The Internet Computer Protocol (ICP) is redefining blockchain performance with its capability to routinely handle over 2 million Ethereum-equivalent transactions per second (TX/s). This metric is remarkable as it normalises the complexity of transactions across blockchains, allowing for a clearer comparison of the computational power and efficiency between ICP and Ethereum (ETH). In addition to scaling to unprecedented levels, ICP can run intricate computations in a fraction of the time and cost required by ETH, supporting its goal to serve as a “world computer” for truly decentralised applications and services.
Blockchain networks typically boast transaction speeds as a measure of performance. However, these figures only reflect true network capability if transactions perform similar tasks; otherwise, they’re not an accurate comparison. ICP’s architecture stands out, as its transactions are generally more complex and computationally richer than those on traditional blockchains like Ethereum. Instead of just counting raw transactions, the metric of ETH-equivalent transactions provides a way to measure ICP’s high work intensity per transaction and its exceptional execution speed.
Both Ethereum and ICP support general-purpose smart contracts that enable diverse applications. At the core, smart contracts are translated into lower-level instructions for execution: Ethereum uses the Ethereum Virtual Machine (EVM), and ICP uses a WebAssembly-compatible runtime (Wasm). Both sets of instructions contain arithmetic operations (e.g., addition, multiplication, division), but they also support specific instructions tailored to smart contract use, such as memory access. Memory access operations are typically more resource-intensive, meaning they consume more computational power, which in turn influences the transaction costs. Ethereum’s gas model for EVM operations and ICP’s cycles model for Wasm instructions each track the cost and resource consumption of these operations, but the cost efficiency and speed between the two differ significantly.
Understanding the workload capacity of ETH and ICP in terms of transactions is complex. Ethereum’s throughput, for example, depends heavily on the daily gas used in the network, while ICP’s throughput is measured in instructions executed per second. Ethereum processed around 108.3 billion gas units on a single day (Sept 16, 2024), which represents the aggregate computational work done across the network. For simplicity, if we assume each instruction equals one gas unit, Ethereum’s network performs under 109 billion instructions daily. In comparison, ICP is capable of executing over 110 billion instructions every second, meaning ICP could theoretically handle Ethereum’s entire daily computational workload in less than a second. This reveals ICP’s significant computational advantage, particularly in high-demand settings.
Ethereum’s daily transactions offer another perspective on the networks’ relative computational capacities. With 1.102 million transactions on Sept 16, 2024, and an average of approximately 98,000 instructions per transaction, Ethereum handles substantial computational tasks daily. In contrast, ICP performed an average of more than 12.9 million instructions per transaction on the same day, indicating that each ICP transaction carries out roughly 130 times the computational work of an average Ethereum transaction. These figures suggest that ICP is not only faster but also significantly more capable of executing complex, computation-heavy tasks in a single transaction.
Network efficiency, particularly when scaling, is a critical area where ICP demonstrates considerable strength. Ethereum’s replication factor—a measure of how often data or transactions are replicated for security—is approximately 1.5 million, while ICP’s replication factor averages around 13. Although replication ensures data integrity and security, Ethereum’s high replication factor consumes considerable computational resources. ICP’s replication is notably more efficient, leading to a level of efficiency that outpaces Ethereum by an estimated 15.5 million times, further underscoring ICP’s scalability and resource-conscious approach.
One practical way to evaluate the two networks’ performance and cost efficiency is through EdDSA signature verification. Signature verification is a routine task in blockchain operations, such as validating transactions. On Ethereum, EdDSA verification costs around 500,000 gas, which at a rate of 40 Gwei per gas unit and an ETH price of approximately USD 1800 translates to a cost of USD 36 per verification. In contrast, ICP’s EdDSA verification through a smart contract written in Rust costs just 4,211,120 cycles. Converting this into USD with the ICP cycle rate set at USD 1.3476 per XDR, the cost per verification on ICP is approximately USD 0.00000567490. This cost disparity illustrates that ICP is over 6 million times more cost-efficient than Ethereum for standard computational tasks like EdDSA signature verification.
ICP’s speed, low-cost transactions, and efficiency are part of its larger mission to become a general-purpose world computer, offering direct web service capabilities without the need for traditional web2 cloud providers. This positions ICP as a unique player in the blockchain space, capable of serving fully decentralised applications directly to user browsers. Such an infrastructure enables developers to build applications that deliver a web-like experience without relying on centralised entities for storage or computation, which is often the case in the traditional internet infrastructure dominated by companies like AWS or Google Cloud.
This mission towards decentralisation is facilitated by ICP’s Wasm-compatible execution layer, which supports high-speed smart contracts that can handle the complexity of web applications. The Wasm instructions are specifically designed for efficient performance on modern hardware, making it easier to scale applications as demand grows. This stands in contrast to Ethereum’s EVM, which, although robust, was initially designed for simpler blockchain applications and is less optimised for high-speed execution.
The differences between Ethereum’s and ICP’s execution models highlight a broader evolution within the blockchain space. Ethereum, for instance, has been moving toward a rollup-centric roadmap to improve scalability. Rollups are Layer 2 solutions that process transactions off-chain before confirming them on the Ethereum mainnet. While rollups significantly improve Ethereum’s scalability, they add layers of complexity and costs that users and developers must navigate. ICP, by comparison, doesn’t require these additional layers due to its efficient Wasm-based execution and low replication requirements, allowing it to process transactions at scale with less operational overhead.
The demand for scalable, interoperable blockchain solutions has driven ICP’s design, resulting in a network capable of not only handling high transaction volumes but also executing intricate, computation-heavy tasks. Traditional blockchains, including Ethereum, have primarily focused on transaction throughput, but ICP aims to provide both throughput and computational power. This approach caters to developers and users seeking more than simple token transfers, positioning ICP as a comprehensive solution for complex dApps that demand both speed and computational intensity.
In the competitive landscape of blockchain networks, ICP’s innovative approach to transaction processing, cost efficiency, and decentralised infrastructure offers a compelling alternative. By achieving transaction speeds and cost efficiency at levels far beyond Ethereum, ICP demonstrates the scalability required to support large-scale, real-world applications while maintaining decentralisation. With the advent of ETH-equivalent transactions, ICP sets a new benchmark in the blockchain world, underscoring the importance of computational capability in evaluating a network’s performance.
As decentralised finance (DeFi), Web3, and other blockchain applications continue to grow, ICP’s advantages in computation and cost will likely draw increased interest from developers and users alike. While Ethereum remains a dominant force, particularly in DeFi, ICP’s architecture appeals to those seeking advanced functionality, speed, and reduced reliance on web2 providers. By providing a robust, low-cost alternative to traditional web services, ICP stands poised to drive a new wave of decentralised innovation.
ICP’s ETH-equivalent transactions per second metric highlights its formidable performance capabilities compared to Ethereum, especially in handling complex, computation-intensive tasks. Its efficiency in terms of both speed and cost presents a compelling case for the future of blockchain, where computational power and scalability will increasingly define success. As blockchain technology evolves, ICP’s focus on delivering high-throughput, low-cost transactions across a decentralised network positions it as a key player in the pursuit of a truly open, web-based world computer.
