A purely peer-to-peer graphic processing would allow computers to have high level processing power without relying on local machine spec. As represented by Moore’s Law, the information processing capability of computers has developed dramatically.

We propose a mechanism to distribute computer information processing capabilities and equalize disparity, thereby benefiting all nodes on the peer-to-peer network. Angelium enables decision making based on the proof-of-work consensus algorithm and distribution on the network of computing power used for proof-of-work of Bitcoin.

For distribution, we propose a compression process (Emojithon) for efficient computation processing and a distributed network connecting each node, block-based reward design for contributing nodes.

1. Introduction

We solve the problem of the waste of computing processing used for proof-of-work by collaborative peer-to-peer graphics processing.


High resolution 3D graphics require high computing power. Although the range of expression increases with 3D graphics, waiting time increases to obtain it. The decentralized consensus algorithm represented by Bitcoin uses computing processing capability as proof-of-work. This large amount of computing processing is only used for verification of transaction. There is a certain criticism that the calculation to perform proof-of-work requires a lot of electric power and it is used only for the proof of the electronic transaction and its is won’t be used elsewhere.

There are three major problems in computing processing ability used for proof-of-work.

The first problem is power consumption. The hashing rate of the Bitcoin continues to increase radically, and the consumed electric power is used only for the work certification. Excessive power consumption affects the global environment.  Bitcoin also wastes chip making resources (ASIC is a one-trick-pony)

The second problem is governance and scalability. It took a long time to solve the Bitcoin block size problem. This resulted in an increase in settlement time and fee.

The third problem is that big mining companies monopolize. Even the decentralized network is now controlled by just two large mining pool companies.  Bitcoin has changed from highly decentralized, to super centralized with the power in the hands of just two groups (they can collaborate behind the scenes to pull off small 51% attacks.

Proof of Waste

There are many technologies that improve the proof-of-work problem, such as PoS, PoI. There are also projects that suppress excessive competition by changing the hash algorithm such as Ethash, Equihash. With these solutions, the amount of work proof can be reduced efficiently. However, the fact that energy is used only for proof-of-work remains unchanged, only a few examples of useful-proof-of-work exist like: PrimeCoin.

Our solution

To enable to construct an efficient distributed ecosystem by converting the computing processing ability for certifying the work for approving the transaction to the value in the network. We provide a mechanism that the nodes contribute for the computing processing power required by other nodes certifies the task. Nodes contributing to the network can receive reward from the network. The basic consensus building algorithm and the scheme of compensation follow the Bitcoin. In addition, a hashing algorithm combining fractal structure and Equihash prevents major mining companies from monopolizing.

2. Proof-of-Rendering

In a game with many players, each player’s computer often performs the same graphics processing in each local environment. Other computers can undertake graphics processing and share it with other players, thereby eliminating waste and providing players with a higher-level user experience.

Description of Proof of Rendering

Uses computation processing of users on the network for work certification for transaction approval and provides results of arithmetic processing as valuable data on the network. To the nodes on the network, two types of tools are provided, namely a computing engine (mining node) and a receiving port (normal node) depending on the computer spec. A node with a high computer specification contributes more to the network. A node with a low computer specification exists as a normal node, and receives data processed by a mining node using a BitTorrent.

Rendering power share

The share of rendering power can be explained from two aspects.

The first is input / output as a device. By saving the data created by the 3D engine in advance, it is efficiently transmitted to the recipient’s computer. The 3D engine uses Blender which has open source and powerful processing capability.

The second is network processing as a node. The normal node queries the data on the network, the mining node provides the target data, and the client checks the hash (checksum) to see if it matches the checksum inside the blockchain for that given 3d object. This process solves the graphics processing in a P2P manner in a simultaneous connection manner.

Transmission and reception of 3D engine and receiver

Sharing the rendering power by sending the data rendered by the 3D engine to the receiver.

The share of the rendering power is communicated in the following flow.

  1. The recipient queries the 3D engine for data already processed
  2. When data exists in the 3D engine, data is sent from the 3D engine to the recipient
    1. First, 3D polygon data with a small number of underlying polygons is transmitted
    2. Send texture data with low resolution
    3. Send animation data
    4. Send texture data in resolution
    5. Send 3D polygon data with many polygons
    6. Send texture data with high resolution
    7. Send lighting and shade data
  3. Repeat 1 and 2 to draw 3D objects in a wide range

Function of node

When the mining node is connected to the network, it shares the rendering power to the normal node. Processing on the network uses P2P network of BitTorrent.

