In January 2024, Andreas Gasmann, a developer at Acurast, hosted a workshop for the CTRL+Hack+ZK hackathon, the inaugural hackathon organized by Aleph Zero. Speaking expressly about the manifold problems associated with present cloud providers and the blockchain quadrilemma and its negative impact on distributed applications, he presented Acurast. This Layer-1 blockchain addresses these shortcomings with a novel decentralized and serverless approach, and this article will go into depth about its solutions and how it solves these problems.
What is Cloud computing?
Cloud computing is a technology that involves delivering various computing services over the internet. Users can access these services on demand, often paying for only the resources they consume. Examples of on-demand cloud services include "Infrastructure as a Service (IaaS)" offered by providers like Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP). Additionally, there are "Platform as a Service (PaaS)" platforms such as Heroku and Google App Engine, enabling users to build, deploy, and manage applications without dealing with the underlying infrastructure. "Software as a Service (SaaS)" applications like Salesforce, Microsoft 365, and Dropbox are also notable, being delivered over the internet for easy access and usage. Cloud storage services, including Google Cloud Storage, and Microsoft Azure Storage, provide scalable and on-demand storage solutions, exemplifying the flexibility and convenience inherent in cloud computing, thereby enabling collaborative work environments.
Contrasting with Decentralized, Serverless Cloud
The evolution from Web2 to Web3 signifies a pivotal shift. Web2-based cloud services rely on centralized entities, fostering concerns over data control and privacy. In contrast, Web3, embodied in decentralized, serverless cloud solutions, prioritizes user ownership and privacy, removing intermediaries. This transition from Web2 to Web3 encapsulates a broader move from centralized trust and control to a democratized, permissionless, and user-centric Internet experience. The serverless cloud, operating on a peer-to-peer network, ensures robust security and confidentiality through smart contracts, albeit facing challenges in scalability and managing decentralized resources.
Spotlight on Cloud Provider Hurdles
Cloud providers contend with 4 challenging aspects of Cloud computing:
Cost: Often expensive due to centralized infrastructure and maintenance costs. Cloud services are dominated by a few major players like Google, Microsoft, and others, creating a competitive landscape that can limit options for smaller businesses. This concentration can result in higher prices as these major companies have significant control over the market. Also, factoring in the cost of potential cloud outages which is a commonplace occurrence, a 2023 report from Parametrix Insurance - pioneers in monitoring and modeling of cloud providers and cloud-based services - concluded that a 24-hour outage of mission-critical services from AWS us-east-1 - the cloud region with the largest number of Fortune 500 companies relying on it - could cost $3.4 billion in direct revenue.
Transparency: The lack of transparency in pricing for cloud services is a common concern. Many traditional cloud providers often have complex pricing models with various factors such as data transfer, storage, and compute resources, making it challenging for users to understand the true cost implications of their usage. In contrast, there's a growing demand for more transparent pricing models in decentralized and serverless services. Clear, straightforward pricing structures allow users to better anticipate and manage costs, fostering trust and enabling more informed decision-making.
Centralization: The centralization of cloud services, particularly dominated by a few major U.S.-based providers, introduces concerns related to control, potential single points of failure, and legal jurisdiction. Reliance on a small number of providers may expose businesses to disruptions if a major provider experiences downtime. Moreover, the concentration of providers in a specific geographic location can subject users to the regulatory framework of that jurisdiction, potentially raising concerns about data privacy and security. This lack of geographic diversity limits options for businesses seeking alternatives that align with different legal and regulatory environments.
Confidentiality: Confidentiality concerns within cloud services involve worries about unauthorized data access or exposure. Users trust cloud providers to implement robust security measures, such as encryption, access controls, and secure authentication, to safeguard their sensitive information. The potential for insider threats within the cloud provider organization raises questions about who might access or misuse the data. Solutions like Acurast aim to address these concerns by providing a network with trust assumptions that align with blockchains' inherent permissionless ethos, offering a low cost of computation with an added layer of effectiveness and confidentiality.
Limitless computing with Acurast
Although blockchain development has been around for over a decade, one might more willingly volunteer to undertake the mythical task of slaying the Lernaean Hydra from Greek mythology, one of the twelve labors of Hercules, than to design a blockchain that is simultaneously secure, scalable, and decentralized while offering computational effectiveness - the ability to run complex decentralized computations at affordable costs - and an additional layer of confidentiality.
