032 s/op 1.1826 s/op 1.02
pass gossip attestations to forkchoice per slot 2.5137 ms/op 2.6238 ms/op 0.96
forkChoice updateHead vc 100000 bc 64 eq 0 387.14 us/op 419.52 us/op 0.92
forkChoice updateHead vc 600000 bc 64 eq 0 2.3920 ms/op 4.2020 ms/op 0.57
forkChoice updateHead vc 1000000 bc 64 eq 0 3.8646 ms/op 4.2272 ms/op 0.91
forkChoice updateHead vc 600000 bc 320 eq 0 2.3448 ms/op 2.5366 ms/op 0.92
forkChoice updateHead vc 600000 bc 1200 eq 0 2.3932 ms/op 2.6542 ms/op 0.90
forkChoice updateHead vc 600000 bc 7200 eq 0 3.9255 ms/op 3.0497 ms/op 1.29
forkChoice updateHead vc 600000 bc 64 eq 1000 2.3523 ms/op 2.5060 ms/op 0.94
forkChoice updateHead vc 600000 bc 64 eq 10000 2.4600 ms/op 2.6947 ms/op 0.91
forkChoice updateHead vc 600000 bc 64 eq 300000 6.9811 ms/op 7.3626 ms/op 0.95
computeDeltas 1400000 validators 0% inactive 11.943 ms/op 13.028 ms/op 0.92
computeDeltas 1400000 validators 10% inactive 11.332 ms/op 12.328 ms/op 0.92
computeDeltas 1400000 validators 20% inactive 10.486 ms/op 11.166 ms/op 0.94
computeDeltas 1400000 validators 50% inactive 8.5849 ms/op 9.2846 ms/op 0.92
computeDeltas 2100000 validators 0% inactive 17.681 ms/op 19.471 ms/op 0.91
computeDeltas 2100000 validators 10% inactive 17.120 ms/op 18.599 ms/op 0.92
computeDeltas 2100000 validators 20% inactive 15.734 ms/op 17.192 ms/op 0.92
computeDeltas 2100000 validators 50% inactive 10.563 ms/op 14.171 ms/op 0.75
altair processAttestation - 250000 vs - 7PWei normalcase 2.1616 ms/op 1.9039 ms/op 1.14
altair processAttestation - 250000 vs - 7PWei worstcase 3.2249 ms/op 2.6774 ms/op 1.20
altair processAttestation - setStatus - 1/6 committees join 102.51 us/op 104.48 us/op 0.98
altair processAttestation - setStatus - 1/3 committees join 197.86 us/op 201.60 us/op 0.98
altair processAttestation - setStatus - 1/2 committees join 293.75 us/op 292.61 us/op 1.00
altair processAttestation - setStatus - 2/3 committees join 377.33 us/op 373.27 us/op 1.01
altair processAttestation - setStatus - 4/5 committees join 542.68 us/op 534.19 us/op 1.02
altair processAttestation - setStatus - 100% committees join 650.95 us/op 630.37 us/op 1.03
altair processBlock - 250000 vs - 7PWei normalcase 5.8111 ms/op 4.7716 ms/op 1.22
altair processBlock - 250000 vs - 7PWei normalcase hashState 25.550 ms/op 21.518 ms/op 1.19
altair processBlock - 250000 vs - 7PWei worstcase 27.363 ms/op 25.179 ms/op 1.09
altair processBlock - 250000 vs - 7PWei worstcase hashState 60.005 ms/op 52.081 ms/op 1.15
phase0 processBlock - 250000 vs - 7PWei normalcase 1.5094 ms/op 1.2978 ms/op 1.16
phase0 processBlock - 250000 vs - 7PWei worstcase 21.728 ms/op 20.365 ms/op 1.07
altair processEth1Data - 250000 vs - 7PWei normalcase 313.62 us/op 296.94 us/op 1.06
getExpectedWithdrawals 250000 eb:1,eth1:1,we:0,wn:0,smpl:16 3.1350 us/op 3.1590 us/op 0.99
