Key takeaways
- Web3 services alone will require almost 90 billion zero-knowledge proofs to be performed in the year 2030
- This would deliver 83 thousand transactions per second
- We project the market for generating zero-knowledge proofs will reach $10 billion in 2030
Zero-knowledge proofs are cryptographic tools that provide key building blocks to drive significant expansion in Web3 services and can add value to traditional industries including finance, health care, data storage, gaming, government applications and more. An application using zero-knowledge proof mathematics would enable the proving of certain information without passing along the specific details of that information, enhancing privacy and enabling robust solutions protecting user identity and data content. Another use case allows for the outsourcing of computing power to a third party, while confirming that the calculations were completed correctly.
Quantifying the potential size of this nascent market in the near and medium term is a key step in ensuring that the capital providers fully appreciate the opportunity to invest in architectures and hardware – the “picks and shovels” –to support the generation of the complex zero-knowledge proof mathematics, enabling the entire ecosystem.
We created a framework to analyze the market size and the hardware requirements for scaling Zero-knowledge proof generation, starting from the most developed use case of scaling the Ethereum blockchain, and then estimated the market size by projecting the proportion ZKPs to be settled on other non-Ethereum L1 chains. We also developed a market clearing price for a decentralizing proving entity to profitably generate these proofs, which is necessary for the long-term viability of the ecosystem.
Based on our work, the zero-knowledge proving market is projected to reach $75 million in revenue in 2024, and has the potential to exceed $10 billion in revenue by the year 2030. Web3 applications alone are expected to require almost 90 billion zK proofs in 2030, and we expect the average market clearing price per proof to fall from $0.21 in 2024 to $0.12 by 2030.
Prediction: 87 billion zero-knowledge proofs by 2030
Ethereum L1 has been designed for maximum decentralization and security at the expense of scaling. As a result, the execution environment where users will experience cheap, fast transactions would consist of Layer 2 solutions – namely Zero-knowledge proof Rollups that are compatible with Ethereum - many of which are being launched on mainnet in 2023. In this thought piece, we derive a potential market size for zK proof by assuming that Ethereum is used significantly by zkEVMs and zKVMs, and then we looked at the potential for other L1 chains to settle transactions.
Because zkEVMs are very new (e.g. two major candidates just launched mainnet versions in March 2023), we have yet to see mass adoption and have limited data. We will therefore highlight some important assumptions for our analysis:
- Ethereum as the cornerstone of the analysis: The analysis assumes that many L2 rollups - blockchain scaling protocols - will choose the Ethereum network to settle zK proofs, given the blockchain’s security, transparency, and immutability.
- zK transactions per Ethereum block: The target gas limit for each Ethereum block is 15 million Gwei - Ethereum’s gas unit - with each zK-proof transaction using about 500,000 Gwei. Each Ethereum block currently can support an average of 30 ZKP transactions.
- zK proofs generated per year on Ethereum: At a potential of 30 proofs per zK block, 7,200 blocks per day could mean a maximum potential of 216,000 ZKPs settled on Ethereum L1 per day, or 78.8mm ZKPs per year. As the industry ramps up, we see 25% of the Ethereum block space being used by zK proofs in 2024, growing to 90% by the year 2030. Layer-2 scaling protocols will be able to acquire space on the Ethereum blockchain because they can divide the gas fee among many zK proofs and transactions.
- Scaling zK proofs per block: Each zK transaction in an Ethereum block has the potential over time to be a rollup of multiple ZKPs. Additionally, the number of proofs per zK can potentially be expanded as new approaches and protocols are developed (some of which may result from adjustments to the Ethereum protocol itself). The analysis assumes that initially 15 ZKPs would be settled per zK block settled at L1, growing to 810 ZKPs by 2027.
- Other non-Ethereum L1 chains: The analysis assumes that many L2 rollups - blockchain scaling protocols - will choose the Ethereum network to record zK proofs, however other L1 zK native chains are being launched as well. This model assumes that 90% of zK proving will take place on Ethereum in 2024 for the reasons discussed above, dropping to 66% by 2030.
The analysis projects that 87 billion zero-knowledge proofs will be conducted to support Web3 applications in 2030, which would encompass 2.6 trillion transactions – about 83 thousand transactions per second.
For comparison, the Visa network has processed up to 65,000 transactions per second, and the Alipay network has processed over 250,000 transactions per second[1]. When you consider that this projection does not include off-chain zK proofs for commercial applications in finance, health care, etc., this can be seen as a conservative projection for the market size for zK proving.
Prediction: $0.12 market price per zK proof in 2030
To complete the calculation of the financial size of the market, we need to determine the market clearing price per zK proof calculated[2]. This is the price per proof whereby a company or organization would be able to cover operating costs, amortization of investments, the cost of capital, and a reasonable level of profit to the stakeholders. At a lower price, there will not be enough investment in supply to meet the market needs, and higher prices would be pushed down by competition. In the short term, price and supply would fluctuate, but over time this equilibrium price clearing mechanism provides a strong framework for calculating expected revenue, especially as the size of the market reaches significant scale.
