Article

Authors:
Article

# The Evolution of Bitcoin Hardware

If you want to read the PDF, try requesting it from the authors.

## Abstract

Since its deployment in 2009, Bitcoin has achieved remarkable success and spawned hundreds of other cryptocurrencies. The author traces the evolution of the hardware underlying the system, from early GPU-based homebrew machines to today’s datacenters powered by application-specific integrated circuits. These ASIC clouds provide a glimpse into planet-scale computing’s future.

## No full-text available

... Usually has partnership like business to business, every node has an authority to choose in advance such as Hyperledger). [6][7][8][9] Some BC types are permissioned (contain semi-trusted members, all BC is decentralized (no need of trusted-third party), transparent (data is recorded by every node in BC), open-source (people can create their own applications), autonomy (anyone can transfer and update data safely), immutable (once the data is written it cannot be changed, unless someone gets 51% control on the node), anonymity (only person's BC address is needed, its real identity is kept anonymous) [11]. Table 1 presents different Blockchain platforms and compares these on the basis of different parameters. ...
... Users interact with the BC via a pair of private/public keys [46].They use their private key to sign their own transactions, and they are addressable on the network via their public key. Every signed transaction is broadcasted by a user's node to its one-hop peers [8]. The intermediate peers assure the received transaction is valid before broadcasting it over whole the network, the invalid transactions are rejected by the peers. ...
... The government should make analogous regulations for this technology, organizations should be ready to grasp BCTs, averting it brings too much influence to recent systems. The BCT has the possibility to renovate banking structure, speed reimbursements, and modernize stock exchanges while providing the transparency needed in this modern era of high-tech if we can secure this technology more [8]. Recent research endeavors have applied BC in Cryptocurrencies, Cloud distributed environments, e-health, Finance Smart Contracts and even in Databases. ...
Article
Full-text available
Similar to decentralized communication systems, a new technology called Blockchain (BC) has the potential to store and manage data in a decentralized manner. By removing the role of third party, all member in the chain has equal access to data. The concept BCT (Blockchain Technology) is not just limited to the cryptocurrencies, but it has been implemented in other areas like e-health, voting, finance, education, smart contract and even in Databases. This paper discusses various Blockchain applications and platforms and then compares these platforms on basis of different parameters. Despite of the advantages, Blockchain faces a significant issue of privacy. This paper examines various security related issues and challenges and presents an account of known possible attacks.
... Nonce generation in PoW is fast. An INTEL XEON X5660 2.8 GHZ XMR-STAK-CPU has a hash rate (or nonce generation rate) of about 450hash/s, which is considered a weak system for mining [30]. [32]. ...
... The introduction of the GPUs and FPGAs brought about parallelisation and greater optimisation lacking with the CPU hardware. FPGA, in particular, was more efficient as it could replicate multiple SHA256 hash functions and unrolled them [30]. ...
Article
Proof-of-Work (PoW) algorithm is a popular blockchain algorithm employed in many blockchain applications such as Bitcoin. Cryptographic hashing is the foundation of the PoW algorithm and blockchain technology in general. Unfortunately, the use of hashing in PoW has led to huge computational requirements. Researchers and industrialists are aware of the immense energy consumed by the PoW algorithm in blockchain-based cryptocurrencies. For instance, Bitcoin currently consumes above 110 TWh of electricity annually. This vast amount of energy is used to calculate non-valuable cryptographic hashes, which eventually becomes a waste when the right nonce value is found. Bioinformatic researchers depend on molecular docking simulation, which is effective but requires heavy computing resources. This paper proposes a solution to the above issues by taking advantage of the wasted computing power harnessed by the PoW algorithm and subsequently channelling the computing resources towards molecular docking simulations for drug discovery. With the new proposed framework, molecular docking is introduced into PoW algorithm where energy from a PoW system like bitcoin can be used to help researchers crunch bioinformatics data for computer aided drug discovery. A target protein receptor and ligands are introduced at the beginning of a new mining period, rather than computing random computer bits to discover a nonce, mining nodes will perform docking simulations with the receptor and ligands to generate docked conformations for drug creation research. This research does not seek to reduce the computational energy of the PoW rather utilise this energy for valuable drug creation process. Keywords— Blockchain, Bitcoin, PoW, Ligand, Molecular docking, RMSD, Score
... However, researchers encounter roadblocks when intending to scale up blockchain by orders of magnitude, i.e., the resource efficiency still remains in situ. Simultaneously, Moore's law-based hardware evolution, from central processing unit (CPU) to graphics processing units (GPU), field-programmable gate array (FPGA), and applicationspecific integrated circuit (ASIC), seems to mainly intensify the energy consumption rather than providing scalability optimizations [13]. Effectively scaling up blockchain, without affecting its properties of decentralization and security, constructs a "trilemma" and poses daunting challenges for researchers. ...
... Naturally, this strategy is targeted to PoW-based blockchain. Taking Bitcoin as an example, during 2009-2019, the equipment of Bitcoin mining experienced multiple rounds of revolution [13], [42]. Initially, Bitcoin miners used CPUs since they are widely adopted and general-purpose processors. ...
