The Render Network is a global GPU marketplace built on the blockchain to bring efficient computation to the masses. Suppliers of idle GPU capacity can plug into the network and earn passive income in the form of RNDR tokens for completing rendering jobs. The RNDR token is up ~330% YTD which correlates with its increase in network activity. In this article, we’ll discuss the Render Network, how it works, and how users can plug in their GPUs and earn RNDR.
The Render Network is a decentralized blockchain-based peer-to-peer compute platform leveraging industry-leading OTOY Inc. software to bring together GPU Providers and GPU Requestors. Providers are users who have unused or idle GPU capacity that can be used to contribute to Render’s global network and earn Render (RNDR) tokens in exchange for rendering data for GPU Requestors. Requestors are users looking for affordable and efficient GPU processing power to render anything from 3D animation to processing complex tasks.
Render was founded in 2016 by Jules Urbach, CEO of OTOY, a software company responsible for developing OctaneRender. He has over 25 years of experience in computer graphics, streaming, and 3D rendering. In addition to eight core team members, eight advisors are behind the Render Network. Several prominent names include:
GPU Requesters (Creators) submit rendering jobs to the RNDR Network with their output and preferences. GPU Providers, called Node Operators with the minimum recommended hardware requirements, can select jobs from the marketplace. Each rendering job is denominated in fiat and can be purchased with fiat or the proportion in RNDR. Creators then have 24-48 hours to approve a rendering job and pay upfront for services by depositing Render tokens (RNDR) into a smart contract. The RNDR Network facilitates coordinating and allocating computing tasks between Creators and Node Operators.
Source: Render Network
The Network optimizes and load balances rendering jobs based on GPU types to match job requests and Creator's requirements. These jobs are rendered in parallel across multiple Node Operators to improve rendering speeds and completion times. After a Creator reviews their completed rendering(s) and approves, RNDR tokens held in the smart contract are released pro-rata to all participating Node Operators within 7-10 days.
The Render Network (RNDR) is built on two core technologies: OctaneRender and ORBX.
The hashing of ORBX on the RNDR Network creates hashed records of all assets and settings in the render graph, using the hash to establish provenance for every rendering processed on the Network. The immutability and traceability inherited on the blockchain can facilitate tracking ownership transfer, recomposition, and re-usage rights.
OctaneRender and ORBX are extensible, supporting a wide range of 3D content creation tools, and can be expanded to new tools via Octane's SDK.
The pricing for compute power on the RNDR Network is based on a unit of work known as OctaneBench, measured in seconds and concurrent OctaneBench power. The RNDR token is mapped to this unit of work, and the amount of OctaneBench work varies according to the user's selected preferences and the overall GPU supply and demand on the Network.
OctaneBench allows users to benchmark their GPUs using OctaneRender, standardizing the comparability of a GPU's performance and ultimately dictating the pricing of rendering tasks. This ensures that Creators experience consistent performance-based pricing when requesting work from Node Operators. A reputation scoring system discussed below is in place for both Node Operators and Creators to facilitate efficient job assignment and mitigate network congestion from failed renders or malicious attacks.
The RNDR Network offers a Multi-Tier Pricing (MTP) framework to maximize efficiency for Creators and Providers. Additionally, the MTP framework works with the Networks' Reputation, Allocation, and Governance systems to ensure optimal utilization.
Source: Render Network
The MTP framework enables creators with different preferences (speed, cost, or security) to utilize the Network optimally and node operators with varying levels of trust and hardware capabilities to be appropriately rewarded. Additionally, the pricing algorithm for tiers is updated periodically based on changes in GPU performance, current GPU cloud rendering costs, operational costs, and the Networks' supply and demand.
Reputation scoring plays a crucial role in resource allocation and maintaining the integrity of the Network. Creators with higher reputation scores gain access to more concurrent nodes while rendering, increasing their efficiency. Node operators with better reputation scores receive job assignments faster than nodes with lower reputation scores. Users with better reputation scores get priority during periods of limited available resources (e.g., GPU compute).
Suppose a creator consistently rejects work from nodes with high success rates. Excessive failures can lead to job termination and negatively impact the Creator's reputation score. Similarly, nodes that repeatedly fail to process frames adequately may receive lower rankings, negatively impacting their reputation scores.
The Network incentivizes users to build positive reputation scores and discourages reputation gaming. Creator accounts are tied to unique Octane Accounts, which include software license entitlements. Better reputation scores grant access to more nodes, creating an incentive to build a positive history. Starting over with a new account is disadvantageous due to the reputation system.
The combination of human job confirmation and algorithmic rendering checks used to validate work on the Network is referred to as "proof-of-render." It is essential in assessing the reputation score and resource allocation within the system.
Resource allocation on the Render Network falls into Job Allocation and Node Allocation.
