docs: add compression circuit outline

[LHerskind]

Fixes #4558.

Adds a section to the yellow-paper on message compression circuits for L1 -> L2 message following #4250

[LHerskind]

• docs: add compression circuit outline #4599 👈
• master

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

Join @LHerskind and the rest of your teammates on Graphite

[AztecBot]

Benchmark results

Metrics with a significant change:

• note_trial_decrypting_time_in_ms (8): 107 (+1322%)

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.

Values are compared against data from master at commit cd1ca2e1 and shown if the difference exceeds 1%.

L2 block published to L1

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

Metric	8 txs	32 txs	64 txs
l1_rollup_calldata_size_in_bytes	5,700	18,884	36,452
l1_rollup_calldata_gas	65,976	239,024	469,856
l1_rollup_execution_gas	193,924	500,330	909,114
l2_block_processing_time_in_ms	1,196	4,503 (-1%)	8,977
note_successful_decrypting_time_in_ms	196 (-2%)	546 (+1%)	1,006 (+2%)
note_trial_decrypting_time_in_ms	⚠️ 107 (+1322%)	67.8 (+14%)	37.2 (-70%)
l2_block_building_time_in_ms	15,808 (-1%)	62,827 (-1%)	126,094 (-2%)
l2_block_rollup_simulation_time_in_ms	12,016 (-1%)	47,766 (-1%)	96,085 (-2%)
l2_block_public_tx_process_time_in_ms	3,762	14,984 (-3%)	29,841 (-1%)

L2 chain processing

Each column represents the number of blocks on the L2 chain where each block has 16 txs.

Metric	5 blocks	10 blocks
node_history_sync_time_in_ms	14,011 (-1%)	26,911
note_history_successful_decrypting_time_in_ms	1,345 (+6%)	2,385 (-2%)
note_history_trial_decrypting_time_in_ms	91.3 (-9%)	149 (-12%)
node_database_size_in_bytes	18,792,528	35,590,224 (+1%)
pxe_database_size_in_bytes	29,923	59,478

Circuits stats

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

Circuit	circuit_simulation_time_in_ms	circuit_input_size_in_bytes	circuit_output_size_in_bytes
private-kernel-init	248	44,736	28,001
private-kernel-ordering	178	52,625	14,627
base-rollup	1,278 (-1%)	177,932	933
root-rollup	71.1 (+1%)	4,192	825
private-kernel-inner	315 (+1%)	73,715	28,001
public-kernel-app-logic	191 (-1%)	32,254	25,379
merge-rollup	5.78	2,712	933

Tree insertion stats

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

Metric	1 leaves	2 leaves	8 leaves	16 leaves	32 leaves	64 leaves	128 leaves	512 leaves	1024 leaves	2048 leaves	4096 leaves
batch_insert_into_append_only_tree_16_depth_ms	9.85	10.2 (-3%)	12.5 (-8%)	16.2	22.3 (+2%)	35.4	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_16_depth_hash_count	16.9	17.5	23.0	31.6	47.0	79.0	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_16_depth_hash_ms	0.571	0.568 (-2%)	0.531 (-8%)	0.502 (+1%)	0.466 (+1%)	0.441	N/A	N/A	N/A	N/A	N/A
batch_insert_into_append_only_tree_32_depth_ms	N/A	N/A	N/A	N/A	N/A	45.8	72.3	229	444	869	1,724
batch_insert_into_append_only_tree_32_depth_hash_count	N/A	N/A	N/A	N/A	N/A	96.0	159	543	1,055	2,079	4,127
batch_insert_into_append_only_tree_32_depth_hash_ms	N/A	N/A	N/A	N/A	N/A	0.469	0.446	0.419	0.416	0.413	0.413
batch_insert_into_indexed_tree_20_depth_ms	N/A	N/A	N/A	N/A	N/A	53.9	108	337	662	1,311	2,626 (+1%)
batch_insert_into_indexed_tree_20_depth_hash_count	N/A	N/A	N/A	N/A	N/A	104	207	691	1,363	2,707	5,395
batch_insert_into_indexed_tree_20_depth_hash_ms	N/A	N/A	N/A	N/A	N/A	0.479	0.487	0.456	0.455	0.455	0.455 (+1%)
batch_insert_into_indexed_tree_40_depth_ms	N/A	N/A	N/A	N/A	61.9 (+1%)	N/A	N/A	N/A	N/A	N/A	N/A
batch_insert_into_indexed_tree_40_depth_hash_count	N/A	N/A	N/A	N/A	109	N/A	N/A	N/A	N/A	N/A	N/A
batch_insert_into_indexed_tree_40_depth_hash_ms	N/A	N/A	N/A	N/A	0.540 (+1%)	N/A	N/A	N/A	N/A	N/A	N/A

Miscellaneous

Transaction sizes based on how many contracts are deployed in the tx.

Metric	0 deployed contracts
tx_size_in_bytes	19,179

Transaction processing duration by data writes.

Metric	0 new note hashes	1 new note hashes
tx_pxe_processing_time_ms	2,614	1,354

Metric	0 public data writes	1 public data writes
tx_sequencer_processing_time_ms	0.0306 (-5%)	467 (-1%)

[benesjan]

Mostly minor issues

[sklppy88]

I think I'm missing the big picture here as to what is the exact reason why this mechanism works for making it easier for users to build proofs. I understand from this that SHA256 is not stark friendly but what makes this circuit stark friendlier ? Please ignore this if this specific background knowledge is assumed already from the outset, and / or the why is not relevant. Aside from this, I think Jan's comments are all fair + should be addressed.

[benesjan]

Why the "+1" here?

[LHerskind]

It is the height of the outgoing tree, which is the heigh of the transactions tree + 1, because each tx have 2 messages. Should probably be something more like math.ceil(log2(MAX_NEW_L2_TO_L1_MSGS_PER_TX))

[benesjan]

Ok ok, could you add a comment there? I think this can confuse more people.

[benesjan]

Looks good. Just some nits. Feel free to merge once addressed

[benesjan]

Suggested change

	In reality larger trees would have more layers between the leafs and the root.
	In reality larger trees would have more layers between the leaves and the root.

[benesjan]

This sentence is redundant to the one on line 33.

[benesjan]

Suggested change

	The circuit must simply compute the two trees using the same inputs, and then we ensure that the elements of the trees match the inbox later at the [state transitioner](./../l1-smart-contracts/index.md#overview).
	The circuit computes the two trees using the same inputs, and then we ensure that the elements of the trees match the inbox later in the [state transitioner](./../l1-smart-contracts/index.md#overview).