Article
Full-text available
As an emerging technology, blockchain has achieved great success in numerous application scenarios, from intelligent healthcare to smart cities. However, a long-standing bottleneck hindering its further development is the massive resource consumption attributed to the distributed storage and consensus mechanisms. This makes blockchain suffer from insufficient performance and poor scalability. In this article, we evaluate numerous representative blockchain proposals in the past decade and demonstrate that the potential of the widely adopted consensus-based scaling is seriously limited, especially in the current era when Moore’s law-based hardware evolution is about to end. We achieve this by developing an open-source benchmarking tool, called $\text{Prism} , with which we further investigate the resource-related factors that hinder the current scaling attempts. Furthermore, we discuss various topology and hardware innovations which could help to scale up blockchain. Generally speaking, this article provides blockchain researchers many valuable insights because it explores the next-generation scaling strategies by conducting a large-scale benchmarking. ... Personal data that is rested in the hands of third parties poses a greatest risk as they might compromise our data for their selfinterests. Goal is to develop a system where data produced by IoT devices is stored in a decentralized distributed system and privacy must be guaranteed [10]. This system leaves the decision of viewing, sharing, controlling, accessing the data to the people who are real data owners. ... Article To provide secure and reliable services using the internet of things (IoT) in the smart cities/villages is a challenging and complex issue. A high throughput and resilient services are required to process vast data generated by the smart city/villages that felicitates to run the applications of smart city. To provide security and privacy a scalable blockchain (BC) mechanism is a necessity to integrate the scalable ledger and transactions limit in the BC. In this paper, we investigated the available solutions to improve its scalability and efficiency. However, most of the algorithms are not providing the better solution to achieve scalability for the smart city data. Here, proposed and implemented a hybrid approach to improve the scalability and rate of transactions on BC using practical Byzantine fault tolerance and decentralized public key algorithms. The proposed Normachain is compares our results with the existing model. The results show that the transaction rate got improved by 6.43% and supervision results got improved by 17.78%. ... Due to the "random characteristics" of hash functions, if a miner has more computing power than another one, the former has more chances to nd its nonce rst. That leads to miners paying more specialised hardware [148]. The race to be the fastest and having more powerful machines leads to a high energy-consumption; Bitcoin is not environment friendly, which is one of the rst and main problems of proof-of-work. ... Thesis Blockchains are one of the most appealing technologies over the last years, both for scientists and for the general public. Blockchains are distributed ledgers that aim to offer transparency, integrity and many more advantages over their centralised counterparts. Blockchains were “revealed” and became popular thanks to the creation and rise of the cryptocurrency Bitcoin. Over the years, blockchain technologies become more and more popular with an exceptional peak in 2017. Blockchains are becoming mainstream technologies, as there is an observatory for blockchains established by the European Commission, blockchain forums in many countries, blockchain start-ups are flourishing, scientific conferences are discussing the topic, and even some scientific conferences are now specifically dedicated to the technology, etc. The blockchain technology promises, thanks to its integrity and transparency properties to be useful and interesting in various domains, and not only for financial systems. However, many questions and doubts float around it. Is it environmentally viable? Is the technology even ensuring its promises? Can they be used in real-life settings, etc. In this thesis, using the lens of distributed systems, we study and define the properties of committee-based blockchains and their fairness; that definition allows formalising and helping correct one of the most used blockchain of that class. Furthermore, adding lenses from game theory, we propose a methodology to analyse the rational behaviours of participants in a blockchains system. Using that methodology, we show, under different mechanisms of rewards, the necessary conditions needed to ensure the blockchain properties. ... The hardware mining process has evolved from a CPUbased system to GPU, FPGA, and ASICs [44]. The authors in [45] prove that ASIC Miners-based systems are more efficient than CPU, GPU, and FPGA in terms of area and energy consumption for blockchain-based applications. ... Article Full-text available Blockchain is among the most promising new technologies due to its unique features, encompassing security, privacy, data integrity, and immutability. Blockchain applications include cryptocurrencies such as Bitcoin. Recently, many other applications have begun to deploy blockchain in their systems. These applications include internet of things (IoT) environments. Although deploying blockchain in IoT architecture has yielded numerous advantages, issues and challenges have arisen that require further research. Most IoT devices and platforms have limited storage capacity, low battery power, and limited hardware resources for computation and network communication. Thus, energy efficiency is a critical factor in these devices. On the other hand, blockchain requires extensive resources and high computational capabilities for mining and communication processes. Balancing computation complexity and IoT resources is a fundamental design challenge in implementing blockchain functions, including the hash function, which is crucial to blockchain design for the mining process. In this study, we present a literature review on the common hash functions used in blockchain-based applications, in addition to the lightweight hash functions available in literature. We evaluate and test the common lightweight hash functions (SPONGENT, PHOTON, and QUARK) on FPGA platforms to determine which is most suitable for blockchain-IoT devices. Moreover, we assess lightweight hash functions in terms of area, power, energy, security, and throughput. The results show tradeoffs between these hash functions. SPONGENT performs best on security and throughput. QUARK consumes the least power and energy but has the lowest security parameters. PHOTON utilizes less area and offers a balance between multiple performance metrics (area, energy, and security), rendering it the most suitable lightweight hash function. ... This results in faster operations, due to operations parallelizing [261] and in energy savings [262]. When hardware based mining solutions took over the computing power dedicated in mining activities experienced, despite strong fluctuations, an exponential growth [263]. It worth nothing that alternative PoW-schemes try to compensate the incredible waste of energy with useful work at an academic level; Primecoin [264] searches for prime numbers chains (Cunningham chain [265]), NooShare [266] executes Monte-Carlo simulations, Shoker [267] proposes matrix-product problems to solve while in [268] authors propose to replace PoW hashing function with alternative one-way functions satisfying additional properties. ... Thesis Blockchain is a technology making the shared registry concept from distributed systems a reality for a number of application domains, from the cryptocurrency one to potentially any industrial system requiring decentralized, robust, trusted and automated decision making in a multi-stakeholder situation. Nevertheless, the actual advantages in using blockchain instead of any other traditional