DecentraCompute vs Cloud Platform
Proprietary cloud platforms have clear economic incentives to create closed software service ecosystems, making it extremely difficult to switch platforms, effectively locking customers into the platform. If customers need multiple cloud services, it is easier to purchase them from the same provider because the integration costs between different providers are usually high. In addition, providers capture customers by setting pricing barriers for data: inbound bandwidth is cheap, but outbound bandwidth is expensive, thus reinforcing the lock-in effect.
Once customers are captured, their businesses are vulnerable to platform risks because they rely on specific providers. When businesses expand and utilize more cloud resources, drastic price fluctuations can put significant pressure on business continuity. Furthermore, many companies realize too late that switching code repositories to another cloud provider or private infrastructure requires a significant amount of time and resource investment.
The DecentraCompute protocol liberates developers from the constraints of proprietary tools and allows applications to switch providers freely. DecentraCompute enables developers to use any connected provider and easily switch between them. Additionally, the DecentraCompute protocol enforces cryptographic proofs for the execution of any client code, thereby enhancing security and resilience.
Unlike traditional clouds, DecentraCompute provides infrastructure managed by Web3 native organizations such as DAOs. Digital organizations can make payments using multi-signature wallets, update their code repositories through collective voting, and invite community members to provide infrastructure for their projects through the DecentraCompute network.
DecentraCompute vs Blockchain Application Platforms
Blockchain application platforms are based on transactional replicated ledgers and require consensus algorithms to update the ledger. This design is useful for "digital value" use cases such as cryptocurrencies, decentralized finance, NFTs, or DAOs because it can prevent malicious updates to the ledger and protect against tampering. However, the consensus model is suitable for relatively simple, deterministic computations and handling limited data. Additionally, blockchain introduces a pricing model for each transaction, and users must pay gas fees for each on-chain operation.
DecentraCompute takes a different approach to decentralized computing, where applications are not dependent on distributed ledgers. Applications are hosted off-chain, similar to centralized cloud platforms, but because the protocol is open, applications can be hosted by multiple providers, and developers can dynamically switch providers. This architecture significantly reduces the importance of any specific provider and drives down prices. Furthermore, DecentraCompute adds computational verifiability by increasing the execution proofs. Each provider in the DecentraCompute network must submit cryptographic proofs to demonstrate that they are correctly serving the application. Failure to submit proofs results in a reduction of their stake and cessation of payments.
Additionally, DecentraCompute takes into account the complexity of computation, similar to gas in blockchain, but by default, the developers deploying the application are responsible for paying the hosting fees, not the end-users. If developers choose, they can also transfer the responsibility of paying hosting fees to multi-signature wallets, DAOs, or end-users.
Similar to decentralized storage protocols like Filecoin, which store data off-chain but use blockchain to track, verify, and compensate storage, computation on DecentraCompute is performed on off-chain network nodes, while proofs and payments are submitted and verified on-chain.
DecentraCompute vs Rollups (Optimistic and Zero-Knowledge)
Rollups provide a scaling mechanism for blockchains by creating additional block space and transaction processing capacity outside of L1. To be compatible with wallets and L1 applications, Rollups also use blocks and transactions and adopt the same execution model (such as EVM). Rollups generate proofs (Merkle or zero-knowledge) for their chain state and verify these proofs on the relevant L1 to ensure the validity of the Rollup at a specific point in time.
DecentraCompute has a different execution and verification model because it does not put execution into blocks and transactions but runs applications similar to traditional clouds, with the requirement that providers submit execution proofs on-chain. This architecture makes it suitable for distributed or non-deterministic computations and scenarios that are unnatural for the "pay-per-transaction execution" model. Examples include backend APIs, data processing, indexing, message routing, multi-party computation, IoT, etc. These are not suitable for on-chain or Rollups.
DecentraCompute Comparison Table
Scalability Censorship Resistance Verifiability Payment Model Cost
Blockchain Low Yes Yes Pay per transaction $$$$
Rollups Medium Yes* Yes Pay per transaction $$$
Other decentralized computing Medium Yes** No** Pay per transaction
DecentraCompute High Yes Yes Pay per resource $
Rollups may face potential censorship if operated by a single node or a small number of nodes.
** Decentralized computing implementations vary, but they generally do not provide strong cryptographic proofs of execution correctness.
*** Cloud computing bills computation, memory, and data bandwidth directly or using composite metrics.