fff

FFF White Paper

Version: 1.0.2

2021/10/25

A blueprint for fully distributed cloud computing based on blockchain

Building a blockchain-based distributed cloud platform

FFF aims to provide a distributed application that runs on the blockchain with the scalable, secure, and easy access to services, datasets, and computational resources needed to execute massively parallel and distributed applications at one-tenth the cost of traditional supercomputers through the main chain FFF.NETWORK network. FFF relies on a mature, reliable, open source public chain. It implements all the features needed for desktop networking software: fault tolerance, multi-application, multi-user, hybrid public/private infrastructure, deployment of virtual disk images, data management, security and auditing, and more. FFF is developing a new proof-of-contribution protocol that will allow consensus outside the blockchain. Through proof of contribution, external resource providers will certify the use of their resources directly on the blockchain. Ultimately FFF will deploy a scalable, high-performance, secure and manageable blockchain that will facilitate new forms of distributed management. Leaders in the high performance computing, big data, and cloud computing industries will soon join in this transformation in the future. We believe that big data and high performance computing applications, high value data sets and computing resources will profit from a highly transparent, resilient and secure blockchain in the future of decentralized infrastructure and marketplace networks. FFF is the key platform driving the future

Challenges for blockchain computing

Similar to others, blockchains offer a new way to run distributed applications, collectively referred to as DApps in the following, allowing programmers to write smart contracts, i.e. code that executes on a blockchain virtual machine whenever some transaction is triggered. This means a new revolution in building and executing in areas such as investment, finance, crowdfunding, the Internet of Things, insurance, prediction markets, gaming, distributed data processing, etc., and in essence, a sea change in the way existing businesses traditionally compute. While the outlook is bright, the average blockchain has very limited computing power to run distributed applications: a few KB of storage, relatively inefficient virtual machines, and a protocol with fairly high access latency. Of course blockchain technology will overcome these problems in the future, but it will also need to provide additional resources to meet more demanding applications

Existing cloud computing infrastructure cannot meet the needs of DApps, which require a fully decentralized infrastructure to operate. At the same time, there is a growing demand for computing power to run large applications and process large volumes of data in the scientific field and across industries. The computational power to run Big Data applications is typically provided by high performance computing and cloud infrastructure. However, cloud infrastructure and high-performance computing are overly complex and costly to operate. Innovative small businesses often do not have the industry baseline and expertise to acquire and operate HPC platforms, and cloud providers like AWS remain very expensive for demanding applications such as GPU rendering. In addition, data processing centers often consume a lot of energy to run servers and cooling systems, which can be very costly and have a negative impact on the environment. Therefore, we need a new form of distributed cloud infrastructure to enable blockchain computing with lower operational costs

Technology Overview

FFF is dedicated to supporting the operation of emerging blockchain-based distributed applications and enabling low-cost, high-yield, high-performance computing by building distributed cloud infrastructures.

Blockchain-based distributed cloud computing infrastructure will allow on-demand, secure, and low-cost access to the most competitive computing infrastructure. dApps will automatically retrieve, find, provision, use, and release all required computing resources, such as applications, data, and servers, through FFF

As a result, FFF has built a new ecosystem in which companies providing storage space, computing factories, data providers, web hosts, and SaaS applications can do business with each other (shown). Distributed cloud computing will open up new markets for the efficient use of existing computing infrastructure. Servers can be moved out of data centers in order to reduce the energy required to run servers and air conditioning systems. By simplifying access to servers, distributed cloud computing significantly reduces heat loss in data centers while making it easier for data providers and consumers to access the computing resources they need

Introduction to core values

FFF addresses the needs of all distributed computing ecosystem participants: DApps providers can perform off-chain computing as needed. Application providers can fundamentally reduce the operational costs of distributed applications by using a secure, robust and reliable FFF infrastructure Data providers can expand their potential market size by integrating their services with the FFF cloud platform Server providers can leverage underutilized computing resources and improve the return on investment in existing infrastructure by seeking higher margins on servers provided to the FFF distributed cloud.

