⛰️Technology base

1. Safer user experience (safety guarantee from the user's perspective)

NJWC provides flexible key management functions. For example, users can rotate (rotate) keys and provide transaction pre-execution functions, so that users can predict the execution results of transactions before they actually sign the transaction, ensuring the transparency of the results.

2. High throughput and low latency

NJWC subdivides transaction processing into five modules: transaction dissemination, block metadata ordering, parallel transaction execution, and batch storage , ledger certification. This modularization plus pipeline processing method improves the parallelism of the entire transaction processing process.

3. Block-STM parallel execution engine

In the Transaction parallel execution module, NJWC uses the Block-STM engine to realize the maximum parallel execution that supports atomicity without the need for the programmer to pre-declare the characteristics of data read and write operations.

4. Support upgrade

NJWC's modular architecture design provides an embedded on-chain change management protocol, making upgrades easy to implement.

5. Sharding

Both NJWC's modular design and parallel execution engine naturally support fragmentation, which facilitates the horizontal expansion of NJWC's throughput.

6. GAN

GAN is a deep neural network architecture consisting of a generative network and a discriminative network. The generating network generates "fake" data and tries to deceive the discriminative network; the discriminative network verifies the authenticity of the generated data and tries to correctly identify all the "false" data.

7. Diffusion Model Diffusion Model

Diffusion models are a new class of generative models that generate a variety of high-resolution images. They have attracted a lot of attention after OpenAI, Nvidia and Google managed to train large models. Example architectures based on diffusion models include GLIDE, DALLE-2, Imagen and the fully open source stable diffusion. Diffusion models already hold the potential to be representative of the next generation of image generation models. Taking DALL-E as an example, it can generate images directly through text descriptions, allowing computers to have human creativity.