getExpectedWithdrawals 250000 eb:0.95,eth1:0.1,we:0.05,wn:0,smpl:220 20.474 us/op 20.233 us/op 1.01
getExpectedWithdrawals 250000 eb:0.95,eth1:0.3,we:0.05,wn:0,smpl:43 5.5060 us/op 5.8330 us/op 0.94
getExpectedWithdrawals 250000 eb:0.95,eth1:0.7,we:0.05,wn:0,smpl:19 3.7840 us/op 3.6330 us/op 1.04
getExpectedWithdrawals 250000 eb:0.1,eth1:0.1,we:0,wn:0,smpl:1021 89.333 us/op 92.295 us/op 0.97
getExpectedWithdrawals 250000 eb:0.03,eth1:0.03,we:0,wn:0,smpl:11778 1.4452 ms/op 1.4554 ms/op 0.99
getExpectedWithdrawals 250000 eb:0.01,eth1:0.01,we:0,wn:0,smpl:16384 1.9437 ms/op 1.9109 ms/op 1.02
getExpectedWithdrawals 250000 eb:0,eth1:0,we:0,wn:0,smpl:16384 2.1983 ms/op 1.9083 ms/op 1.15
getExpectedWithdrawals 250000 eb:0,eth1:0,we:0,wn:0,nocache,smpl:16384 6.1997 ms/op 3.7452 ms/op 1.66
getExpectedWithdrawals 250000 eb:0,eth1:1,we:0,wn:0,smpl:16384 2.2173 ms/op 2.1479 ms/op 1.03
getExpectedWithdrawals 250000 eb:0,eth1:1,we:0,wn:0,nocache,smpl:16384 5.6909 ms/op 4.1255 ms/op 1.38
Tree 40 250000 create 422.59 ms/op 327.31 ms/op 1.29
Tree 40 250000 get(125000) 93.234 ns/op 96.912 ns/op 0.96
Tree 40 250000 set(125000) 1.2025 us/op 1.0679 us/op 1.13
Tree 40 250000 toArray() 17.739 ms/op 11.659 ms/op 1.52
Tree 40 250000 iterate all - toArray() + loop 19.973 ms/op 14.844 ms/op 1.35
Tree 40 250000 iterate all - get(i) 41.992 ms/op 41.583 ms/op 1.01
Array 250000 create 2.7052 ms/op 2.7041 ms/op 1.00
Array 250000 clone - spread 846.15 us/op 822.08 us/op 1.03
Array 250000 get(125000) 0.30200 ns/op 0.30400 ns/op 0.99
Array 250000 set(125000) 0.30200 ns/op 0.30600 ns/op 0.99
Array 250000 iterate all - loop 58.615 us/op 58.471 us/op 1.00
phase0 afterProcessEpoch - 250000 vs - 7PWei 39.553 ms/op 62.582 ms/op 0.63
Array.fill - length 1000000 6.0921 ms/op 4.1754 ms/op 1.46
Array push - length 1000000 16.801 ms/op 15.830 ms/op 1.06
Array.get 0.21094 ns/op 0.21205 ns/op 0.99
Uint8Array.get 0.25109 ns/op 0.23891 ns/op 1.05
phase0 beforeProcessEpoch - 250000 vs - 7PWei 21.299 ms/op 17.554 ms/op 1.21
altair processEpoch - mainnet_e81889 248.43 ms/op 218.84 ms/op 1.14
mainnet_e81889 - altair beforeProcessEpoch 24.547 ms/op 23.113 ms/op 1.06
mainnet_e81889 - altair processJustificationAndFinalization 5.8440 us/op 4.4750 us/op 1.31
mainnet_e81889 - altair processInactivityUpdates 6.3578 ms/op 3.6115 ms/op 1.76
mainnet_e81889 - altair processRewardsAndPenalties 19.633 ms/op 18.623 ms/op 1.05
mainnet_e81889 - altair processRegistryUpdates 574.00 ns/op 585.00 ns/op 0.98
mainnet_e81889 - altair processSlashings 147.00 ns/op 145.00 ns/op 1.01
mainnet_e81889 - altair processEth1DataReset 142.00 ns/op 139.00 ns/op 1.02
mainnet_e81889 - altair processEffectiveBalanceUpdates 4.8147 ms/op 1.7427 ms/op 2.76
mainnet_e81889 - altair processSlashingsReset 718.00 ns/op 719.00 ns/op 1.00
