chore(avm): use commit_sparse

[fcarreiro]

Ran on a ~2^19 trace with 7% non-zero elems (the long keccak).

BEFORE

prove/execute_log_derivative_inverse_commitments_round_ms: 1598
prove/execute_wire_commitments_round_ms: 4992

AFTER

prove/execute_log_derivative_inverse_commitments_round_ms: 1826
prove/execute_wire_commitments_round_ms: 2092

[fcarreiro]

• chore(avm): use commit_sparse #7581 👈
• master

This stack of pull requests is managed by Graphite. Learn more about stacking.

Join @fcarreiro and the rest of your teammates on Graphite

[AztecBot]

⚠️ Performance Alert ⚠️

Possible performance regression was detected for benchmark 'C++ Benchmark'.
Benchmark result of this commit is worse than the previous benchmark result exceeding threshold 1.05.

Benchmark suite	Current: c52c8ba	Previous: 0e3ce84	Ratio
nativeClientIVCBench/Full/6	16799.669832999996 ms/iter	14245.174773000002 ms/iter	1.18
nativeconstruct_proof_ultrahonk_power_of_2/20	5199.691017999996 ms/iter	4744.4964180000115 ms/iter	1.10
wasmClientIVCBench/Full/6	44624.76105 ms/iter	40542.955264 ms/iter	1.10
wasmconstruct_proof_ultrahonk_power_of_2/20	16867.301229 ms/iter	14940.837621 ms/iter	1.13

This comment was automatically generated by workflow using github-action-benchmark.

CC: @ludamad @codygunton

[Maddiaa0]

It depends some lookups are, as they are not active on every row

[Maddiaa0]

Grand product polys for the copy constraints are dense though so we can treat those differently

[fcarreiro]

Ah good point. I'll merge this one but we can consider later.

I did try with commit_sparse here and did see an improvement of some % points, but not as big as the other case. I preferred to err on the side of caution mem-wise, since the current implementation copies the poly.

[ledwards2225]

A bit tougher for you guys since you have so many polys that aren’t distinguished from one another. For example, we use the sparse commit for only a subset of the log derivative inverse polys, and notably not for the table lookup inverses because they’re generally used heavily.

[AztecBot]

Benchmark results

Metrics with a significant change:

• avm_simulation_time_ms (Token:assert_minter_and_mint): 93.1 (+44%)
• avm_simulation_time_ms (Token:_increase_public_balance): 158 (+90%)
• l2_block_building_time_in_ms (4): 15,278 (+48%)
• l2_block_building_time_in_ms (16): 51,474 (+30%)
• l2_block_rollup_simulation_time_in_ms (4): 15,278 (+48%)
• l2_block_rollup_simulation_time_in_ms (16): 51,473 (+30%)
• l2_block_public_tx_process_time_in_ms (4): 13,539 (+57%)
• l2_block_public_tx_process_time_in_ms (16): 49,425 (+31%)

Detailed results

All benchmarks are run on txs on the Benchmarking contract on the repository. Each tx consists of a batch call to create_note and increment_balance, which guarantees that each tx has a private call, a nested private call, a public call, and a nested public call, as well as an emitted private note, an unencrypted log, and public storage read and write.

This benchmark source data is available in JSON format on S3 here.

Proof generation

Each column represents the number of threads used in proof generation.

Metric	1 threads	4 threads	16 threads	32 threads	64 threads
proof_construction_time_sha256_ms	5,747	1,551	708	760 (+1%)	763 (-2%)
proof_construction_time_sha256_30_ms	11,842	3,172	1,410	1,436 (-1%)	1,461
proof_construction_time_sha256_100_ms	44,020	12,000	5,439 (+1%)	5,447 (+1%)	5,385
proof_construction_time_poseidon_hash_ms	78.0	34.0	34.0	58.0 (-2%)	88.0 (-1%)
proof_construction_time_poseidon_hash_30_ms	1,529	418	200	228 (+5%)	267
proof_construction_time_poseidon_hash_100_ms	5,662	1,524	684 (+1%)	731 (+1%)	749

L2 block published to L1

Each column represents the number of txs on an L2 block published to L1.