solution (such as centralized databases) are not completely understood to date, or at least there is a strong need for a vademecum guiding designers toward the right decision about when to adopt blockchain or not, which kind of blockchain better meets use-case requirements, and how to use it. At first, we aim at providing the community with such a vademecum, while giving a general presentation of blockchain that goes beyond its usage in Bitcoin and surveying a selection of the vast literature that emerged in the last few years. We draw the key requirements and their evolution when passing from permissionless to permissioned blockchains, presenting the differences between proposed and experimented consensus mechanisms, and describing existing blockchain platforms. Furthermore, we present the B-VMOA blockchain to secure virtual machine orchestration operations for cloud computing and network functions virtualization systems applying the proposed vademecum logic. Using tutorial examples, we describe our design choices and draw implementation plans. We further develop the vademecum logic applied to cloud orchestration and how it can lead to precise platform specifications. We capture the key system operations and complex interactions between them. We focus on the last release of Hyperledger Fabric platform as a way to develop B-VMOA system. Besides, Hyperledger Fabric optimizes conceived B-VMOA network performance, security, and scalability by way of workload separation across: (i) transaction execution and validation peers, and (ii) transaction ordering nodes. We study and use a distributed execute-order-validate architecture which differentiates our conceived B-VMOA system from legacy distributed systems that follow a traditional state-machine replication architecture. We parameterize and validate our model with data collected from a realistic testbed, presenting an empirical study to characterize system performance and identify potential performance bottlenecks. Furthermore, we present the tools we used, the network setup and the discussion on empirical observations from the data collection. We examine the impact of various configurable parameters to conduct an in-dept study of core components and benchmark performance for common usage patterns. Namely, B-VMOA is meant to be run within data center. Different data center interconnection topologies scale differently due to communication protocols. Enormous challenges appear to efficiently design the network interconnections so that the deployment and maintenance of the infrastructure is cost-effective. We analyze the structural properties of several DCN topologies and also present some comparison among these network architectures with the aim to reduce B-VMOA overhead costs. From our analysis, we recommend the hypercube topology as a solution to address the performance bottleneck in the B-VMOA control plane caused by gossip dissemination protocol along with an estimate of performance improvement. ... Over time, Bitcoin miners have turned to increasingly specialized hardware equipment with higher computing power (Bedford Taylor, 2017). Whereas miners initially used central processing units (CPUs) to find PoWs, they quickly realized that graphic processing units (GPUs) were better equipped for the task. ... Article Bitcoin’s increasing energy consumption has triggered a passionate debate about the sustainability of the digital currency. And yet, most studies have thus far ignored that Bitcoin miners cycle through a growing amount of short-lived hardware that could exacerbate the growth in global electronic waste. E-waste represents a growing threat to our environment, from toxic chemicals and heavy metals leaching into soils, to air and water pollutions caused by improper recycling. Here we present a methodology to estimate Bitcoin’s e-waste and find that it adds up to 30.7 metric kilotons per annum, per May 2021. This number is comparable to the amount of small IT and telecommunication equipment waste produced by a country like the Netherlands. At peak Bitcoin price levels seen early in 2021, the annual amount of e-waste may grow beyond 64.4 metric kilotons in the midterm, which highlights the dynamic trend if the Bitcoin price rises further. Moreover, the demand for mining hardware already today disrupts the global semiconductor supply chain. The strategies we present may help to mitigate Bitcoin’s growing e-waste problem. ... Practically, individual miners receive a reward only occasionally, relative to their computational power. As more miners join the network [2] and with faster mining hardware available [3], the total computational power of miners (called the hash rate) has been growing rapidly [4]. The mining difficulty is set by the protocol relative to the hash rate [5] so the probability of mining a new block with the same hardware becomes lower as hash rate increases. ... Conference Paper Full-text available We provide an empirical analysis of pool hoppingbehavior among 15 mining pools throughout Bitcoin’s history.Mining pools have emerged as major players to ensure that theBitcoin system stays secure, valid, and stable. Individual minersjoin mining pools to benefit from a more predictable income.Many questions remain open regarding how mining pools haveevolved throughout Bitcoin’s history and when and why minersjoin or leave mining pools. We propose a heuristic algorithm toextract the payout flow from mining pools and detect the pools’migration of miners. Our results showed that payout schemesand pool fees influence miners’ decisions to join, change, orexit from a mining pool, thus affecting the dynamics of miningpool market shares. Our analysis provides evidence that miningactivity becomes an industry as miners’ decisions follow classicaleconomic rationale. ... In the public blockchain scenario, miners get rewards for verifying transactions. The difficulty level of mining is always adjusted in a certain time interval based on the expected time and actual time taken to mine a certain number of blocks [15]. However, in the case of the proposed system, this approach might not work. ... Article The Internet of Unmanned Vehicles (IoUV) is regarded as an emerging technology for military applications not only to make surveillance systems and battlefield operations fully coordinated and automated, but also to provide significant strategic advantages. All the UVs form a coordinated network for exchanging information in IoUV, which enhances context awareness and risk analysis, and improves response time to make the mission effective. This article provides a perspective of fusion of blockchain technology with software defined IoUV, also known as extended 5G, for the battlefield scenario. Software defined IoUV dynamically configures the network parameters and provides network visibility for better security and manageability of the network in IoUV. Block-chain technology helps to provide trustworthy command and control operations in IoUV and stores those operations in tamper-resistant digital ledgers in the form of transactions. Unmanned vehicles (UVs) with high computing, storage, battery life and transmission power act as miners that validate every transaction and are responsible for creating blocks of the blockchain. However, current blockchain technology suffers from scalability; thus, we propose sharding-enabled blockchain to address the issue where shards of lightweight UVs maintain the required number of miners/ auditors to handle the issues when a miner gets destroyed or damaged in the battlefield. Some participating UVs are used as wireless stations to provide persistent wireless connectivity in IoUV. The proposed framework aims to increase the trust and accountability, and to reduce business friction among different units involved in the battlefield. ... The column 'maintained' identifies if the ASIC for the hash function is still under development. As a continuation of the work in [104], it can be observed that the opportunities have diminished to implement novel ideas for executing the algorithm, such as it was during the