Binius and Circle STARKs were the two major highlights last year. 

but if this is true, then does it make sense to keep working in the circle STARKs direction? everyone should move to multi linear polynomials

4/ One takeaway: 256-bit or binary fields are better than 31-bit or 64-bit fields for SNARK performance. One reason among many: committing to 0s is basically free with elliptic curves and 30x faster in binary fields than in 31-bit fields.

5/ The key performance dichotomy in SNARK design isn’t “big fields vs. small fields” or “hashing vs. curves”. It’s the sum-check protocol and multilinear polynomials vs. univariate polynomials.

Justin Thaler

Binius and Circle STARKs were the two major highlights last year. 

but if this is true, then does it make sense to keep working in the circle STARKs direction? everyone should move to multi linear polynomials

https://x.com/succinctjt/status/1882435768632947031?s=46

Circle starks was not a highlight fam.

What was the performance increase in the paper?

perhaps only an improvement when compared to big field STARKs.

so, maybe removing FFTs entirely is the way to go. idk

Working on building zkTLS infra/products. A ruthless empiricist and meme connoisseur.

/opacity
/eulerlagrange

EC and sumcheck means only MSM no FFT. I understand the small field work for scaling blockchain is with fast hashed based monolithic provers… right now.

But tend to agree with Thaler on his view more generally. How much is the speed up engineering efforts? Why does he claim JOLT is the fastest zkVM when both JOLT and HyperNova only commit to the witness? They have the same prover time.