Skip to content

feat: configure new param set [skip-line-limit] #1576

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
ctrlc03 merged 12 commits into from
Jun 4, 2026
Merged

feat: configure new param set [skip-line-limit] #1576

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
12 commits
Select commit Hold shift + click to select a range
106cb02
adjust event filter for C7 integration test
0xjei Jun 3, 2026
80a221f
executed benches and report
0xjei Jun 3, 2026
e4e195e
executed medium bench with 3 mod
0xjei Jun 3, 2026
7fa05c1
update report
0xjei Jun 3, 2026
0cf46c3
remove esi_per_ct and hardcoded lambda
0xjei Jun 4, 2026
84827a7
add large committee size
0xjei Jun 4, 2026
334123a
update to param set 3+ and run micro again
0xjei Jun 4, 2026
c2b2831
Merge branch 'main' into bench/medium-3mod
0xjei Jun 4, 2026
024a17f
fix bug on report regeneration due to ordering of script calls
0xjei Jun 4, 2026
298321b
regenerate micro report
0xjei Jun 4, 2026
99a7583
switch to insecure for local CI test
0xjei Jun 4, 2026
7302ea6
introduce insecure search params
0xjei Jun 4, 2026
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
14 changes: 14 additions & 0 deletions circuits/benchmarks/results_secure_agg_medium/benchmark_run_meta.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,14 @@
{
"benchmark_mode": "secure",
"bfv_preset_subdir": "secure-8192",
"committee": "medium",
"proof_aggregation": true,
"multithread_jobs": 13,
"verbose": true,
"nodes_spawned": 20,
"committee_size_n": 10,
"committee_size_h": 8,
"committee_threshold_t": 4,
"network_model": "in_process_bus",
"testmode_harness": true
}
107 changes: 107 additions & 0 deletions circuits/benchmarks/results_secure_agg_medium/crisp_verify_gas.json
View file
Open in desktop

Large diffs are not rendered by default.

244 changes: 244 additions & 0 deletions circuits/benchmarks/results_secure_agg_medium/integration_summary.json
View file
Open in desktop

Large diffs are not rendered by default.

215 changes: 215 additions & 0 deletions circuits/benchmarks/results_secure_agg_medium/report.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,215 @@
# Enclave ZK Circuit Benchmarks

**Generated:** 2026-06-03 21:00:47 UTC

**Git Branch:** `bench/medium-3mod`
**Git Commit:** `80a221f2152a8cf4cc4d65d0905c555d18da1f02`

**Committee Size:** `H=8`, `N=10`, `T=4`

## Run configuration

Settings for this benchmark run (integration test + Nargo circuit benches on the same host).

### Integration test (`test_trbfv_actor`)

| Setting | Value |
| ----------------------------------------------------- | -------------------------------------------- |
| Benchmark mode | `secure` |
| BFV preset (artifacts) | `secure-8192` |
| BFV preset (enum) | `SecureThreshold8192` |
| λ (smudging / error) | 55 |
| Nodes spawned (builder) | 20 |
| Network model | `in_process_bus` |
| Testmode harness | true |
| `proof_aggregation_enabled` | true |
| `BENCHMARK_MULTITHREAD_JOBS` (max concurrent ZK jobs) | 13 |
| Rayon worker threads | 13 |
| CPU cores (host) | 14 |
| `dkg_fold_attestation_verifier` (EIP-712) | `0x7969c5eD335650692Bc04293B07F5BF2e7A673C0` |
| Verbose logging (`run_benchmarks.sh --verbose`) | true |

### Hardware & software (Nargo / Barretenberg host)

| | |
| ---------------- | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| **CPU** | Apple M4 Pro |
| **CPU cores** | 14 |
| **RAM** | 48.00 GB |
| **OS** | Darwin |
| **Architecture** | arm64 |
| **Nargo** | nargo version = 1.0.0-beta.16 noirc version = 1.0.0-beta.16+2d46fca7203545cbbfb31a0d0328de6c10a8db95 (git version hash: 2d46fca7203545cbbfb31a0d0328de6c10a8db95, is dirty: false) |
| **Barretenberg** | 3.0.0-nightly.20260102 |

---

## Audit status

On-chain verify gas: **complete** (CRISP Π_user + Enclave Π_DKG / Π_dec replay).

---

## Measurement methodology

| Metric kind | Source | Meaning | Do **not** use for |
| -------------------- | ------------------------------------------------ | ------------------------------------------------------------------------------------------ | -------------------------------------------------------------- |
| **wall_clock** | `test_trbfv_actor` phase timers / HLC event span | End-to-end wait in the in-process test harness | Production WAN latency; per-node deployment cost |
| **isolated_nargo** | `benchmark_circuit.sh` per circuit | Single `bb prove` on oracle witness, one circuit at a time | Full protocol pipeline (different witness path) |
| **tracked_job_wall** | `MultithreadReport` per `ComputeRequest` | Wall time of each job on the shared Rayon pool (≤ `BENCHMARK_MULTITHREAD_JOBS` concurrent) | End-to-end time — **sums exceed wall clock** when jobs overlap |

**Harness limits (integration):** all ciphernodes share one process and bus
(`network_model: in_process_bus`); sortition registers extra nodes; `testmode_*` enabled. Compare
runs only with the same `benchmark_mode`, proof-aggregation flag, `BENCHMARK_MULTITHREAD_JOBS`,
commit, and hardware.

