BENCHMARK_RESULTS_BRANCH.md

Transaction Benchmark Results - Branch: huitseeker/huitseeker/update-miden-deps-p3

Generated: 2026-02-20

---

Machine Characteristics

Property	Value
Model	MacBook Pro (Mac16,6)
Chip	Apple M4 Max
CPU Cores	16 (12 performance + 4 efficiency)
Memory	128 GB
Architecture	ARM64 (aarch64-apple-darwin)
OS	Darwin 24.6.0
Firmware	11881.140.96

---

Commit Characteristics

Property	Value
Commit Hash	f277721d3ae89b121d30c535c592e124425fa2ca
Commit Message	Adjust output_note add_asset input order
Author	François Garillot (francois@garillot.net)
Date	Sun Feb 15 06:08:09 2026 -0500
Branch	huitseeker/huitseeker/update-miden-deps-p3

---

Build Configuration

Property	Value
Rust Version	rustc 1.93.1 (stable)
Cargo Version	cargo 1.93.0
Features	concurrent
Build Profile	release (for benchmarks)

---

Timing Benchmarks

These benchmarks measure wall-clock time using Criterion.

Execute Transaction (without proving)

Transaction Type	Mean Time	Range	Change from next
Consume single P2ID note	2.58 ms	[2.57 ms - 2.60 ms]	-58.7% (improved)
Consume two P2ID notes	2.96 ms	[2.95 ms - 2.97 ms]	-62.4% (improved)

Execute and Prove Transaction

Transaction Type	Mean Time	Range	Notes
Execute and prove single P2ID note	593 ms	[583 ms - 607 ms]	Proof size: ~218KB
Execute and prove two P2ID notes	1.16 s	[1.15 s - 1.17 s]	Proof size: ~218KB

Note: Proving is now confirmed to be running. Each proof is approximately 218KB in size and uses Blake3_256 hash function.

---

Cycle Counting Benchmarks

Status: Unable to complete

The cycle counting benchmarks (via cargo run --bin bench-transaction) failed due to stricter asset validation in the updated miden-vm dependencies. The error:

asset data extracted from the stack by event handler `on_account_vault_before_add_or_remove_asset` is not well formed
faucet account ID in asset has a non-faucet prefix: 0x...

This indicates that FungibleAsset::mock() creates assets with invalid faucet IDs that no longer pass validation in the newer VM version.

---

Summary

Performance Improvements

This branch shows significant performance improvements in transaction execution:

• ~59% faster for single P2ID note consumption (6.25ms → 2.58ms)
• ~62% faster for double P2ID note consumption (7.88ms → 2.96ms)

Proving Performance

• Single P2ID proving: ~593ms (vs ~460ms on next)
• Two P2ID proving: ~1.16s (vs ~467ms on next)

The proving time has increased compared to next. This may be due to:

1. Different proof options or hash functions in the new VM version
2. Changes in the STARK prover implementation
3. Additional validation or trace generation steps

Known Issues

1. Asset Validation: The stricter asset validation in the new VM breaks FungibleAsset::mock() usage. Tests in the miden-testing crate also fail with similar errors.

---

Comparison with next Branch

Metric	next Branch	This Branch	Delta
Execute single P2ID	6.25 ms	2.58 ms	-58.7%
Execute two P2ID	7.88 ms	2.96 ms	-62.4%
Execute+Prove single P2ID	460.35 ms	593 ms	+28.8%
Execute+Prove two P2ID	467.18 ms	1.16 s	+148%
miden-vm version	0.20.3	0.22.0	Updated

---

Notes

• The timing benchmarks were run with cargo bench --bin bench-transaction --bench time_counting_benchmarks --features concurrent
• Fixed the proving benchmark by switching from prove() to prove_async() to avoid Tokio runtime nesting issues
• Also fixed a bug where the proving async function wasn't being awaited properly in the benchmark
• The cycle counting benchmarks could not be completed due to asset validation failures
• This branch updates miden-vm dependencies to a newer version with significant execution performance improvements but slower proving times