We based our analysis on a standard “hardware stack” that could handle a certain number of Zero-knowledge proofs of a projected median difficulty. We constructed a pro-forma “income statement” for a ZKP operation, layering in projected operational parameters and incorporating operating expenses, investments in equipment and infrastructure, and other costs. These assumptions were based on our experience as infrastructure providers, and were discussed with key players in the industry.
- Complexity of zK proofs: Many kinds of zK proofs exist, with more complex proofs having more advanced hardware and computing requirements to handle the zero-knowledge proof mathematics. We recognize that there will be a range of required hardware, from the latest generation of high-performance GPUs for complex ZKPs, to smaller configurations to handle simpler proofs. For this analysis, we assumed a core technology stack that could handle median difficulty ZKPs.
- Hardware and configuration costs: The model assumes a baseline required hardware stack based on 3 GPUs would be capable of processing average zK proofs. These GPUs would cost about $2,000 each, for a total of $6,000 for three units. (Using high-performance GPUs would significantly increase the cost of this tech stack.) The costs of other processing power, memory, hardware, and installation bring the total for hardware and configuration to an estimated $15,640.
- Additional infrastructure and installation costs: Experience indicates that the cost of necessary additional infrastructure and installation is about 50% of the cost to purchase and configure the hardware. This assumption tacks on $7,820 to the hardware and configuration costs, for an estimated total upfront investment of $23,640.
- Number of zK proofs per day: The expected capacity of this baseline tech stack would be an average of 3,000 zero-knowledge proofs per day, or 1,095,000 proofs per year.
- Capacity utilization: The assumed capacity utilization is 50%, or about 547,500 zero-knowledge proofs per year. This utilization rate is considered appropriate (or possibly a bit high) for highly specialized hardware configurations, which may experience significant fluctuations in processing demand and can be impacted by network issues, variability in protocol performance, etc.
- Hardware amortization period: The target payback period for the initial investment is two years. The model assumes that the riskiness of the business creates investor appetite for a fast payback. Amortizing the capital investment of $23,640 over two years puts a charge on the income statement of $11,730 per year.
- Unit economics analysis: An operating entity in this business should expect operating costs (power, bandwidth, rent, etc.) of 20% of revenue, headcount and administrative costs of 50%, hardware amortization of 10%, and a resulting profit of 20% of revenue. Based on this formula, the company would need to generate revenue of $117,300 per year to attract investment and create a sustainable business.
A zero-knowledge proving company earning revenue of $117,300 per year to complete 540,000 zero-knowledge proofs would receive a market clearing price of $0.22 per proof, which would support a profitable ecosystem of zK provers, even with operating costs and amortization expenses being considered.
Scaling a technology and its supporting ecosystem tends to decrease average market prices, although mitigating factors for zK proofs can maintain upward price pressure. Adoption zero-knowledge proof mathematics for increasingly complex applications would require significantly increased processing power, supporting a higher proof price. An expected significant increase in the functionality of zero-knowledge proofs would increase the demand for complex proofs using zK protocols. For our projection, we expect the price per proof to decline by 10% per year to $0.12 by the year 2030.
Prediction: $10 billion market for zero-knowledge proving by 2030
This analysis indicates that the zero-knowledge proof market is likely to generate $10.2 billion of revenue in 2030 - that’s 87 billion zero-knowledge proofs completed at $0.12 each. Key factors to be followed as this market develops include the emergence of the use of increasingly complex zero-knowledge proof mathematics, the ability to scale zero-knowledge proofs within block transactions, and especially the growth of zero-knowledge proving beyond the Ethereum blockchain. Predicting a zero-knowledge proof market worth more than $10 billion in 2030 is likely still conservative.
There are several issues to consider as this analysis adapted for a particular implementation or business instance, and will need to be addressed at the micro (protocol / business) and macro (Dapps / industry) level:
- There are many kinds of zK proofs and a host of different implementations. What are the implications for infrastructure based on the approach, business requirement, etc.?
- What are the limitations of current hardware to support zK proving, especially given the high cost and design restrictions of the latest generation of GPUs?
- What is the impact of variable / bursty demand on infrastructure demand / costs, and what will be the implications for network pricing and performance for end users?
- Initially many of the provers will be provided by the zK based chains themselves. How can / will the proving market evolve to enable efficient, decentralized provisioning of proofs.
We hope that this analysis will help support the efforts to invest in the growth of this exciting market, and are sure that many questions and issues will be raised. We are available to discuss our approach in further detail, and will appreciate getting your feedback and insights.
About Aligned.co -> Aligned’s proprietary integrated hardware and software approach brings exceptional speed and efficiency in processing massive data sets on demand. Built on our current fleet of 4,000 FPGAs, our solution delivers high performance, low-latency, and energy efficient compute. Today, the company is providing remote sealing as a service for Filecoin storage providers (generating the zK proofs as required by the protocol), and is expanding into generating ZKPs for other Web3 and non-Web3 customers.
