arnaucube/sonobe-docs
1
0
Fork 0
mirror of https://github.com/arnaucube/sonobe-docs.git synced 2026-02-10 05:06:45 +01:00
Code Issues Packages Projects Releases Wiki Activity

update augmentedfcircuit constr

Browse Source
This commit is contained in:
arnaucube
2024-04-24 10:16:03 +02:00
parent 2709c2950e
commit fc66215bd8
1 changed files with 4 additions and 4 deletions
Download Patch File Download Diff File Expand all files Collapse all files

8
src/design/nova-decider-onchain.md
View File

@@ -78,9 +78,9 @@ The idea is that we check in a R1CS circiut the RelaxedR1CS relation ($Az \circ
<u>Estimated costs of the full decider-onchain circuit:</u> <u>Estimated costs of the full decider-onchain circuit:</u>
*(`x` is the number of constraints of the circuit that we're folding, and the AugmentedFCircuit takes ~80k constraints)* *(`x` is the number of constraints of the circuit that we're folding, and the AugmentedFCircuit takes ~52k constraints)*
1. $U_{n+1}$ relation: `3(x+80k)` 1. $U_{n+1}$ relation: `3(x+52k)`
2. $u_n$ check: `<1000` 2. $u_n$ check: `<1000`
3. $u_n.x$ hash check: `2634` 3. $u_n.x$ hash check: `2634`
4. Pedersen check of $U_{EC,n}$ commitments (native): `<5M` both commitments (including the inputs allocations). This is a couple of native MSMs of <1500 elements each one) 4. Pedersen check of $U_{EC,n}$ commitments (native): `<5M` both commitments (including the inputs allocations). This is a couple of native MSMs of <1500 elements each one)
@@ -89,8 +89,8 @@ The idea is that we check in a R1CS circiut the RelaxedR1CS relation ($Az \circ
7. check that the KZG evaluations are correct 7. check that the KZG evaluations are correct
8. check that the given NIFS challenge $r$ is indeed well computed 8. check that the given NIFS challenge $r$ is indeed well computed
Total: 3 * (x + 80) + 1000 + 2634 + 4_967_155 + 5_146_236 + 7708 Total: 3 * (x + 52_252) + 1000 + 2634 + 4_967_155 + 5_146_236 + 7708
eg: for a circuit of `500k` constraints the decider circuit would take approximately `11.6M` constraints. eg: for a circuit of `500k` constraints the decider circuit would take approximately `11.7M` constraints.
As can be seen, most of the costs come from the Pedersen commitments verification and the $U_{EC,n}$ relation As can be seen, most of the costs come from the Pedersen commitments verification and the $U_{EC,n}$ relation