mainnet_e81889 - altair processRandaoMixesReset 988.00 ns/op 928.00 ns/op 1.06
mainnet_e81889 - altair processHistoricalRootsUpdate 166.00 ns/op 145.00 ns/op 1.14
mainnet_e81889 - altair processParticipationFlagUpdates 467.00 ns/op 448.00 ns/op 1.04
mainnet_e81889 - altair processSyncCommitteeUpdates 114.00 ns/op 114.00 ns/op 1.00
mainnet_e81889 - altair afterProcessEpoch 42.037 ms/op 42.337 ms/op 0.99
capella processEpoch - mainnet_e217614 759.92 ms/op 688.86 ms/op 1.10
mainnet_e217614 - capella beforeProcessEpoch 70.199 ms/op 67.526 ms/op 1.04
mainnet_e217614 - capella processJustificationAndFinalization 4.5800 us/op 4.6770 us/op 0.98
mainnet_e217614 - capella processInactivityUpdates 16.243 ms/op 13.203 ms/op 1.23
mainnet_e217614 - capella processRewardsAndPenalties 93.093 ms/op 95.492 ms/op 0.97
mainnet_e217614 - capella processRegistryUpdates 4.4740 us/op 4.5350 us/op 0.99
mainnet_e217614 - capella processSlashings 143.00 ns/op 144.00 ns/op 0.99
mainnet_e217614 - capella processEth1DataReset 142.00 ns/op 141.00 ns/op 1.01
mainnet_e217614 - capella processEffectiveBalanceUpdates 15.168 ms/op 9.4884 ms/op 1.60
mainnet_e217614 - capella processSlashingsReset 685.00 ns/op 703.00 ns/op 0.97
mainnet_e217614 - capella processRandaoMixesReset 947.00 ns/op 960.00 ns/op 0.99
mainnet_e217614 - capella processHistoricalRootsUpdate 143.00 ns/op 145.00 ns/op 0.99
mainnet_e217614 - capella processParticipationFlagUpdates 463.00 ns/op 455.00 ns/op 1.02
mainnet_e217614 - capella afterProcessEpoch 106.42 ms/op 113.88 ms/op 0.93
phase0 processEpoch - mainnet_e58758 222.47 ms/op 210.28 ms/op 1.06
mainnet_e58758 - phase0 beforeProcessEpoch 53.143 ms/op 48.455 ms/op 1.10
mainnet_e58758 - phase0 processJustificationAndFinalization 5.4550 us/op 4.6070 us/op 1.18
mainnet_e58758 - phase0 processRewardsAndPenalties 17.209 ms/op 16.556 ms/op 1.04
mainnet_e58758 - phase0 processRegistryUpdates 2.2180 us/op 2.3280 us/op 0.95
mainnet_e58758 - phase0 processSlashings 143.00 ns/op 141.00 ns/op 1.01
mainnet_e58758 - phase0 processEth1DataReset 139.00 ns/op 158.00 ns/op 0.88
mainnet_e58758 - phase0 processEffectiveBalanceUpdates 806.16 us/op 834.28 us/op 0.97
mainnet_e58758 - phase0 processSlashingsReset 902.00 ns/op 907.00 ns/op 0.99
mainnet_e58758 - phase0 processRandaoMixesReset 1.0170 us/op 965.00 ns/op 1.05
mainnet_e58758 - phase0 processHistoricalRootsUpdate 143.00 ns/op 143.00 ns/op 1.00
mainnet_e58758 - phase0 processParticipationRecordUpdates 1.0290 us/op 940.00 ns/op 1.09
mainnet_e58758 - phase0 afterProcessEpoch 33.888 ms/op 34.481 ms/op 0.98
phase0 processEffectiveBalanceUpdates - 250000 normalcase 972.57 us/op 1.0125 ms/op 0.96
phase0 processEffectiveBalanceUpdates - 250000 worstcase 0.5 1.5714 ms/op 1.2350 ms/op 1.27
altair processInactivityUpdates - 250000 normalcase 14.335 ms/op 11.126 ms/op 1.29