---

## Protocol Summary

### Circuit Benchmarks (isolated Nargo + Barretenberg)

Single-circuit `bb prove` on the benchmark oracle witness (not the integration actor pipeline).

| Circuit | Constraints | Prove (s) | Verify (ms) | Proof (KB) |
| -------------------- | ----------- | --------- | ----------- | ---------- |
| C0 | 287764 | 1.42 | 24.34 | 15.88 |
| C1 | 2432076 | 9.28 | 25.01 | 15.88 |
| C2a | 4813061 | 17.97 | 24.23 | 15.88 |
| C2b | 6255714 | 20.59 | 24.25 | 15.88 |
| C3a | 3563517 | 10.72 | 24.68 | 15.88 |
| C3b | 3563517 | 10.72 | 24.68 | 15.88 |
| C4a | 3115129 | 10.19 | 25.03 | 15.88 |
| C4b | 3115129 | 10.19 | 25.03 | 15.88 |
| C5 | 6073618 | 20.41 | 25.71 | 15.88 |
| user_data_encryption | 1688723 | 5.74 | 25.35 | 15.88 |
| C6 | 3001845 | 10.24 | 25.30 | 15.88 |
| C7 | 191104 | 0.86 | 26.53 | 15.88 |

### Artifacts

| Artifact | Proof size | Public input size | Verify gas | Calldata gas | Total gas |
| -------- | ---------- | ----------------- | ---------- | ------------ | --------- |
| Π_DKG | 10.69 KB | 0.94 KB | 3154367 | 181920 | 3336287 |
| Π_user | 15.88 KB | 0.12 KB | 2972929 | 193348 | 3166277 |
| Π_dec | 10.69 KB | 3.75 KB | 3658402 | 190692 | 3849094 |

### Role / Phase / Activity

| Role | Phase | Activity | Metric | Duration | Proof size | Bandwidth |
| --------------- | ----- | ----------------------------------------- | -------------- | --------- | ---------- | --------- |
| Each ciphernode | P1 | one-time DKG participation (test harness) | wall_clock | 6558.88 s | 127.00 KB | 130.81 KB |
| Aggregator | P2 | C5 + Π_DKG fold (aggregator span) | wall_clock | 1380.96 s | 10.69 KB | 11.62 KB |
| User | P3 | per user input | isolated_nargo | 10.95 s | 15.88 KB | 16.00 KB |
| Each ciphernode | P4 | per computation output (C6) | isolated_nargo | 10.24 s | 15.88 KB | 16.00 KB |
| Aggregator | P4 | C7 + Π_dec fold (full publish→aggregate) | wall_clock | 382.94 s | 10.69 KB | 14.44 KB |
| Aggregator | P4 | C7 + fold only (pending→plaintext span) | wall_clock | 92.61 s | 10.69 KB | 14.44 KB |

_P2 **tracked_job_wall** sum (ZkDkgAggregation + ZkPkAggregation, parallelizable): **155.61 s** —
not comparable to P2 wall_clock row above._

## Integration test (`test_trbfv_actor`)

### End-to-end phase timings (integration test)

| Phase | Metric | Duration (s) |
| ------------------------------------------------------------------ | ------------ | ------------ |
| Starting trbfv actor test | `wall_clock` | 0.00 |
| Setup completed | `wall_clock` | 2.78 |
| Committee Setup Completed | `wall_clock` | 20.09 |
| Committee Finalization Complete | `wall_clock` | 0.00 |
| Aggregator P2: PkAggregation pending -> PublicKeyAggregated (wall) | `wall_clock` | 1380.96 |
| ThresholdShares -> PublicKeyAggregated | `wall_clock` | 6558.88 |
| E3Request -> PublicKeyAggregated | `wall_clock` | 6559.39 |
| Application CT Gen | `wall_clock` | 0.28 |
| Running FHE Application | `wall_clock` | 0.00 |
| Aggregator P4: Aggregation pending -> PlaintextAggregated (wall) | `wall_clock` | 92.61 |
| Ciphertext published -> PlaintextAggregated | `wall_clock` | 382.94 |
| Entire Test | `wall_clock` | 6965.49 |