first to third generations of mining and in the early years of the fourth generation. Figure 7 shows the number of watts of electricity for 1000 hash calculations per second, sorted by the ASIC technology evolution from 55 nanometers until the current designs of 7 nanometers, between March 2014 and May 2020. ... Preprint Full-text available Precision, validity, reliability, timeliness, availability, and granularity are the desired characteristics for data and information systems. However due to the desired trait of data mutability, information systems have inherently lacked the ability to enforce data integrity without governance. A resolution to this challenge has emerged in the shape of blockchain architecture, which ensures immutability of stored information, whilst remaining in an online state. Blockchain technology achieves this through the serial attachment of set-sized parcels of data called blocks. Links (liken to a chain) between these blocks are implemented using a cryptographic seal created using mathematical functions on the data inside the blocks. Practical implementations of blockchain vary by different components, concepts, and terminologies. Researchers proposed various architectural models using different layers to implement blockchain technologies. In this paper, we investigated those layered architectures for different use cases. We identified essential layers and components for a generalised blockchain architecture. We present a novel three-tiered storage model for the purpose of logically defining and categorising blockchain as a storage technology. We envision that this generalised model will be used as a guide when referencing and building any blockchain storage solution. ... This would mean improved security and more profit for independent miners contributing to the security of the network as it becomes more decentralized, compared with the mining pools of the Bitcoin network centralizing hash power and resources of the network. Nonetheless, Ethereum dismisses a memoryhard requirement for the participating miners to encourage miners using affordable and general GPU instead of relying on ASIC (Application-Specific Integrated Circuit) producers, as explained in [12], and thus preventing ASIC producers and miners from forging a 51% share or attack to Ethereum network. ... Preprint Full-text available Blockchain is a continuously developing technology that has made digital transactions and related computing operations more transparent and secure through globally distributed and decentralized management of states, as well as the strong immutability of blocks mined and transactions validated in a network enabled by the blockchain technology. This manuscript is aimed at elaborating the concept of blockchain technology alongside its coordination and implementation with other emerging technologies, such as smart contract, which works with different blockchain frameworks, as well as enabling anonymous transactions and decentralized consensus amongst different untrusting parties. The discussion of blockchain forks is also covered in this manuscript, depicting fork events created in the blockchain process, their brief history, types, and impacts upon the blockchain development and operation. ... The common ones amongst these models are "mixture of distributions" models. Taylor (2017), "asymmetric information" models and "differences in opinion" models. According to the "asymmetric information" models, investors submit trades based on available private information. ... Article Full-text available This paper offers a plausible response to “what explains the sporadic volatility in the price of Bitcoin?” We hypothesized that market “fundamentals” and “information demands” are key drivers of Bitcoin’s unpredictable price fluctuation. We adopt the transfer-function [Autoregressive Distributed Lag, ARDL] model and its Bounds testing approach to verify how the volatility of the price of Bitcoin responds to its transaction volume, cryptocurrency market capitalisation, world market equity index and Google search. We found the existence of long-run cointegration relation and observed that all the variables except the equity index positively explain the volatility of Bitcoin price. The result established evidence that market fundamentals drive erratic swing in Bitcoin price than information. ... Furthermore, in the implementation of smart contracts, in capacity of transactions or in validation of protocols, design limitations are faced by the technology of blockchain. Additionally, to resolve the tendency of centralized approach [61], some methods should be introduced. • Governance, legal, organizational and regulatory aspects ... Chapter In the world of technology, the Internet of Things (IoT) is a network to link entire things, that is, people, devices, and systems, with each other through an approach of common networking. This technology constructed a way, where many of the routine devices or things are interrelated and easily communicated with, their surroundings to gather or transfer the information over the network without the need of any human-to-system communication or human-to-human communication. It is born with features such as dynamics, scalability, and heterogeneity, and only that network solution can adapt to it which has strategy to incorporate its features. And here comes data centric interaction paradigm, it applies an approach of data naming to comprise the dynamics, scalability, and heterogeneity features to adapt to IoT and composes NDN of things, that is, Named Data Networking of Things (NDNoT). This paradigm also familiarizes the readers with the various kinds of security irruptions that occur in the network, and blockchain acts as a solution for those security attacks in NDNoT. The technology of blockchain is born with a cryptocurrency known as Bitcoin. It is an undoubtedly amazing and innovative invention in the world of technology. The cryptographic algorithms used by blockchains would make consumer data more private. So the technology of blockchain could perhaps be the silver bullet required by the industry of Internet of Things. ... Miners must carefully weigh the costs of investing in mining hardware. A simple solution is comparing the purchase price and operating expenses (power, maintenance, rent, and so on), converted into BTCs, to the net mining returns in BTCs at the end of the machine's life [64]. ... Article Full-text available Blockchain is a very attractive technology since it maintains a public, append-only, immutable and ordered log of transactions which guarantees an auditable ledger accessible by anyone. Blockchain systems are inherently interdisciplinary since they combine various fields such as cryptography, multi-agent systems, distributed systems, social systems, economy, and finance. Furthermore, they have a very active and dynamic ecosystem where new blockchain platforms and algorithms are developed continuously due to the interest of the public and the industries to the technology. Consequently, we anticipate a challenging and interdisciplinary research agenda in blockchain systems, built upon a methodology that strives to capture the rich process resulting from the interplay between the behavior of agents and the dynamic interactions among them. To be effective, however, modeling studies providing insights into blockchain systems, and appropriate description of agents paired with a generic understanding of their components are needed. Such studies will create a more unified field of blockchain systems that advances our understanding and leads to further insight. According to this perspective, in this study, we propose using a generic multi-agent organizational modeling for studying blockchain systems, namely AGR4BS. Concretely, we use the Agent/Group/Role (AGR) organizational modeling approach to identify and represent the generic entities which are common to blockchain systems. We show through four real case studies how this generic model can be used to model different blockchain systems. We also show briefly how it can be used for