altair processInactivityUpdates - 250000 worstcase 14.854 ms/op 11.200 ms/op 1.33
phase0 processRegistryUpdates - 250000 normalcase 2.2020 us/op 2.2840 us/op 0.96
phase0 processRegistryUpdates - 250000 badcase_full_deposits 135.37 us/op 144.50 us/op 0.94
phase0 processRegistryUpdates - 250000 worstcase 0.5 58.251 ms/op 61.280 ms/op 0.95
altair processRewardsAndPenalties - 250000 normalcase 14.493 ms/op 16.376 ms/op 0.89
altair processRewardsAndPenalties - 250000 worstcase 14.232 ms/op 15.821 ms/op 0.90
phase0 getAttestationDeltas - 250000 normalcase 6.3304 ms/op 5.7507 ms/op 1.10
phase0 getAttestationDeltas - 250000 worstcase 7.7739 ms/op 5.8089 ms/op 1.34
phase0 processSlashings - 250000 worstcase 58.294 us/op 62.165 us/op 0.94
altair processSyncCommitteeUpdates - 250000 9.6010 ms/op 10.092 ms/op 0.95
BeaconState.hashTreeRoot - No change 176.00 ns/op 177.00 ns/op 0.99
BeaconState.hashTreeRoot - 1 full validator 81.162 us/op 77.599 us/op 1.05
BeaconState.hashTreeRoot - 32 full validator 873.50 us/op 889.07 us/op 0.98
BeaconState.hashTreeRoot - 512 full validator 8.2824 ms/op 8.1333 ms/op 1.02
BeaconState.hashTreeRoot - 1 validator.effectiveBalance 98.558 us/op 95.864 us/op 1.03
BeaconState.hashTreeRoot - 32 validator.effectiveBalance 1.4622 ms/op 1.4311 ms/op 1.02
BeaconState.hashTreeRoot - 512 validator.effectiveBalance 18.788 ms/op 16.356 ms/op 1.15
BeaconState.hashTreeRoot - 1 balances 81.771 us/op 76.249 us/op 1.07
BeaconState.hashTreeRoot - 32 balances 773.93 us/op 792.05 us/op 0.98
BeaconState.hashTreeRoot - 512 balances 6.5207 ms/op 5.8638 ms/op 1.11
BeaconState.hashTreeRoot - 250000 balances 144.19 ms/op 125.54 ms/op 1.15
aggregationBits - 2048 els - zipIndexesInBitList 19.580 us/op 20.488 us/op 0.96
regular array get 100000 times 22.371 us/op 23.426 us/op 0.95
wrappedArray get 100000 times 22.228 us/op 23.548 us/op 0.94
arrayWithProxy get 100000 times 9.3922 ms/op 10.138 ms/op 0.93
ssz.Root.equals 20.890 ns/op 21.812 ns/op 0.96
byteArrayEquals 21.001 ns/op 21.640 ns/op 0.97
Buffer.compare 8.6820 ns/op 8.9540 ns/op 0.97
processSlot - 1 slots 11.392 us/op 8.8930 us/op 1.28
processSlot - 32 slots 2.3405 ms/op 2.1556 ms/op 1.09
getEffectiveBalanceIncrementsZeroInactive - 250000 vs - 7PWei 7.0699 ms/op 3.9788 ms/op 1.78
getCommitteeAssignments - req 1 vs - 250000 vc 1.6041 ms/op 1.9002 ms/op 0.84
getCommitteeAssignments - req 100 vs - 250000 vc 3.2711 ms/op 3.4900 ms/op 0.94
getCommitteeAssignments - req 1000 vs - 250000 vc 3.5118 ms/op 3.7410 ms/op 0.94
findModifiedValidators - 10000 modified validators 798.57 ms/op 826.66 ms/op 0.97
findModifiedValidators - 1000 modified validators 549.27 ms/op 582.18 ms/op 0.94
findModifiedValidators - 100 modified validators 374.57 ms/op 353.25 ms/op 1.06