Existing infrastructure providers

FFF can enable home users to quickly monetize their existing computing resources, as well as provide additional monetization opportunities for existing infrastructure providers, such as miners. With the ability to share idle CPU compute cycles, use servers as compensation, and use resources from different providers without complex resource management, FFF users can easily achieve more profitability

Distributed application or cloud infrastructure users

FFF's cost of providing computing resources for distributed applications is much lower than the cost of computing resources for traditional blockchains, which will help them bring more value to their customers. The good reputation of reliable resource providers will bring additional benefits and provide the required level of quality control for computing resource integration services. Supporting different resource providers and having a full understanding of each provider's portion of the contribution can also help improve transparency across the platform.

Key Technical Advantages

The development of robust distributed computing marketplace networks requires breaking several technologies: development of proof-of-contribution protocols to provide provable consensus, traceability, and credit; development of smart contracts to enable the acquisition and provisioning of computing resources, and automatic post-execution payments; development of technologies to allow distributed applications to access off-chain computing resources on demand; development of technologies to advertise and use computing resources on marketplace networks; development of solid, systematic and qualitative services. Complete validation by tracking resource usage and providing Service Level Agreements (SLAs) to customers, vendors to enable support for Service Level Agreements for resource utilization. FFF's upcoming solution will enable it to quickly grow into the world's premier distributed computing market, with its proof-of-concept system now well on line.

FFF's upcoming solution will enable it to quickly evolve into the world's premier distribution

formal computing market, it now has a proof-of-concept system up and running smoothly.

distributed application market

FFF will provide customers with the basic building blocks that depend on running their applications.

First, we will focus on the needs of DApps providers to enable them to perform off-chain computations. Different lines of

Entrepreneurs in the industry have shown great interest in these highly promising applications. Based on their characteristics, these should

The usage process is decentralized and hence ideal for FFF applications. The blockchain application market is growing at a CAGR of

It is estimated to reach 61.5% and is expected to generate $2.3 billion in benefits by 2021. Current major blockchain technologies

Technology vendors include Microsoft (US), IBM (US), Deloitte (US) and Ripple (US).

We then allow traditional cloud services to operate in a new way in a distributed cloud computing platform, enhanced by FFF side chaining. More specifically, FFF will provide compute and service management services through configurable resource management services, allowing the resources therein to be used for specific purposes and perform related tasks. The IaaS market has grown steadily in recent years. As a result, FFF is entering a promising growth market. As shown in

The latest statistics show that it will likely reach 700% by 2026, compared to its current market value of $12 billion, as indicated.

The FFF team will identify the best areas for its distributed cloud market network to compete with existing cloud infrastructure providers and focus on building the foundation platform. The existing technology base provides FFF with a first-mover advantage into the market, and FFF is now positioned as a cutting-edge computing provider capable of delivering the killer distributed applications of the future

Examples of commercial applications

To meet the needs of a specific distributed application, we elaborate on a business example of a financial trading framework, FFF, that requires high performance computing power to perform complex computational methods. FFF is a framework designed to help small investors improve their trading decisions through different services, such as obtaining a portfolio of stocks based on complex computational methods. Currently complex calculations need to be performed through high performance computing, a service that is typically only available to large financial institutions. By using the FFF distributed cloud computing platform as a virtual supercomputer, FFF provides a realistic, low-cost and secure service to improve the quality of trading decisions.

Improve the investment decisions of clients.

Each new service developed in FFF can be transacted directly on the blockchain, similar to a cloud system

in a software and services (SaaS) approach.FFF's customers use dedicated smart contracts, which will clearly

The functions and access rights of FFF on FFF.

The world continues to produce massive amounts of data and is a great source of business, technical and technological innovation. We have seen the emergence of blockchain-based data marketplaces, such as Ledgys, Inc. FFF is able to connect Kaiko as a data provider to FFF as an application provider, running FFF applications on a specific user-defined portfolio. Computing Resource Providers Any computing resource provider can provide relevant computing resources to the FFF blockchain. The current computing resources fall into the following categories: 1) traditional cloud service providers such as Amazon, AWS, Microsoft Azure; 2) decentralized cloud service providers with aggressive cutting-edge technologies such as Qarnot Computing and Stimergy in France; 3) blockchain mining companies that are constantly optimizing the efficiency of their resources, such as Genesis Mining, which have enormous computing power and can run high-performance computations in combination with other blockchains. In this example, FFF customers choose FFF as the application provider, Kaiko as the data provider, and Stimergy as the resource provider. Their applications, data and resources are represented as smart contracts deployed on the blockchain with their terms of use embedded, and the interaction model between them