modeling three well-known attacks on blockchain systems. ... In Bitcoin network, most of the mining power is controlled by mining pools, and most of the hash rate is produced by "mining factories" that equip ASIC mining machines [1]. Since ASIC mining machines have limited purposes other than mining, the economic behavior of a miner depends on its longterm prediction rather than a short-term benefit. ... ... This would mean improved security and more profit for independent miners contributing to the security of the network as it becomes more decentralized, compared with the mining pools of the Bitcoin network centralizing hash power and resources of the network. Nonetheless, Ethereum dismisses a memoryhard requirement for the participating miners to encourage miners using affordable and general GPU instead of relying on ASIC (Application-Specific Integrated Circuit) producers, as explained in [12], and thus preventing ASIC producers and miners from forging a 51% share or attack to Ethereum network. ... Preprint Full-text available Blockchain is a continuously developing technology that has made digital transactions and related computing operations more transparent and secure through globally distributed and decentralized management of states, as well as the strong immutability of blocks mined and transactions validated in a network enabled by the blockchain technology. This manuscript is aimed at elaborating the concept of blockchain technology alongside its coordination and implementation with other emerging technologies, such as smart contract, which works with different blockchain frameworks, as well as enabling anonymous transactions and decentralized consensus amongst different untrusting parties. The discussion of blockchain forks is also covered in this manuscript, depicting fork events created in the blockchain process, their brief history, types, and impacts upon the blockchain development and operation. ... The miners are the ones who keep the record of all the payments. Mining is typically done on a dedicated computer [77], as it requires a fast CPU and higher electricity usage, and more heat generated than typical computer operations [76]. ... Article Full-text available Network slicing is one of the fundamental tenets of Fifth Generation (5G)/Sixth Generation (6G) networks. Deploying slices requires end-to-end (E2E) control of services and the underlying resources in a network substrate featuring an increasing number of stakeholders. Beyond the technical difficulties this entails, there is a long list of administrative negotiations among parties that do not necessarily trust each other, which often requires costly manual processes, including the legal construction of neutral entities. In this context, Blockchain comes to the rescue by bringing its decentralized yet immutable and auditable lemdger, which has a high potential in the telco arena. In this sense, it may help to automate some of the above costly processes. There have been some proposals in this direction that are applied to various problems among different stakeholders. This paper aims at structuring this field of knowledge by, first, providing introductions to network slicing and blockchain technologies. Then, state-of-the-art is presented through a global architecture that aggregates the various proposals into a coherent whole while showing the motivation behind applying Blockchain and smart contracts to network slicing. And finally, some limitations of current work, future challenges and research directions are also presented. ... For computing SHA-256, miners use mining ASIC (application-specific integrated circuit) machines. ASIC is by far more efficient than all-purpose computers, and mining is profitable only when miners use mining ASIC machines (Taylor, 2017). Mining machines are often compatible with multiple cryptocurrencies: Most SHA-256 ASIC machines list BTC, BCH, and BSV as "minable currencies" in their advertised specs. ... Preprint Full-text available Proof-of-Work cryptocurrencies, such as Bitcoin and its forks, hire miners (freelance contributors) to maintain the system by algorithmically setting the reward. Therefore, the nature of miners' labor supply is essential for the cryptocurrency's stability. We develop a short-run supply-side model of the multicurrency mining market and estimate miners' labor supply elasticity by exploiting the discontinuity created by an event called halving. The stability of Bitcoin hinges on external factors lowering the labor supply elasticity, such as the interaction with competing currencies. Upgrading algorithm can stabilize Bitcoin regardless of external factors and improve the mining market's energy consumption rate by 2.9%. ... Central processing units, graphics processing units, FPGAs, and ASICs are just a few of the mining-specific equipment that are employed. In this article, Mining with the help of Graphics Processing Units to deploy the blockchain technology [24]. ... Conference Paper Full-text available Blockchain Technology is considered as an attention-seeking technology and still in its early years according to many researchers. Cryptocurrencies like Bitcoin, Ethereum, Dogecoin and many more have emerged as new area of research for the fintech community as they are using blockchain technology. Bitcoin (BTC) emerged as the first real-world application of the blockchain technology. Bitcoin and other cryptocurrencies depend on the mining of the raw data which is going to be stored in the blocks. These cryptocurrencies depend on crypto mining, which is a complex computation of mathematical equations and problems. To solve these equations and problems, highpowered computing is required. The Graphics Processing Units work on the principle of high-power and highperformance computing. They are also extremely useful for solving complex computation of mathematical equation and problems. We have observed the usage of GPUs to deploy blockchain for mining of cryptocurrencies like Bitcoin, Ethereum, and Dogecoin. The research describes about the applications of GPU and FPGA for deployment of Blockchain for Cryptocurrency. ... In many real-world computing tasks, the relevant metric for performance is not the total computing power of the system, but rather the computations-per-watt. A compelling example is the adoption of specialized hardware designed for Bitcoin mining, which can easily exceed 40 times the performance-per-watt of a graphics processing unit (GPU) [34]. Cytomorphic chips operate at tens of milliwatts, yet in many cases still perform equal or faster simulations than desktop computers operating at tens of watts, representing a more than 1000-fold improvement in performance-per-watt for general networks [3,6,8,9,11]. ... Article Full-text available The explosive growth in semiconductor integrated circuits was made possible in large part by design automation software. The design and/or analysis of synthetic and natural circuits in living cells could be made more scalable using the same approach. We present a compiler which converts standard representations of chemical reaction networks and circuits into hardware configurations that can be used to simulate the network on specialized cytomorphic hardware. The compiler also creates circuit-level models of the target configuration, which enhances the versatility of the compiler and enables the validation of its functionality without physical experimentation with the hardware. We show that this compiler can translate networks comprised of mass-action kinetics, classic enzyme kinetics (Michaelis-Menten, Briggs-Haldane, and Botts-Morales formalisms), and genetic repressor kinetics, thereby allowing a large class of models to be transformed into a hardware representation. Rule-based models are particularly well-suited to this approach, as we demonstrate by compiling a MAP kinase model. Development of specialized hardware and software for simulating biological networks has the potential to enable the simulation of larger kinetic models than are currently feasible or allow the parallel simulation of many