Metric	4 txs	8 txs	16 txs
l1_rollup_calldata_size_in_bytes	708	708	708
l1_rollup_calldata_gas	6,576	6,590	6,600
l1_rollup_execution_gas	583,038	583,203	583,364
l2_block_processing_time_in_ms	764 (+1%)	1,436	2,720
l2_block_building_time_in_ms	⚠️ 15,278 (+48%)	20,946 (+4%)	⚠️ 51,474 (+30%)
l2_block_rollup_simulation_time_in_ms	⚠️ 15,278 (+48%)	20,945 (+4%)	⚠️ 51,473 (+30%)
l2_block_public_tx_process_time_in_ms	⚠️ 13,539 (+57%)	19,050 (+5%)	⚠️ 49,425 (+31%)

L2 chain processing

Each column represents the number of blocks on the L2 chain where each block has 8 txs.

Metric	3 blocks	5 blocks
node_history_sync_time_in_ms	7,067	10,156 (-1%)
node_database_size_in_bytes	12,705,872	16,871,504
pxe_database_size_in_bytes	16,254	26,813

Circuits stats

Stats on running time and I/O sizes collected for every kernel circuit run across all benchmarks.

Circuit	simulation_time_in_ms	witness_generation_time_in_ms	input_size_in_bytes	output_size_in_bytes	proving_time_in_ms	proof_size_in_bytes	num_public_inputs	size_in_gates
private-kernel-init	97.4 (+4%)	398 (-2%)	21,744	54,085	N/A	N/A	N/A	N/A
private-kernel-inner	179 (+5%)	693	81,209	54,264	N/A	N/A	N/A	N/A
private-kernel-tail	384 (+4%)	864	61,565	62,127	N/A	N/A	N/A	N/A
base-parity	6.89 (+5%)	654	160	96.0	1,358 (+1%)	12,676	3.00	131,072
root-parity	120 (+12%)	107 (+1%)	64,860	96.0	27,922	12,676	3.00	4,194,304
base-rollup	3,589 (+10%)	5,023 (-2%)	181,083	632	40,860	13,476	28.0	4,194,304
root-rollup	155 (+1%)	109 (+3%)	51,261	652	25,389 (+1%)	13,412	26.0	4,194,304
public-kernel-setup	103 (+1%)	2,328	113,048	80,822	15,753 (+1%)	119,428	3,339	2,097,152
public-kernel-app-logic	117 (+10%)	3,322 (+1%)	113,048	80,822	8,870 (-1%)	119,428	3,339	1,048,576
public-kernel-tail	694 (+9%)	25,959 (+4%)	410,181	10,814	61,912	25,860	415	8,388,608
private-kernel-reset-tiny	237 (+5%)	933 (-1%)	77,005	53,954	N/A	N/A	N/A	N/A
private-kernel-tail-to-public	4,891	1,385	773,216	1,632	N/A	N/A	N/A	N/A
public-kernel-teardown	91.4 (+1%)	3,318 (+2%)	113,048	80,822	16,421 (+1%)	119,428	3,339	2,097,152
merge-rollup	60.4 (+5%)	N/A	33,566	632	N/A	N/A	N/A	N/A
undefined	N/A	N/A	N/A	N/A	151,732 (-13%)	N/A	N/A	N/A

