is_prime on sp1
#814
Draft
is_prime on sp1
#814
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| @@ -0,0 +1,127 @@ | |||
| //! An end-to-end example of using the SP1 SDK to generate a proof of a program that can have an | |||
| //! EVM-Compatible proof generated which can be verified on-chain. | |||
There was a problem hiding this comment.
Why are we going with evm compatible proofs?
Could we just do the simplest thing that SP1 allows? Or is there a good reason to go with something more complicated?
(I don't think our proof is evm-compatible? So it would be fair to require that of SP1.)
There was a problem hiding this comment.
I followed the instructions here. I generated the project via cargo prove new --bare is_prime.
- I don't know why there's anything pertaining to
evmthere, it surprised me too. - It generated a Fibonacci project in a folder called
is_prime:). I modified the logic, but didn't bother with all the modules, paths, etc. I changed them now.
|
My results for sp1: and Ceno You can see the number of cycles differs ( |
|
If I enable link time optimisation for SP1, I get 325_334 cycles. That takes care of most of the difference in cycles, since it's only about 2.7% more than Ceno. (We can investigate that a bit more.) diff --git a/is_prime/Cargo.toml b/is_prime/Cargo.toml
index 3915b209..9d246545 100644
--- a/is_prime/Cargo.toml
+++ b/is_prime/Cargo.toml
@@ -8,3 +8,14 @@ resolver = "2"
[workspace.dependencies]
alloy-sol-types = "0.7.7"
+
+[profile.release]
+lto = "fat"
+panic = "abort"
+codegen-units = 1
+
+[profile.host]
+inherits = "release"
+lto = "thin"
+panic = "abort"
+codegen-units = 16cargo run --profile=host -- --n=5000 --executeLet me actually run the proving next. |
|
When proving on my machine with LTO turned on for SP1 (see comment above), I get the following: SP1CenoFurther resultsWith n = 20_000 I get this for Ceno And this for SP1: |
|
@mcalancea Could you please do the sorting benchmark for SP1 and ceno? Not the quadratic one, just the one that uses Vec's built-in sorting, please. I suspect that's more representative of real world workloads than 'is-prime' or the quadratic sorting. Thanks! |
Draft
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Sandbox branch, shouldn't be merged.
Based on #813 . Hardcodes
is_primeintobin/and adds theis_prime/sp1project at the top level. I've been running:and
Results are very interesting 👀 👀 . But can we confirm them @matthiasgoergens ?
Later Edit:
Ignore the Fibonacci names, it's auto-generated by
sp1.Consult https://docs.succinct.xyz/docs/getting-started/quickstart for more.