-
Notifications
You must be signed in to change notification settings - Fork 160
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.
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
Move all buses and tables to LogUp-GKR #1505
base: logup-gkr
Are you sure you want to change the base?
Conversation
5e9f719
to
bc89f1f
Compare
The preliminary results look pretty bad with the current branch more than doubling the proving time. The entire program on
To fix this performance gaps, as previously discussed, we can introduce a The radical increase in fractions is due to how to adapt the constraint equation for "standard multiset checks" (last equation in this section). Adding rows to the table becomes very expensive. That is, when adapted to LogUp becomes which itself adds 11 fractions. next
logup-gkr branch
This branch
|
logup-gkr branch
this branch
|
This with aux segment size halved
Next
|
c87de25
to
b7de6a8
Compare
a18f2b4
to
b4a52a3
Compare
b4a52a3
to
e7093e1
Compare
Latest benchmarks (parallel version, 64 fractions containing all tables except for the chiplets bus and the stack overflow table)TL;DR: the RPO-based benchmark is exactly at par with
This branch (all winterfell PRs merged)
next(only show the RPO-based benchmark; blake3 is unchanged and available in previous comments)
|
8de48f9
to
77bb46d
Compare
77bb46d
to
eafc592
Compare
eafc592
to
e21d358
Compare
Left to do:
MidenLogUpGkrEval
TODO(plafer)