Stats on running time collected for app circuits

Function	input_size_in_bytes	output_size_in_bytes	witness_generation_time_in_ms	proof_size_in_bytes	proving_time_in_ms
ContractClassRegisterer:register	1,312	9,344	377 (+5%)	N/A	N/A
ContractInstanceDeployer:deploy	1,376	9,344	27.0 (+7%)	N/A	N/A
MultiCallEntrypoint:entrypoint	1,888	9,344	572 (+1%)	N/A	N/A
GasToken:deploy	1,344	9,344	481 (-2%)	N/A	N/A
SchnorrAccount:constructor	1,280	9,344	438 (+1%)	N/A	N/A
SchnorrAccount:entrypoint	2,272	9,344	777 (+6%)	N/A	N/A
Token:privately_mint_private_note	1,248	9,344	536 (+1%)	N/A	N/A
FPC:fee_entrypoint_public	1,312	9,344	102 (-3%)	N/A	N/A
Token:transfer	1,280	9,344	1,348	N/A	N/A
AuthRegistry:set_authorized (avm)	19,222	N/A	N/A	95,392	2,335 (-4%)
FPC:prepare_fee (avm)	22,670	N/A	N/A	95,456	2,451 (-1%)
Token:transfer_public (avm)	42,914	N/A	N/A	95,456	3,035 (-4%)
AuthRegistry:consume (avm)	33,100	N/A	N/A	95,424	2,719 (-5%)
FPC:pay_refund (avm)	32,835	N/A	N/A	95,424	7,788 (-5%)
Benchmarking:create_note	1,312	9,344	446 (+8%)	N/A	N/A
SchnorrAccount:verify_private_authwit	1,248	9,344	41.5	N/A	N/A
Token:unshield	1,344	9,344	1,156 (-1%)	N/A	N/A
FPC:fee_entrypoint_private	1,344	9,344	1,516 (-1%)	N/A	N/A

AVM Simulation

Time to simulate various public functions in the AVM.

Function	time_ms	bytecode_size_in_bytes
GasToken:_increase_public_balance	98.4	13,790
GasToken:set_portal	13.3 (+10%)	3,339
Token:constructor	124 (+2%)	23,692
FPC:constructor	92.8 (+1%)	13,592
GasToken:mint_public	81.7	10,158
Token:mint_public	58.3 (-4%)	19,034
Token:assert_minter_and_mint	⚠️ 93.1 (+44%)	12,925
AuthRegistry:set_authorized	13.3 (-4%)	7,812
FPC:prepare_fee	138 (+5%)	11,068
Token:transfer_public	35.2 (-15%)	31,218
FPC:pay_refund	122 (-3%)	21,266
Benchmarking:increment_balance	1,119 (+12%)	11,281
Token:_increase_public_balance	⚠️ 158 (+90%)	15,006
FPC:pay_refund_with_shielded_rebate	108 (-1%)	22,361

Public DB Access

Time to access various public DBs.

Function	time_ms
get-nullifier-index	0.159 (+1%)

Tree insertion stats

The duration to insert a fixed batch of leaves into each tree type.

Metric	1 leaves	16 leaves	64 leaves	128 leaves	256 leaves	512 leaves	1024 leaves
batch_insert_into_append_only_tree_16_depth_ms	10.8 (+4%)	17.5 (+4%)	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_16_depth_hash_count	16.8	31.7	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_16_depth_hash_ms	0.627 (+4%)	0.538 (+4%)	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_32_depth_ms	N/A	N/A	53.0 (+9%)	76.2 (+1%)	131 (+1%)	248 (-3%)	471 (+1%)
batch_insert_into_append_only_tree_32_depth_hash_count	N/A	N/A	95.9	159	287	543	1,055
batch_insert_into_append_only_tree_32_depth_hash_ms	N/A	N/A	0.542 (+9%)	0.468 (+1%)	0.451 (+1%)	0.449 (-3%)	0.440 (+1%)
batch_insert_into_indexed_tree_20_depth_ms	N/A	N/A	65.3 (+8%)	111 (-1%)	184 (+1%)	356 (-1%)	696 (+1%)
batch_insert_into_indexed_tree_20_depth_hash_count	N/A	N/A	109	207	355	691	1,363
batch_insert_into_indexed_tree_20_depth_hash_ms	N/A	N/A	0.553 (+9%)	0.499	0.488 (+1%)	0.482	0.477 (+1%)
batch_insert_into_indexed_tree_40_depth_ms	N/A	N/A	76.9 (+5%)	N/A	N/A	N/A	N/A
batch_insert_into_indexed_tree_40_depth_hash_count	N/A	N/A	133	N/A	N/A	N/A	N/A
batch_insert_into_indexed_tree_40_depth_hash_ms	N/A	N/A	0.549 (+5%)	N/A	N/A	N/A	N/A

Miscellaneous

Transaction sizes based on how many contract classes are registered in the tx.

Metric	0 registered classes	1 registered classes
tx_size_in_bytes	77,362	668,542

Transaction size based on fee payment method

| Metric | |
| - | |