站在巨人的肩膀上：分层设计在 Nervos 生态系统中的 layer 2 区块链上的实际应用

1.

2. Layer 2 on Nervos CKB 肖雪洁 @ 秘猿科技

3. Or 站在巨人的肩膀上：分层设计在 Nervos 生态系统中的 layer 2 区块链上的实际应用

4. About me • 肖雪洁 • https://xuejie.space/ • CTO @ 秘猿科技

5. 秘猿科技

6. Nervos

7. Blockchains are gaining popularity

8. Scaling Problem in Blockchains Source: https://howmuch.net/articles/crypto-transaction-speeds-compared

9. Sharding: a popular choice Source: https://vitalik.ca/general/2021/04/07/sharding.html

10. Sounds familiar? Source: https://www.logicbig.com/tutorials/core-java-tutorial/java-multi-threading/java-intrinsic-locks.html

11. Sounds familiar? Source: https://kgrvamsi.wordpress.com/2013/05/28/riak-in-depth/

12. Is sharding the best solution?

13. TCP/IP Source: https://fiberbit.com.tw/tcpip-model-vs-osi-model/

14. NewSQL Source: https://medium.com/rabiprasadpadhy/google-spanner-a-newsql-journey-or-beginning-of-the-end-of-the-nosql-era- 3785be8e5c38

15. Layers in NewSQL Source: https://www.cockroachlabs.com/docs/v21.1/architecture/overview#layers

16. Layers in NewSQL Source: https://docs.pingcap.com/tidb/stable/tidb-architecture

17. Linux Source: http://www.versolibre.org/versus/lookatlinux

18. We didn’t choose layered design to be different. Layers long exist in the software industry.

19. Layered blockchain design dapp Layer 2 dapp Layer 1

20. CKB: layer 1 blockchain built for layer 2 • Proof of Work • Custom designed P2P propagation network • CKB VM • Cell(generalized UTXO) based model

21. P2P component is also a layered design Source: https://talk.nervos.org/t/p2p/1742

22. CKB VM • RISC-V ISA • A thin, secure layer on top of native hardware • Performant implementation to support crypto algorithms • Vector programming is on the way

23. Custom cryptographic primitives

24. UTXOs Source: https://academy.horizen.io/technology/expert/utxo-vs-account-model/

25. Accounts Source: https://academy.horizen.io/technology/expert/utxo-vs-account-model/

26. UTXOs vs accounts Source: https://academy.horizen.io/technology/expert/utxo-vs-account-model/

27. UTXOs are naturally paralleled Source: https://en.wikipedia.org/wiki/Thread_pool

28. Programming on UTXOs is HARD!

29. Layer 2 for the rescue

30. Introducing Godwoken A generic rollup framework on CKB

31. Godwoken

32. Optimistic Rollup

33. Optimistic Rollup (reverts)

34. polyjuice • An Ethereum compatible layer runs on Godwoken • Compiles evmone, an EVM implementation directly into CKB VM runnable code • Maximum compatibility can be achieved: • Official EVM implementation is used, no unsupported EVM opcodes • The exact signing scheme from EVM is supported, Metamask can be used with no changes • Other VMs are also possible, such as Move from libra

35. Things we don’t need in layer 2 • No hassles dealing with UTXOs • No forks • No P2P network for block propagation • No disputes • Layer 1 VM works as the final judge • Minimal consensus needed

36. And there’s more on CKB • Axon: CKB’s sidechain/parachain design • CKB’s channel network for user defined tokens • State channels for games

37. Aim at higher performance with layer 2 • CKB: ~100 TPS • Godwoken: ~600 TPS • Godwoken with BLS: ~2000 TPS • Axon: >100000 TPS

38. Latency is also a metric • CKB: 15 seconds * 4 blocks • Godwoken: 90 seconds • Channel network: 0 seconds

39. The Blockchain Trilemma Source: https://www.quora.com/What-is-the-blockchain-scalability-trilemma

40. The Blockchain Trilemma, layered edition ✔ Layer 2 ✔ Layer 1 Source: https://www.quora.com/What-is-the-blockchain-scalability-trilemma

41. Looking into the future • Software are more and more complex, blockchains are no exceptions • The only viable solution for the future, is to encapsulate complexities via abstractions/layers • We firmly believe more blockchains will take the pill of layers in the future

42. Computer Science is a science of abstraction -creating the right model for a problem and devising the appropriate mechanizable techniques to solve it. - A. Aho and J. Ullman

43.