findModifiedValidators - 10 modified validators 309.67 ms/op 296.41 ms/op 1.04
findModifiedValidators - 1 modified validators 301.17 ms/op 249.50 ms/op 1.21
findModifiedValidators - no difference 225.32 ms/op 194.70 ms/op 1.16
RootCache.getBlockRootAtSlot - 250000 vs - 7PWei 3.3800 ns/op 3.6600 ns/op 0.92
state getBlockRootAtSlot - 250000 vs - 7PWei 385.07 ns/op 276.27 ns/op 1.39
computeProposerIndex 100000 validators 1.3207 ms/op 1.3767 ms/op 0.96
getNextSyncCommitteeIndices 1000 validators 2.8110 ms/op 2.8467 ms/op 0.99
getNextSyncCommitteeIndices 10000 validators 24.704 ms/op 25.406 ms/op 0.97
getNextSyncCommitteeIndices 100000 validators 88.342 ms/op 87.199 ms/op 1.01
computeProposers - vc 250000 552.01 us/op 572.44 us/op 0.96
computeEpochShuffling - vc 250000 38.617 ms/op 38.879 ms/op 0.99
getNextSyncCommittee - vc 250000 9.3237 ms/op 9.3419 ms/op 1.00
nodejs block root to RootHex using toHex 93.050 ns/op 103.79 ns/op 0.90
nodejs block root to RootHex using toRootHex 60.658 ns/op 66.496 ns/op 0.91
nodejs fromHex(blob) 801.70 us/op 934.61 us/op 0.86
nodejs fromHexInto(blob) 615.31 us/op 652.34 us/op 0.94
nodejs block root to RootHex using the deprecated toHexString 637.24 ns/op 625.11 ns/op 1.02
nodejs byteArrayEquals 32 bytes (block root) 25.546 ns/op 25.353 ns/op 1.01
nodejs byteArrayEquals 48 bytes (pubkey) 36.635 ns/op 36.665 ns/op 1.00
nodejs byteArrayEquals 96 bytes (signature) 34.010 ns/op 40.320 ns/op 0.84
nodejs byteArrayEquals 1024 bytes 43.371 ns/op 43.351 ns/op 1.00
nodejs byteArrayEquals 131072 bytes (blob) 1.7401 us/op 1.7419 us/op 1.00
browser block root to RootHex using toHex 141.93 ns/op 146.37 ns/op 0.97
browser block root to RootHex using toRootHex 128.25 ns/op 128.99 ns/op 0.99
browser fromHex(blob) 1.7250 ms/op 1.6804 ms/op 1.03
browser fromHexInto(blob) 632.88 us/op 651.73 us/op 0.97
browser block root to RootHex using the deprecated toHexString 409.60 ns/op 410.79 ns/op 1.00
browser byteArrayEquals 32 bytes (block root) 27.601 ns/op 27.310 ns/op 1.01
browser byteArrayEquals 48 bytes (pubkey) 38.782 ns/op 38.499 ns/op 1.01
browser byteArrayEquals 96 bytes (signature) 73.459 ns/op 72.003 ns/op 1.02
browser byteArrayEquals 1024 bytes 746.03 ns/op 733.10 ns/op 1.02
browser byteArrayEquals 131072 bytes (blob) 93.856 us/op 92.639 us/op 1.01

by benchmarkbot/action

@GrapeBaBa GrapeBaBa changed the title perf(state-transition): use native PTC sampling (WIP)perf(state-transition): use native PTC sampling Aug 24, 2026
@krisoshea-eth krisoshea-eth mentioned this pull request Aug 24, 2026
Open
22 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

Lodestar Team Coordination
Status: In Progress

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

Gloas/Heze mainnet sanity/slots/historical_accumulator spec test is ~7.5x slower than pre-Gloas forks

1 participant

@GrapeBaBa