Our competitors are mainly limited to the blockchain space, and in particular to these projects: off-chain computing, data storage, and computing resources. Some projects allow the execution of computations on untrusted resources, such as Enigma, Truebits, etc. These designs are interesting from a research point of view, but they often rely on solutions that severely limit their applicability, such as Enigma multi-party computation. FFF does not compete with existing blockchain-based online storage solutions (e.g., StorJ, Filecoin, Sia). Oraclize acts as an intermediary between the smart contract and the data source, ensuring that no data is incorrectly pushed in the smart contract. These technologies are complementary and synergies can be found on FFF. Gridcoin has created a cryptocurrency based on computational resources for Boinc-based volunteer projects, which is mainly limited to altruistic contributions to scientific projects. Golem aims to first assemble a virtual "supercomputer" to attract regular high performance computing users to its platform, while FFF focuses first on supporting DApps to build a distributed cloud computing platform that will eventually be competitive enough to attract cloud and high performance computing users. In addition, compared to existing and future competitors

FFF has the following advantages.

1. Significant reduction in time to market based on proven technology.

2. Many features for businesses.

3. Incentives based on proof of contribution for rapid network growth and optimal use of the platform.

4. Each version of the development roadmap has a corresponding revenue model

The shell uses underutilized computing resources at a tenth of the cost of traditional supercomputers to execute very large scale parallel applications. The following three features of shell computing technology make it a good platform for fully distributed cloud computing.

5. Resilience: if some nodes fail, the computation continues to run on other working nodes.

6. Efficiency: applications get the best performance even with a wide variety of compute nodes.

7. Easy to deploy: no specific configuration is required to use any node, even those at the edge of the network.

Allows the use of any type of computing resource to execute compute-intensive and data-intensive applications.

Over the last decade, we have also developed a number of technologies for distributed computing: MPICH-V parallel computing, BitDew large scale data management, SQS applications providing quality of service, the first implementation for Internet computing, and many more.

Our team has acquired unique expertise and experience in keeping the Desktop Grid running smoothly and applying this technology to various scientific communities, startups and innovative companies.

infrastructure

FFF relies on blockchains to coordinate access to computational resources in distributed applications. This approach will lead to a steady stream of innovations in traditional blockchain technology. Some of the more unique of these innovations are proof-of-contribution consensus protocols and blockchains for specialized domains. Proof of contribution Traditional blockchains such as mainstream blockchains rely on proof-of-work (PoW) mechanisms to ensure that token transactions occurring between blockchain participants are verified by the cryptographic challenges employed by a large number of nodes, while FFFs employ proof-of-contribution protocols where token transactions between participants are triggered by off-chain actions, such as providing datasets in real time, transferring files, performing computations to provide professional services, and other activities. Therefore, a new protocol is needed to prove that contributions have occurred accurately and that the corresponding token transactions can be performed on the blockchain. We call this consensus mechanism a proof-of-contribution mechanism. There are several platforms that use similar protocols such as ripple and IPFS, which allow consensus building between the blockchain and off-chain resources. In comparison, proof of contribution will be more general and allow for more behaviors to be verified. Verification of contributions requires particular attention, as some malicious users will forge contributions to get illegal payouts. To prevent similar situations, the FFF will rely on a reputation mechanism [Araujo], proof of interest [Peercoin], and a 7-14 day transaction fallback window with contribution authentication and spot-checking [Sarmenta]

are allowed to deploy their own smart contracts. In order to prevent potential security issues, FFF will follow a more conservative approach, which we call a dedicated domain-oriented blockchain. Dedicated domain oriented means that we will change the blockchain to meet the needs of distributed infrastructure management. There may be requirements to handle large numbers of transactions (large number of task submissions) or low latency (real-time communication/information). In these cases, sidechain technologies with specific features can help handle these situations. Currently, matching description languages have been detailed in several academic papers and are used in software. In this work, our team plans to design and adopt a simplified version of the well-known software ClassAd, which controls the CondorHTC distributed system, developed by the University of Wisconsin.

FFF Smart Contracts: Multi-Criteria Scheduling on the Blockchain

In a distributed system, the scheduling algorithm assigns a number of tasks to be executed on the relevant computational resources. The scheduling process

Sequences are an important part of distributed computing systems, and the performance of application execution depends largely on its

Effectiveness. A particular challenge for schedulers is to design multi-criteria scheduling, i.e., an algorithm with multiple

policies to select computational resources and schedule tasks.

