THUBA research report｜render decentralized computing power - also talks about the depin track

Introduction to Render

• Render Network is a decentralized rendering platform that connects people who need GPU computing power with providers who have excess computing power, mainly for artificial intelligence, virtual reality, and media and content creation. Officially launched publicly in April 2020.

• Render Network's pricing model is dynamic and competitive, taking into account the complexity, urgency and available resources of the job .

• Render Network distributes workloads among available GPU providers through a peer-to-peer network of RNDR tokens, incentivizing nodes to provide excess Calculate ability.

• In December 2023, Render completed the migration from Ethereum to Solana, bringing new capabilities to it, such as Live streaming and dynamic NFTs, and state compression.

Introduction to the concept of DePIN

DePIN (Decentralized Physical Infrastructure Networks), decentralized physical infrastructure. Messari's research report divides DePIN into two main areas: digital resource network and physical resource network. Digital resource networks include storage, computing, and bandwidth, while physical resource networks focus on hardware-related areas such as wireless networks, geospatial networks, mobile networks, and energy networks.

PoPW (Proof of Physical Work), physical workload proof mechanism. Incentivize individuals to produce verifiable work that promotes the development and efficient use of real-world infrastructure.

According to the definition of "decentralized physical infrastructure", almost all distributed networks can be classified as DePIN:

Filecoin, Arweave: a decentralized hard drive

Ethereum: a decentralized computer

< p style="text-align: left;">Bitcoin: Decentralized Ledger

Figure 1 DePIN track mapping-1

Figure 2 DePIN track mapping-2

Based on homogeneity or not, The concept can be subdivided:

DeREN "decentralized resource network" vs DePIN "decentralized infrastructure network"

DePIN: A network model with geographical location as a key element, using an incentive mechanism to deploy geographically specific hardware devices to generate non-homogeneous resources. How do you understand non-substitutable? Mobile phone base stations (China Tower) can be seen deployed throughout the city. But for the same base station, base station A is in Tsim Sha Tsui and base station B is in Sha Tin. However, due to their different geographical locations, they cannot be replaced with each other. In this sense, each base station is a unique and non-homogeneous resource.

DeREN: a resource-based network that is weakly dependent on geographical location. It establishes a market through incentive mechanisms and increases the availability of existing or idle alternative resources. supply. These resources are generated by geographically independent hardware. In the field of cloud services, DeRENs are similar to cloud infrastructure as a service (IaaS). For example, with Avware storage, the user experience is often less tied to the user's geographic location and the service is highly homogeneous.

DePIN Industry Research< /h2>

Current status of the traditional ICT industry

The current ICT industry is facing the following Two obvious dilemmas:

1. High barriers to entry, insufficient competition, and pricing monopolized by giants

Building data storage and cloud computing clusters requires a large investment in hardware purchase, site rental and personnel hiring. Such high costs make almost only giant companies capable of joining this industry. For example, the combined market share of AWS, Microsoft Azure, Google Cloud, and Alibaba Cloud in the field of cloud computing and data storage is close to 70%. This results in high costs.

2. The resource utilization rate of centralized infrastructure is low

The low utilization rate of centralized infrastructure resources is very important today. is a major challenge in business operations. This problem is particularly acute in cloud computing environments, where companies typically allocate large budgets to cloud services. According to a recent Flexera report (2022), a worrying trend has emerged where, on average, 32% of a company’s cloud budget is wasted, meaning that one-third of a company’s resources are left idle after spending on the cloud. This resulted in huge financial losses.

Advantages of DePIN

1. From Capital-intensive industries transform into P2P/P2B models

Lower the threshold for enterprise participation, break the price monopoly, and provide users with more economical and affordable choices< /p>

2. Reuse idle resources and improve resource utilization efficiency

In the traditional economic model, many resources are idle and fail to realize their potential value. Many users hold many idle resources, whether storage, computing power or data. The key lies in how to mobilize these resources. Through the incentive mechanism, DePIN encourages users to share and utilize their resources, maximizing resource utilization.

Take the cloud as an example. According to a Flexera report, the effective usage rate of cloud purchases by enterprises in 2022 is only 68%, which means that 32% of cloud resources are wasted. . Considering Gartner expects cloud spending to reach nearly $500 billion in 2022, that means a rough estimate of $160 billion in cloud spending is wasted. These wasted resources can be utilized through DePIN.