Acurast embraced innovation, tackling the quadrilemma with a decentralized and serverless strategy within its L1 Blockchain architecture network. This audacious move, coupled with a zero-trust security model, not only signifies a departure from conventional approaches but also promises robust benefits. Decentralization enhances resilience, minimizing the risk of a single point of failure and fostering scalability. The adoption of serverless architectures ensures automatic scaling, optimizing resource utilization and potentially lowering costs. This approach not only bolsters flexibility by offering a choice among decentralized platforms but also mitigates the risk of vendor lock-in. Improved accessibility is a natural outcome, as decentralized systems, by their nature, distribute services, resulting in reduced latency and heightened performance across diverse locations for end-users.
Andreas Gasmann, a developer at Acurast, spoke on the transformative impact of new chips, particularly ARM 64-based chips found in mobile phones like the Google Pixel, and highlighted that these chips, with not just CPUs but also GPUs and dedicated machine learning chips, offer substantial computing power. Despite being six times cheaper than traditional server racks, mobile phones exhibit faster CPU benchmarks, and when factoring in additional chips, they can achieve an 80% higher performance. The energy efficiency of these newer chips is noteworthy, requiring about 50 times less energy, resulting in improved overall performance for the same price.
A significant point emphasized was the ability to repurpose old cell phones for the Acurast network, allowing users to participate as processors, contributing compute power at a low cost, sometimes as low as $10. Anyone can join the network, with a simple three-minute onboarding process involving scanning a QR code on the website, making it accessible even to those without technical expertise. This approach has led to the formation of small device farms driving the Acurast backend. In comparison to a Google instance, running on the Acurast network is estimated to be around 30 times cheaper, coupled with the added benefit of confidential computing.
The Architecture of Acurast
Acurast's modular architecture, with distinct layers for consensus, execution, and application, exemplifies a strategic approach that not only champions a broader interoperability framework but also facilitates harmonious collaboration across Web3 and Web2 environments. The modular execution layer, a standout feature of Acurast, empowers the network to leverage secure hardware coprocessors, eliminating the need for trust in third parties and enhancing overall security. This design not only emphasizes universal interoperability but also prioritizes robustness and trustlessness at its core.
Essentially, the consensus layer of Acurast has 2 cores: the purpose-built Orchestrator and a reputation engine. The Orchestrator in Acurast orchestrates consumer jobs, employing an End-to-End Zero Trust Job Execution approach. A "job" refers to a task or computational process defined by a consumer within the Acurast system. It involves specifying details such as the destination for settling the job, selecting deployment templates, and choosing execution processors. These jobs can range from various computational tasks, and the Acurast system manages the entire lifecycle of a job, from definition and deployment to completion.
Acurast's job lifecycle comprises four essential components. First, in the job registration phase, consumers define task details, select deployment templates, and specify execution processors, while payments and integration preferences are settled. Subsequently, in the job acknowledgment stage, processors fetch job details and move through states like MATCHED and ASSIGNED based on fulfillment capabilities. The third phase involves job execution, where the actual computational process occurs within secure runtimes like Acurast Secure Hardware Runtime (ASHR). Finally, in the job fulfillment and reporting phase, the completed output is delivered to specified destinations, settling gas fees for cross-chain transactions. Processors report back to the Acurast Consensus Layer, entering the DONE state (which means job fulfillment is successful), and continuous reliability metrics are fed into the reputation engine to ensure the protocol's robustness.
The reputation engine in Acurast plays a critical role in maintaining the integrity of the system by ensuring accurate updates to the reputation scores of processors and incentivizing honest behavior. It acts as a mechanism to evaluate and track the performance of processors based on their successful job fulfillment or reported failures. By continuously feeding reliability metrics, particularly after job completion or failure, the reputation engine provides a dynamic way to gauge the trustworthiness of processors. This, in turn, creates a powerful incentive structure, encouraging processors to act honestly and efficiently in executing tasks within the Acurast ecosystem. The four components of the Orchestrator layer and reputation engine delineate a comprehensive and secure journey for computational tasks within the Acurast ecosystem.
Conclusion
Addressing the widely recognized challenges of centralization of trust, seamless interoperability, and confidentiality in the execution layer within Web2 and Web3, Acurast presents a disruptive solution to the US cloud monopoly. By leveraging mobile hardware, Acurast democratically decentralizes the cloud, offering everyone the opportunity to participate in this decentralized ecosystem using their mobile phones. This approach grants developers permissionless access to trustless, affordable, and confidential computing resources. As a Layer-1 blockchain, Acurast introduces a novel decentralized and serverless architecture, transforming it into a decentralized and serverless cloud. This modular design not only mitigates the identified shortcomings but also enables seamless, native settlements across ecosystems, ultimately enhancing the efficiency of computation. Acurast emerges as a stalwart vine, reshaping the landscape of decentralized cloud computing.
Comments