smaller networks with better performance than current simulation software. ... The surge of Bitcoin price and mining profitability incentivized increasing computing power to participate in the game. Moreover, specific mining rigs were quickly designed, manufactured and upgraded 5 . Mining sites were purposefully selected and developed. ... Article Full-text available Bitcoin mining is not only the fundamental process to maintain Bitcoin network, but also the key linkage between the virtual cryptocurrency and the physical world. A variety of issues associated with it have been raised, such as network security, cryptoasset management and sustainability impacts. Investigating Bitcoin mining from a spatial perspective will provide new angles and empirical evidence with respect to extant literature. Here we explore the spatial distribution of Bitcoin mining through bottom-up tracking and geospatial statistics. We find that mining activity has been detected at more than 6000 geographical units across 139 countries and regions, which is in line with the distributed design of Bitcoin network. However, in terms of computing power, it has demonstrated a strong tendency of spatial concentration and association with energy production locations. We also discover that the spatial distribution of Bitcoin mining is dynamic, which fluctuates with diverse patterns, according to economic and regulatory changes. ... This results in faster operations, due to operations parallelizing [147] and in energy savings [148]. When hardware based mining solutions took over the computing power dedicated in mining activities experienced, despite strong fluctuations, an exponential growth [149]. It worth nothing that alternative PoW-schemes try to compensate the incredible waste of energy with useful work at an academic level ; Primecoin [150] searches for prime numbers chains (Cunningham chain [151]), NooShare [152] executes Monte-Carlo simulations, Shoker [153] proposes matrix-product problems to solve while in [154] authors propose to replace PoW hashing function with alternative one-way functions satisfying additional properties. ... Thesis The disruptive technology born in 2008 with Bitcoin and known as blockchain represents a significant quality leap from the distributed database technology. Distributed systems theory provides then models and techniques to analyze some protocols characterizing the technology, however in order to analyze a blockchain system additional considerations on its users need to be done. This thesis aims at analyzing the different behaviors of the users operating in blockchains or more in general in DLTs (i.e., Distributed Ledger Technologies).The latter are considered as rational agents, fully aware of all actions available to them and capable of choosing the one they feel is the best for themselves. Game theory is then used to model situations where users are called to choose and perform certain actions within the DLT environment. This thesis analyzes different users as well as different blockchains with the scope of providing a general overview on the topic and formal results on their behaviors ; users may indeed be honest vis-à-vis of other users or they may behave maliciously (as Byzantine nodes) attacking the blockchain system. ... Over the years, mining became a highly profitable business. Therefore, specialized tools (called ASICs) were developed to mine more efficiently and miners created mining pools to increase their chances (Taylor, 2017). As shown in table 2, the block size of Bitcoin is roughly one megabyte. ... Thesis Over the years, machine learning established its presence in the world of finance. Using machine learning to forecast the price of financial assets is nowadays a common practice. Notably, the prediction of stocks often underlies machine learning algorithms. Over the last two decades, a new form of assets appeared, commonly known as decentralized currencies or simply cryptocurrencies. Although machine learning algorithms were proven to be successful in predicting stocks, they experienced difficulties in forecasting cryptocurrencies. While there are scholars claiming to have developed viable algorithms, there are plenty of other scholars arguing that predicting cryptocurrencies can not be done in an efficient manner. Therefore, this thesis aims at answering the question, "Can a useful prediction algorithm for the price of cryptocurrencies be created?" This thesis analyzed three different approaches. First, an Auto-Regressive Integrated Moving Average model was developed. Second, a Bayesian Neural Network was created. Third, a Long Short-Term Memory Network was designed. Each approach predicted the prices for three different cryptocurrencies: Bitcoin, Ethereum and Litecoin. The results showed that the Bayesian Neural Network could predict the prices of Bitcoin and Ethereum in a useful manner. The achieved profit for Bitcoin was 13'541 USD and for Ethereum 7'646 USD. Considering the investments, an ROI of 24.62% for Bitcoin and of 13.9% for Ethereum resulted. ... Practically, individual miners receive a reward only occasionally, relative to their computational power. As more miners join the network [2] and with faster mining hardware available [3], the total computational power of miners (called the hash rate) has been growing rapidly [4]. The mining difficulty is constantly adjusted by the protocol relative to the hash rate [5], so the probability of mining a new block with the same hardware becomes lower as the hash rate increases. ... Article Full-text available We analyzed 23 mining pools and explore the mobility of miners throughout Bitcoin's history. Mining pools have emerged as major players to ensure that the Bitcoin system stays secure, valid, and stable. Many questions remain open regarding how mining pools have evolved throughout Bitcoin's history and when and why miners join or leave the pools. We investigated the reward payout flow of mining pools and characterized them based on payout irregularity and structural complexity. Based on our proposed algorithm, we identified miners and studied their mobility in the pools over time. Our analysis shows that Bitcoin mining is an industry that is sensitive to external events (e.g., market price and government policy). Over time, competition between pools involving reward schemes and pool fees motivated miners to migrate between pools (i.e., pool hopping and cross pooling). These factors converged toward optimal scheme and values, which made mining activities more stable. ... • Bitcoin mining experiences substantial technological changes in computing power by improving the processing power (hashrate). According to Taylor [2017], the hardware for Bitcoin mining has evolved rapidly since the invention of Bitcoin. ... Article Recent papers that have explored spot and futures markets for Bitcoin have concluded that price discovery takes place either in the spot, or the futures market. Here, we consider the robustness of previous price discovery conclusions by investigating causal relationships, cointegration and price discovery between spot and futures markets for Bitcoin, using appropriate daily data and time-varying mechanisms. We apply the time-varying Granger causality test of Shi, Phillips, and Hurn [2018]; time-varying cointegration tests of Park and Hahn [1999], and time-varying information share methodologies, concluding that futures prices Granger cause spot prices and that futures prices dominate the price discovery process. ... Today, the start up and operational costs, coupled with the intense competition and the ever decreasing rewards considerably outweigh the potential profits from independent mining (Taylor, 2017). Therefore, joining a mining pool may be the only option as it increases miners' financial prospects (Maurer et al., 2013). ... Article Full-text available In this study, I use the Critical Realism perspective of power to explain how the Bitcoin protocol operates as a system of power. I trace the ideological underpinnings of the protocol in the Cypherpunk movement to consider how notions of