The operation of the node is carried out in the following flow

  1. The normal node requests the data information which the normal node needs on the network
  2. The computing engine of the mining node draws and saves the original data by calculation processing
  3. The rendered data is provided on the network through a mining node
  4. The data provided by the mining node and the node requested by the normal node are matched
  5. If it matches the request of the normal node, the data continues to be provided from the mining node

The more mining nodes, the more it is possible to provide a lot of 3D graphic data to normal nodes.

3. New Hash Form/Format

Angelium adopts its own hash format by combining a fractal structure and Equihash adopted in Zcash. Combine the hash value obtained by encrypting the fractal structure + 3d object checksums with the hash value generated by Equihash to create the final hash value. From a high-level perspective this ​​conforms to the Bitcoin transaction structure.

4. Network

The network big data is distributed by BitTorrent. In order to make BitTorrent scalable, we build a numerical allocation network that bundles multiple BitTorrent. The numeral assignment network gives IDs to a plurality of BitTorrent scattered like a tile and manages which data each BitTorrent stores. An upper limit is set for the number of nodes that can participate in each BitTorrent. The BitTorrent seed functions as a mining node and the leecher connects to the network as a normal node.

Procedure of network execution

  1. Hash a part of the data of 3D space and store it in BitTorrent
  2. A node requests data
  3. The seed that stored the data transmits hashed data to the network
  4. The hashed data is transmitted to the node which sent the request through the reacher on the BitTorrent
  5. Complete communication when the node receives all the data

Description on network disconnection

The mining node can distribute the location in the 3D space where the 3D engine operation processing is performed by geolocation. Mining nodes share arithmetic processing based on their respective regions. If the network is temporarily disconnected, it can continue to exist as multiple networks. At this time, the geolocation is reassigned and the effect on the network is kept to a minimum. When the network disconnection is recovered, and all the networks are connected again, the 3D space that existed in different networks exist independently. Then geolocation is again allocated to optimize the impact on the network.

5. Efficient compression of object data

Introduction of Emojithon

Emojithon is a creator tool for creating 3D space. Since it is saved in rich text format, it is possible to simply share 3D art created by creators. Also, unlike conventional 3D art, it can be rendered lightly because it is represented by very small data.

Emojithon features

Emojithon has roughly two functions. First, it is a function that allows creator to create 3D art that moves with small data by typing Emoji code which Emojithon prepared from the beginning. Creators can make 3D art by writing code with WYGWIG directly or by keyboard directly without writing complicated code.  The user can also bypass our WYGWIG editor, and simply type the language in any rich text editor (like MSWord), and later copy and paste their emoji-code into the Emojithon editor.

The second is the ability to create a new Emoji that creators can use with Emojithon. You can create 3D objects of completely new, beautiful castles and strange monsters on Emojithon and assign Emoji code to it. Creators can create 3D space using Emoji made by another creator.

Emojithon's specialty as CGM

Emojithon is unique as CGM is that creators can demonstrate imagination at multiple levels. Creators can design 3D space by setting objects like Minecraft according to the given Emojithon’s assets. Creator who wants to demonstrate more creativity as Emoji creator can create objects with free idea and convert it into light code to produce 3D space never seen by anyone. The most interesting part is that you can combine Emoji created by other creators and create your own Emoji so that you can create a whole new 3D space through secondary creation.

Issues to be solved by Emojithon

Emojithon not only provides a 3D space creation tool that works lightly for creators, it also lightens the rendering process of 3D objects on the network. 3D objects rendered by Emojithon can be written with small code. By overwhelmingly shortening the code, compression of the 3D rendering process was made possible.

6. Conclusion

By utilizing the P2P network in Angelium, we eliminate the waste generated when the computer renders 3D space. There is a problem that calculation processing capability cannot be used for other task certification for transaction approval. This problem is solved by providing the computation processing capacity used for work certification as a value to network participants. As a further excellent part, unlike competitive mining work so far, network participants can cooperate and increase the value of the network. We also propose new scalability when nodes and transactions increase. A large-scale scalability was realized by constructing a network nesting multiple BitTorrent. The P2P network which can respond flexibly to the increase of participants will solve the problems of the non-centralized network so far and provide a platform on which participants can share value.