SuperScalar Ranked №1 in Proof Generation of Taiko Testnet

SuperScalar
2 min readOct 16, 2023

SuperScalar, the pioneer company in ZKP (zero-knowledge proof) acceleration solutions, has been participating as a prover in the Taiko Testnet since Alpha-3, Alpha-4, and Alpha-5. In Taiko A3, SuperScalar has already generated over 300,000 proofs, the largest number of any prover.

Taiko is a Type-1 Ethereum ZK-rollup and the first ZK-rollup project to adopt decentralized sequencers and provers from its inception. SuperScalar provides both GPU and FPGA solutions for Taiko projects.

The SuperScalar GPU solution has been running in Taiko A3, A4, and A5 projects, demonstrating a 10x acceleration compared to the CPU version and serving as the fastest prover in Taiko A3.

SuperScalar’s GPU solution is the fastest prover and ranks №1 in proofs’ share.

The performance improvements of the GPU solution come from accelerating modules such as MSM, NTT, evaluator, and others. Additionally, it incorporates various technologies to avoid or minimize memory copying.

The Taiko A3 Testnet has been operational since June 2023. Through a partnership with ZKPool, SuperScalar has won the largest share of proofs in the Taiko Testnet.

Here are the results achieved by SuperScalar’s solution in Taiko A3 Tetnet:

  • The data is sourced from ZKPool’s statistics.

Recently, SuperScalar launched the K10, the first FPGA ZKP miner for the Aleo project. They have also introduced the K11, the latest version of the FPGA accelerator for zkEVM solutions. The K11 can accelerate Aleo and zkEVM solutions such as Taiko and Scroll.

SuperScalar will continue participating as a prover in the Taiko project, including A5 and later Testnet and Mainnet. SuperScalar is also working on accelerating other ZKP projects and providing GPU, FPGA, or ASIC solutions.

For more information, please get in touch with SuperScalar.

Website: superscalar.io

Email: sales@superscalar.io

Twitter: @SuperScalar_io

--

--

SuperScalar

Accelerate Zero Knowledge Proof Accelerate Scaled and Privacy Computation http://superscalar.io