Source: Render Network
The Render (RNDR) token is an ERC-20 compatible utility token launched in 2017. It is used by Creators to pay for rendering services, reward Node operators for completing GPU-intensive jobs, and vote on governance. In addition to RNDR, Creators can purchase RNDR credits using fiat platforms (e.g., Stripe and Paypal) to use the Network. Since its mainnet launch, Render is now deployed on MATIC to reduce transaction fees incurred when distributing funds to Node operators with plans of expanding to Solana.
The total supply of RNDR tokens was capped at 536,870,912 at the token generation event (TGE) in October 2017. However, in January 2023, the Render community voted to approve RNP-001, discussed below, causing the supply of RNDR to increase by 107,374,182 over time, bringing its maximum supply to 644,245,094.
Node operators are entitled to a percentage of token emissions for jobs completed in each epoch. The Render Network also takes a 5% network fee on all transactions to help cover any infrastructure or operations costs.
The emission schedule for the RNDR token was designed to control the total circulating supply of RNDR while rewarding network participants in a structured manner. The schedule is divided into two periods: Launch and Growth.
It is important to note that the emission schedule is subject to forthcoming governance procedures and may be adjusted. As of this report, approximately ~369.6 million (~57%) RNDR tokens are in circulation.
Most gaming PCs today are built with equipment that could get someone started as a Node Operator on the Render Network. Aside from the monthly cost of securing stable internet connectivity, it’d cost approximately ~$560 to purchase appropriate equipment. Below, we’ll review an example of the infrastructure and approximate costs for an entry-level setup.
For those interested in joining the network, users can sign up and get added to an onboarding queue. Upon approval, the new Node Operator will be required to download the Render Network client, establish an Ethereum wallet address, and have already met the minimum recommended hardware requirements.
The profitability of running a Render Node is determined by the number of GPUs running, equipment type, OctaneBench score (OB), Tier, and operational expenses.
Below we’ve come up with estimated profitability using the Render’s profitability calculator and the following inputs: Average OB result for 1 GTX 1080: 129.75, 175 Watts under load, national average $/kW of $0.23, estimated usage at 10%, and price of RNDR (~$1.70).
According to the Render Foundation, Tier 3 usage has been growing, relative to Tier 2 jobs. It’s worth noting that these figures are estimates and likely unadjusted for network nuances. However, the data suggest a relatively inexpensive setup could be profitable.
The Render Network's adoption could be measured by the number of Node Operators on the Network, the amount of RNDR disbursed by Creators, and the number of projects rendered.
Source: Dune @lviswang
Despite a lingering bear market, creators requesting rendering services grew for three consecutive quarters boasting 480% growth in Q2 2023, with over 407K scenes rendered compared to ~70K in the prior quarter. April saw its largest month, accounting for 44% of scenes rendered in the quarter. As a result, creators paid over 815,000 RNDR to Node operators in Q2 2023, growing by 121% QoQ.
Source: Dune @lviswang
Over the last year, on average, there have been about 291 Nodes online. In Q2 2023, were 313 Nodes on average available to render services, 35% more than in Q1. Although the number of new Nodes connecting to the Network grew by 200% in Q2, with 9 added, the last four quarters saw single-digit Nodes added. However, Node operators online grew by 46% in May, corresponding with the monthly record increase in scenes rendered and RNDR distributed.
Considering there were 325 new nodes added to the Network in May 2021, representing 55% of all new nodes added, it seems plausible this cohort could represent a foundational base of Node operators responsible for performing a vast majority of rendering services.
The Render Network has several key initiatives and improvements planned for the remainder of 2023. The recent establishment of the Render Network Foundation and its Grants Cohort will help build out the community and aid in the expansion of the Network. Some notable additions and improvements in the pipeline are as follows:
Stable Diffusion: The Network will support AI training and AI rendering workloads, which are expected to be a significant source of compute growth in 2023.
Cinema4D / Redshift Integration: This integration will bring multi-render support for creators, allowing seamless workflows across the 3D ecosystem and enabling users to leverage the Network for different rendering pipelines.
Octane X multi-backend and thin client rendering: This development will enable users to create works on M1 and M2 iPads and MacOS, and render them onto thousands of decentralized GPU nodes. This is particularly important for virtual production, Augmented Reality, and mobile rendering.
LightField Rendering (in Beta): This addition expands the Network's intensive rendering services to include precomputing lightfields, NeRFs, and other holographic rendering jobs, potentially increasing per-job rendering demand significantly.
Render has emerged as a leading decentralized cloud GPU rendering platform, leveraging blockchain technology to connect a distributed network of consumer-grade GPUs and offer affordable on-demand render power to creators. The Network has experienced impressive growth despite a market facing regulatory pressures, signaling that non-crypto natives could be taking advantage of the platform. The 40-80% cost savings compared to centralized cloud rendering services and faster turnaround times through parallelized GPU processing are strong incentives to attract new creators.
As the demand for interactive 3D and augmented reality experiences grows, Render is a trusted solution, bridging the gap between creators seeking high-quality renderings and suppliers willing to share their idle GPU resources. Proving the viability of blockchain-coordinated decentralized rendering at scale, Render's unique value proposition and robust roadmap position it favorably for further adoption within the GPU compute sector.