3. Eliminate middlemen and make money flow more efficiently

In traditional O2O projects such as Meituan, Dianping, and Didi, the platforms charge high intermediary fees, which are equivalent to a tax wedge, and social efficiency has not reached optimal status. There have been many projects on the DePIN track trying to solve this problem, such as taxi-hailing app Drife, food delivery app devourgo, Dianping app Ketchup Republic, etc.

Relationship with Solana

depin protocols often require real-time responses and low transaction fees. Many protocols now prefer to use existing ecosystems rather than building their own chains. Currently, many of depin’s leading protocols migrate to Solana.

Solana is a perfect fit for DePIN / PoPW due to its high-speed blockchain and features such as 0.5 second block time and direct processing of validators. Additionally, Solana’s cNFT provides more cost-effective authorization certificates for DePIN/PoPW nodes, which is a common practice.

More and more DePIN/PoPW protocols choose Solana as their base layer. Nearly all segments of the DePIN/PoPW protocol have been successfully integrated with Solana, providing proof of concept to those who were initially skeptical of these protocols.

DePIN market size

Current market value ranking in the DePIN field Judging from the top 10 projects, most of them belong to the storage and computing fields. (Ryze Labs, October 2023)

< h2 style="text-align: left;">Detailed explanation of Render

1. Introduction to the project party

Render Network was founded by OTOY, a company specializing in 3D rendering technology. It is also the world's first commercial rendering software Octane Render that fully utilizes GPUs. Creator, OTOY is a cloud graphics company providing rendering solutions to Hollywood studios and artists. The idea for the Render Network was born in 2017, when OTOY published a white paper outlining the technical details of the network.

You can directly access render through Octane Render's subscription account. (It is also the only way to use the computing power of Render Network)

Any task that can be achieved with GPU rendering using Octane Render can be rendered on Render Network, including movies , animation, games, virtual reality scenes, etc.

2. Token Economics Model

The architecture of Render network

In its current state, Render Network consists of two main layers:

1. Off-chain rendering network, including creators, node operators, rendering network and rendering application layer suppliers. Node operators are the GPU nodes that power the network.

Blockchain layer that handles payments via RENDER and smart contracts.

2. Node operators and creators interact with each other in a distributed network model arbitrated by the server core network infrastructure.

OctaneBench is a proprietary benchmark tool (and the most popular GPU rendering benchmark today) that measures OctaneBench points per hour (OBh) Represents the GPU rendering speed, thereby determining the pricing of rendering tasks. OctaneBench is used to standardize and benchmark GPU performance, ensuring creators receive consistent performance pricing when requesting work from node operators. The cost of any task is measured in OctaneBench units and determined by the Multi-Tier Pricing (MTP) algorithm.

RENDER token

• RNDR acts as a utility token for on-demand GPU rendering services. Creators are paid for rendering work via RNDR. They have access to a variety of GPU providers with varying speed, security, and price levels. Currently, pricing is divided into three tiers: Tier 1 (Trusted Partner), Tier 2 (Preferred), and Tier 3 (Economic). Tier 2 and Tier 3 are already available, with Tier 1 coming in the future.

• Secondly, as a reward mechanism for GPU providers. Node operators can use idle GPU power to render jobs on the rendering network and earn RNDR tokens in return. The number of RNDR tokens they earn depends on the tier, duration, and complexity of the rendering job. Node operators can also stake their RNDR tokens to increase their reputation and gain access to more rendering jobs.

• Third, RNDR is the governance token of the protocol, allowing token holders to influence the governance of the rendering network.

Token allocation:

• 40% for development

• 25% for growth

  < /li>

• 20% for operations

• 10% For reserves

• 5% for creation

Token Economics Model

Dynamic Pricing Model

The rendering network's dynamic pricing mechanism ensures competitive service costs for content creators and fair compensation for GPU providers. The pricing model takes into account factors such as job complexity, urgency and resource availability to determine the best cost to provide services. Its dynamic approach does not rely on static or manual pricing systems, but instead adapts to the needs of its user base and ecosystem.

Render Network rendering service price

3. Pairing algorithm

Uses a matching algorithm to connect content creators with appropriate GPU providers. By evaluating factors such as GPU type, capacity, location, and provider reputation, Render Network ensures that rendering jobs are allocated efficiently across resources to maximize the efficiency, speed, and cost-effectiveness of your rendering services.

Reputation scores of task creators and node operators

The reputation of a node operator is built by getting the job done promptly and accurately.

Creators build reputation scores through their successful use of the web. As creators accumulate a track record of successful work and minimize user error, they gain access to more concurrent GPU nodes.