FFF will propose a new algorithm for result validation using blockchain and smart contract functionality. The method allows users to verify results based on provable

We have a clear reputation and an established budget to choose the right partner from the market. We could build some third-party escrow mechanism in which settlement payments would be delayed until the results are certified. Such mechanisms could be combined with a reputation system stored on the blockchain and only for less trusted

nodes to run redundant computations, thus significantly reducing the resources and costs required for computation Verifiable file transfer

Business content distribution will be a key feature of a distributed application built on the FFF blockchain. Customers will use the blockchain's smart contracts to access high-value data sets (such as genetic or financial data). In addition, FFF will protect data recipients by ensuring that content providers can actually deliver files and that the files have been downloaded before users pay for them.FFF can also protect file providers from malicious downloaders who pretend that the file transfer was unsuccessful in order to refund the fees already paid.

1. There is a clear boundary between public/private access to resources. For example, private resources can only be accessed by proprietary ownership companies or restricted trusted partners. Public resources, on the other hand, can be handled through any host.

2. The data must be as important as the calculations.

3. The allocation of computing resources requires clear judgements about cost and performance.

Designing systems that provide these three features on a fully decentralized infrastructure is a challenge. No

We already have a wealth of research and practical experience in these three areas.

This release will locate to data providers and allow them to join the marketplace network. In addition, more companies will

With this version, the marketplace network will allow different resources to run their applications and DApps through FFF.

Multiple direct connections between source providers coexist.

This release increases FFF's revenue stream by allowing new revenue models to be built for DApps, and this

A higher level of trust and quality of service is required. These applications would benefit from the expertise of selected resource providers.

environment, and its excellent service-level agreement (SLA)-specific quality of service (QoS) capabilities.

field(R) and the financial markets where we already have extensive experience. With respect to server vendors, the focus will be on working with

Partnerships with infrastructure providers, including small cloud providers who rent out servers, individuals and Greek companies.

Miners looking to rent self-built miners, server clusters and home servers

High Performance Computing (Enterprise Edition)

This version allows miners to join the FFF marketplace network as server providers and provides customers with true

Positive supercomputing power.

Currently, mining farms monetize their GPU resources through blockchain consensus computing. Through FFF, this Some vendors will gain access to a new market for blockchain-based high-performance computing applications. Through this, vendors will

There is an opportunity to better leverage its own powerful computing capabilities and expand related businesses.

In addition to miners, FFF high-performance computing version extends application provisioning pool to GPU-based applications

programs. These applications can be used for deep learning, 3D rendering, computational fluid dynamics, molecular dynamics

learning, finance, and many more areas. We will focus on deep learning applications because of their amazingly fast growth long usage rates, and the related players are keen to use GPU cloud computing

Research Edition

The goal of this release is a new use of FFF that goes beyond the distributed cloud. This will be a big leap forward in blockchain computing, as DApps are fully autonomous applications that can access resources, data and applications directly from the blockchain in a completely decentralized manner. For this reason, it is necessary to integrate software and protocols such as swarm, uport, etc., that are available now or may emerge during the course of the project. Combined with the full development of proof-of-contribution, this will open up new frontiers in serverless services, hosted directly on the blockchain. With the distributed cloud, we will open up the market network to new applications deployed exclusively on FFF: IoT, fog/edge computing, smart cities. For example, a recent study showed that (AT&T, Verizon, Orange ......) could cut infrastructure costs in half by distributing small data centers along their networks. the FFF will use this approach to build blocks. More on the technology can be viewed via the white paper open source community

bibliography

[1] people on nxtforum.org (2014) DAG, a generalized

blockchain.https://nxtforum.org/

[2] Yonatan Sompolinsky, Aviv Zohar (2013) Accelerating Bitcoin's Transaction

Processing. Fast Money Grows on Trees, Not Chains.

https://eprint.iacr.org/2013/881.pdf

[3] Yoad Lewenberg, Yonatan Sompolinsky, Aviv Zohar (2015) Breaking free from

chains: "Secure chainless" protocols for Bitcoin.

https://dl.dropboxusercontent.com/u/7426164/Bitcoin/ Bitcoin meetup Chainless.pptx

[4] Sheldon M. Ross (2012) Introduction to Probability Models. 10th ed.

[5] Amir Dembo, Ofer Zeitouni (2010) Large Deviations Techniques and

Applications. Springer.