power shaped the protocol. The protocol by design encompasses structures, namely the Proof of Work and the Block Selection, that reproduce asymmetrical constraints on the entities that comprise it. These constraining structures generate constraining mechanisms, those of cost effectiveness and deanonymisation, which further restrict participating entities’ ‘power to act’, reinforcing others’ ‘power over’ them. In doing so, I illustrate that the Bitcoin protocol, rather than decentralising and distributing power across a network of numerous anonymous, trustless peers, it has instead shifted it, from the traditional actors (e.g., state, regulators) to newly emergent ones. ... Examples of network statistics for each month include the market price, total hash rate, total block reward, total transaction fees, mining difficulty, or the number of transactions. Other external data that I included in Bitcoin statistics are mining hardware evolution [82,151], electricity consumption [14], and the global energy price index [68]. ... Thesis Full-text available Bitcoin is a pioneer cryptocurrency that records transactions in a public distributed ledger called the blockchain. It has been used as a medium for payments, investments, and digital wallets that are not controlled by any government or financial institution. Over the past ten years, transaction activities in Bitcoin have increased rapidly. The volume and evolving nature of its data pose analysis challenges to explore diverse groups of users and different activities on the network. The field of Visual Analytics (VA) has been working on the development of analytical systems that allow humans to interact and gain insights from complex data. In this thesis, I make several contributions to the analysis of Bitcoin mining activity. First, I provide a characterization of the past work and research challenges related to VA for blockchains. From this assessment, I proposed a VA tool to understand mining activities that ensure data integrity and security on the Bitcoin blockchain. I propose a method to extract miners from the transaction data and trace pool hopping behavior. The empirical analysis of this data revealed that emerging mining pools provided a better incentive to attract miners. Simultaneously, miners strategically chose mining pools to maximize their profit. To explore the evolution and dynamics of this activity over the long term, I developed a VA tool called MiningVis that integrates mining behavior data with contextual information from Bitcoin statistics and news. The user study demonstrates that Bitcoin miner participants use the tool to analyze higher-level mining activity rather than mining pool details. The evaluation of the tool proves that it helped participants to relate multiple information and discover historical trends of Bitcoin mining. Article Presents the introductory editorial for this issue of the publication. Conference Paper Full-text available Blockchain is a public ledger which is distributed in nature and has become highly popular. Bitcoin is the most successful application of it. The reason behind Bitcoin's success lies in its Consensus mechanism which ensures the security from any kind of attack. Because of this, no dishonest miner can affect the chain to manipulate it according to their wish. Also, the double spending problem does not occur and there is no need to trust a third party in this network. The most common consensus protocol in bitcoin technology is Proof-of-Work (PoW) which entirely depends on the computation power of miners. Because of this dependency, Application-specific Integrated Circuits (ASIC) is designed for bitcoin mining. Thereafter, it has become a threat to its decentralized nature and has been monopolizing the validation of new blocks. It is not possible to halt the production of ASIC-based devices even if it threatens the decentralized applications. So, different types of consensus protocols are proposed to nullify this threat whereas all of them have failed to fully nullify it. ASIC devices are costly, so only a few miners can afford it and monopolize over blockchain network. In our work, Filtered Proof-of-Work (FPoW) is proposed and its ASIC-resistivity has been evaluated to make it a future-proof ASIC-resistant consensus protocol. Chapter SHA-256 is a completely unpredictable pseudorandom function which generates unique output for a given input ensuring data authenticity and non-repudiation. It is the cornerstone for imparting security and privacy into Blockchain and its efficiency of calculation decides the performance of Blockchain. In this paper, we propose two novel methods to accelerate the calculation of SHA-256 in different situations. To eliminate the useless operations, we present pre-expanded padding blocks and hard coded into the software. On this basis, for a single message containing multiple 512-bit message blocks, we propose Interleaved Multi-Vectorizing Message Scheduling (IMV-MS) to optimize the message schedule stage of SHA-256, which utilize the interleaved multi-vectorizing (IMV) to combine single instruction multiple data (SIMD) vectorization with SHA-256. It splits a vectorized program into multiple states, then it interleaves the execution of vectorized states from those running instances which can make full use of the data parallelism in SIMD. On the other hand, in the situation where we hash several messages simultaneously, we propose the modified SHA-256 which employs SIMD instructions and thread-level parallelism technology together to realize parallel optimization on SHA-256. As experimental results show, IMV-MS and the proposed SHA-256 achieves up to 6.36X, 60.38X better performance compared with the pure SIMD vectorization and the pure scalar implementation, respectively. Article The proliferation of the IoT in connected society is rapidly expanding into vertical industry sectors due to the ever-increasing ties amongst businesses and economies. As the number of IoT nodes utilized in a network increases, decentralized network infrastructure, and security provisioning mechanisms, primarily enabled by blockchain-based technologies, become more beneficial. However, blockchain-based IoT networks experience transaction throughput degradation due to the platform's cryptographically-based security features, where negotiating with ledger maintainers for faster processing is a must. Existing transaction processing schemes are mainly geared towards digital currency applications. In overcoming these challenges, a novel feeless transaction processing algorithm is proposed for non-cryptocurrency blockchain-based IoT networks. The proposed algorithm enables ledger maintainers in achieving desired processing throughputs for select transactions found in a miner's transaction pool. A utility function is designed to select transactions from miners' transaction pools to form blocks that add a desired operational value for achieving pre-determined production outputs over blockchain-based networks. Furthermore, the proposed scheme will utilize an aging process to increase the likelihood of selecting transactions with larger miners' transaction pool residence times. The simulation and implementation results show that the proposed methodologies increase the processing throughput of desired transactions while preventing transaction processing starvation. Article Full-text available The controversies surrounding Bitcoin, one of the most frequently used and advertised cryptocurrency, are focused on identifying its qualities, the advantages and disadvantages of using it and, last but not least, its ability to survive over time and become a viable alternative to the traditional currency, taking into account the effects on the environment of the technology used to extract and trade it. Based on such considerations, this article aims to provide an overview of this cryptocurrency, from the perspective of conducting a systematic review of the literature dedicated to the economic and environmental impact of Bitcoin. Using peer-reviewed articles collected from academic databases, we