Reputation scores help the network distribute work efficiently and reduce unintentional congestion caused by failed rendering or malicious Sybil attacks. The reputation scoring system is regularly updated to increase network efficiency as workload complexity and network service delivery continue to evolve.

Priority of task assignment

Reputation of the creator The score affects its task allocation priority in two ways:

1. Task creators with higher reputation scores can run more concurrent tasks on the network.

2. Add a priority modifier to each job based on the creator's current reputation score, so that creators with high reputation scores have fewer tasks than creators with low reputation Equivalent tasks from score creators are "more attractive" to nodes.

Node allocation priority

Node allocation is for The process by which nodes allocate rendering work to various tasks.

When tasks are sent to the rendering network, they are assigned based on factors such as selected layer, hardware requirements, network time, user reputation, node reputation score, and OctaneBench score. to node operators.

4. Privacy and security

During rendering or processing Privacy and security are of the utmost importance when working with any creative or confidential material. To protect the privacy of creators and assets on the network, the system uses a combination of end-to-end encryption, hashing, virtualization, and secure centralized storage:

Every scenario are broken down into many individual assets, hashed and encrypted when uploaded to the network.

Rendered output is encrypted before being sent over the network. Any assets stored in memory or on disk are always encrypted.

A watermark will be added to each frame before downloading to ensure that payment has been completed before downloading.

5. ORBX format (high compatibility)

Render Network uses the open ORBX file format and streaming framework to support fully distributed rendering. ORBX is a container format for capturing scene data (assets) and XML rendering graphs that describe scene semantics. Just like a web page, it can be cached to an archive (.orbx file) or streamed from a URL or URI via raw UDP/tcp or web wss or https. The ORBX format supports over 20 industry-leading DCC (digital content creation) tools and includes industry-standard sub-formats such as Alembic, OpenVDB, EXR, Open Shader Language (OSSL) and gLTF. ORBX export is increasingly integrated natively into the Octane Render interface, enabling a simplified one-click export process to the Render Network's third-party plug-in host, and improved export speed and compatibility with extended third-party plug-in effects and libraries .

Reference materials

Research report

1. Will the two-way rush between Solana and DePIN bring about a DePIN summer?

https://mp.weixin.qq.com/s/Se4WOYst8-o-JjocjPydxw

2. Ryze Labs 10,000 words research report: comprehensive interpretation of the DePIN track.

https://mp.weixin.qq.com/s/SBWhxyMTlwsBsVIbHS1CuA

3. Web3+AIoT=DePIN, look at the future development of DePIN from the Worldcoin project of the founder of OpenAI.

https://yakihonne.com/article/naddr1qq2kwnmgg424jsm829vhjaee0gkhgs20fuerqq3q08nx6y5gu3dz6urhhlcga0ktxcm9x05pdml2hy3gykrvj0r6qnrqxpqqqp65wzs548q

4 . In-depth interpretation of DePIN track potential projects.

https://mirror.xyz/0x30bF18409211FB048b8Abf44c27052c93cF329F2/kAQZ7kAdJYHi8ZMma0r6H_A81lM4-zSel2itrDhQk90

5. One article to explore investment opportunities in the Depin track | Trend X Research Institute. https://mirror.xyz/0x901eA9e4c3D637C73a368B2bCD0E708992baE49C/PG9n-yKReNcMdFwRBZHkapKsDW6QuH0hx6oH_b4OoQg

6. Mapping the Landscape of Decentralized Physical Infrastructure Networks.

https://iotex.io/blog/depin-landscape-map/

7. ZKML+DePIN will what happened? A Deep Dive Into DePin (Decentralized Physical Infrastructure).

https://coinmarketcap.com/academy/article/a-deep-dive-into-depin-decentralized-physical -infrastructure

9.https://coinmarketcap.com/academy/article/what-is-render-network-rndr

10.https://globalcoinresearch.com/2023/04/26/render-network-scaling-rendering-for-the-future/

11.https://www.coinfeeds.io/research/an-introduction-to-the-render-network

12. https://foresightnews.pro/article/detail/40042

Project link

1. filecoin https://filecoin.io/zh-cn/

2. Arweave https://www.arweave.org /

3. helium https://www.helium.com/

4. render https://rendernetwork.com/

5. livepeer https://livepeer.org/zh

6. space and time

 https://www.spaceandtime.io/

7. EMC https://edgematrix.com/en/

8. PowerPod https://www.powerpod.pro/< /p>

9.kandola https://kandola.network/index.html#home