### Multithread job timings (`tracked_job_wall`)

| Name | Avg (s) | Runs | Total (s) |
| ----------------------------- | ------- | ---- | --------- |
| CalculateDecryptionKey | 0.05 | 10 | 0.46 |
| CalculateDecryptionShare | 0.16 | 10 | 1.58 |
| CalculateThresholdDecryption | 0.25 | 1 | 0.25 |
| GenEsiSss | 0.36 | 10 | 3.65 |
| GenPkShareAndSkSss | 0.26 | 10 | 2.60 |
| NodeDkgFold/c2ab_fold | 25.92 | 10 | 259.25 |
| NodeDkgFold/c3a_fold | 674.19 | 10 | 6741.94 |
| NodeDkgFold/c3ab_fold | 20.81 | 10 | 208.08 |
| NodeDkgFold/c3b_fold | 639.78 | 10 | 6397.83 |
| NodeDkgFold/c4ab_fold | 20.69 | 10 | 206.88 |
| NodeDkgFold/node_fold | 39.74 | 10 | 397.38 |
| ZkDecryptedSharesAggregation | 4.73 | 1 | 4.73 |
| ZkDecryptionAggregation | 87.50 | 1 | 87.50 |
| ZkDkgAggregation | 41.30 | 1 | 41.30 |
| ZkDkgShareDecryption | 90.86 | 20 | 1817.28 |
| ZkNodeDkgFold | 1421.14 | 10 | 14211.40 |
| ZkPkAggregation | 114.31 | 1 | 114.31 |
| ZkPkBfv | 10.21 | 10 | 102.12 |
| ZkPkGeneration | 121.56 | 10 | 1215.57 |
| ZkShareComputation | 96.21 | 20 | 1924.15 |
| ZkShareEncryption | 106.26 | 540 | 57380.53 |
| ZkThresholdShareDecryption | 192.97 | 10 | 1929.69 |
| ZkVerifyShareDecryptionProofs | 0.51 | 10 | 5.13 |
| ZkVerifyShareProofs | 2.06 | 12 | 24.67 |

Sum of tracked job wall time: **93078.28 s** — **not** end-to-end latency (jobs run in parallel up
to `BENCHMARK_MULTITHREAD_JOBS`).

### NodeDkgFold sub-steps (`tracked_job_wall`, per fold prove)

| Step | Avg (s) | Runs | Total (s) |
| --------- | ------- | ---- | --------- |
| c2ab_fold | 25.92 | 10 | 259.25 |
| c3a_fold | 674.19 | 10 | 6741.94 |
| c3ab_fold | 20.81 | 10 | 208.08 |
| c3b_fold | 639.78 | 10 | 6397.83 |
| c4ab_fold | 20.69 | 10 | 206.88 |
| node_fold | 39.74 | 10 | 397.38 |

### Aggregation jobs (`tracked_job_wall`)

| Operation | Avg (s) | Runs | Total (s) |
| ---------------------------- | ------- | ---- | --------- |
| ZkDecryptedSharesAggregation | 4.73 | 1 | 4.73 |
| ZkDecryptionAggregation | 87.50 | 1 | 87.50 |
| ZkDkgAggregation | 41.30 | 1 | 41.30 |
| ZkNodeDkgFold | 1421.14 | 10 | 14211.40 |
| ZkPkAggregation | 114.31 | 1 | 114.31 |

Sum of aggregation job tracked time: **14459.24 s** (parallel CPU work; not P1/P2 wall clock).

### Folded on-chain artifacts (exported for Π_DKG / Π_dec gas)

| Artifact | Proof (bytes) | Public inputs (bytes) |
| --------------------- | ------------- | --------------------- |
| dkg_aggregator | 10944 | 960 |
| decryption_aggregator | 10944 | 3840 |

## Raw circuit benchmark JSON (Nargo)

Source files for the **Circuit Benchmarks** table. Persist this directory with
`crisp_verify_gas.json` (and optional `integration_summary.json`) to regenerate the report without
re-running the integration test.

| File |
| ----------------------------------------------------- |
| `config_default.json` |
| `dkg_e_sm_share_computation_default.json` |
| `dkg_pk_default.json` |
| `dkg_share_decryption_default.json` |
| `dkg_share_encryption_default.json` |
| `dkg_sk_share_computation_default.json` |
| `threshold_decrypted_shares_aggregation_default.json` |
| `threshold_pk_aggregation_default.json` |
| `threshold_pk_generation_default.json` |
| `threshold_share_decryption_default.json` |
| `threshold_user_data_encryption_ct0_default.json` |
| `threshold_user_data_encryption_ct1_default.json` |

## Notes

- All nodes are executed on the same machine in this benchmark run, so inter-node network latency is
effectively 0.
3 changes: 3 additions & 0 deletions circuits/benchmarks/results_secure_agg_micro/benchmark_run_meta.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,11 +1,14 @@
{
"benchmark_mode": "secure",
"bfv_preset_subdir": "secure-8192",
"committee": "micro",
"proof_aggregation": true,
"multithread_jobs": 13,
"verbose": true,
"nodes_spawned": 20,
"committee_size_n": 3,
"committee_size_h": 3,
"committee_threshold_t": 1,
"network_model": "in_process_bus",
"testmode_harness": true
}
Loading
Loading