aimed at synthesizing and critically evaluating the points of view in the scientific literature regarding the doctrinal source of the emergence of Bitcoin, the identity of this cryptocurrency from an economic point of view, following its implications on the economic and social environment. Subsequently, this research offers the opportunity of evaluating the level of knowledge considering the impact of Bitcoin mining process on the environment from the perspective of the energy consumption and CO2 emissions, in order to finally analyze Bitcoin regulation and identify possible solutions to reduce the negative impact on the environment and beyond. The findings suggest that, despite high energy consumption and adverse environmental impact, Bitcoin continues to be an instrument used in the economic environment for a variety of purposes. Moreover, the trend of regulating it in various countries shows that the use of Bitcoin is beginning to gain some legitimacy, despite criticism against this cryptocurrency. Thesis L’Internet des Objets (IdO) est une infrastructure mondiale, visant à connecter une multitude de systèmes de diverses natures, dans le but de générer de nouveaux services. Alors que cette hyperconnectivité laisse entrevoir de nombreuses perspectives dans tous les secteurs d’activité (e.g. la santé, l’industrie, l’énergie etc.), son déploiement constitue un défi majeur en termes de sécurité et de vie privée. L’usage de la cryptographie devient alors primordial, notamment pour assurer la confidentialité des données. Néanmoins, les algorithmes classiques tels que l’AES ne sont pas conçus pour être embarqués dans des systèmes à ressources limitées que sont les objets connectés. Pour répondre à cette problématique, la communauté scientifique s’oriente vers la cryptographie dite « légère ». Malgré ce qualificatif, jusqu’à quel point cette dernière est-elle pertinente et en adéquation avec les ressources des objets connectés ? Quel niveau de sécurité permet-elle d’atteindre ?Au travers de métriques communes, l’objectif du premier axe de cette thèse est de déterminer les gains qu’apporte la cryptographie légère par rapport à son homologue classique, en termes de taille, de performance et de robustesse face aux attaques par canaux auxiliaires, reconnues comme un vecteur d’attaque majeur. Cette évaluation se focalise sur le standard AES-128 et ses proches dérivés légers PRESENT-80/128 et plus récents GIFT-64-128/GIFT-128-128, dont leur structure orientée au niveau bit favorise l’implémentation matérielle. En termes de performances, GIFT-64-128 se positionne comme la meilleure alternative, avec une réduction de taille de près d’un facteur 10, pour un gain en efficacité de 58% par rapport à l’AES-128. Concernant la résistance aux attaques par canaux auxiliaires, ici la CPA, la cryptographie légère se montre une nouvelle fois plus pertinente, avec PRESENT-128 atteignant une robustesse accrue d’un facteur 3 par rapport à l’AES-128.Bien qu’indispensable, la confidentialité seule n’est pas suffisante pour instaurer un environnement sécurisé. Face à l’émergence des cyberattaques, un écosystème de l’IdO se doit d’intégrer des mécanismes de sécurité pour faire face à diverses menaces. En outre, sa topologie réseau devient propice à la décentralisation pour des questions d’efficacités. Sans organisme central, comment assurer à la fois la fiabilité et la sécurité d’un environnement fortement hétérogène ?La seconde partie de cette thèse propose un protocole de confiance, spécialement conçu pour l’IdO. Inspiré du concept de la blockchain, plusieurs optimisations ont été mises en place, permettant de réduire l’utilisation des ressources au minimum. Dans une approche centrée sur les passerelles réseau, les résultats démontrent une réduction de l’espace de stockage d’un facteur 3 000 par rapport à Bitcoin, ainsi qu’une latence divisée par 18, pour une consommation inférieure à celle d’un chargeur de téléphone. Enfin, une extension du protocole est proposée pour être embarquée dans les objets. Le modèle se concentre sur l’évaluation de la confiance dans un environnement proche, se focalisant sur les entités où une communication directe est possible. Avec moins de 400 octets, le protocole est capable d’évaluer la confiance d’une dizaine d’objets, et jusqu’à une trentaine avec seulement 1 Ko. Chapter This chapter explains how Bitcoin works, and its underlying blockchain technology, which delivers publically verifiable, immutable records whose integrity relies on neither trusted central parties nor secret keys. It also gives an overview of Ethereum, extending Bitcoin to a decentralized computing platform supporting what are called smart contracts. End notes provide references to explore underlying principles and further details. Article Full-text available The art industry has commercialised and popularised non-fungible tokens (NFTs), with the volume and value of NFT transactions rapidly growing to US$ 10.7 billion in Q3 2021. The increase in NFT transactions has drawn the attention of the art market to the consequent carbon emissions resulting from verifying transactions in proof-of-work blockchains supporting NFT transactions. With CO2-related deaths attributable to NFT transactions, social pressure from the art market has helped to progress the switch away from the deliberately polluting proof-of-work blockchains to more sustainable consensus protocols. Nonetheless, many popular types of blockchain have resisted the pressure to decrease their environmental impact, including Bitcoin, whose attributed 2021 annual emissions will produce emissions responsible for around 19,000 future deaths. In response, recent global policy interventions have employed legal and fiscal tools to reduce the carbon impact of some or all types of blockchains. Linking the damage caused by proof-of-work blockchains to climate change and human mortality, this study examines the recent policy interventions designed to motivate a shift in blockchain consensus protocols and promote miners' energy efficiency to mitigate environmental damage. This article further explores available policy intervention options that are currently not utilised.
Preprint
Full-text available
We introduce a model of greenhouse gas emissions due to on-chain activity on Ethereum, focusing on cryptoart. We also estimate the impact of individual transactions on the environment, both before and after the London hard fork. We find that with the current fee mechanism, spending one dollar on transaction fees corresponds to emitting at least the equivalent of 1.151 kilograms of CO2. We also describe several techniques to reduce cryptoart emissions, both in the short and long term.
Conference Paper
Bitcoin mining machines become a new driving force to push the physical limitation of semiconductor process technology. Instead of peak performance, mining machines pursue energy and computation efficiency of implementing cryptographic hash functions. Therefore, the state-of-the-art ASIC design of mining machines adopts near-threshold computing, deep pipelines, and uni-directional data flow. According to these design properties, in this paper, we propose a novel clock reversing tree design methodology for bitcoin mining machines. In the clock reversing tree, the clock of global tree is fed from the last pipeline stage backward to the first one, and the clock latency difference between the local clock roots of two consecutive stages maintains a constant delay. The local tree of each stage is well balanced and keeps the same clock latency. The special clock topology naturally utilizes setup time slacks to gain hold time margins. Moreover, to alleviate the incurred on-chip variations due to near-threshold computing, we maximize the common clock path shared by flip-flops of each individual stage. Finally, we perform inverter pair swap to maintain duty cycle. Experimental results show that our methodology is promising for industrial bitcoin mining designs: Compared with two variation-aware clock network synthesis approaches widely used in modern ASIC designs, our approach can reduce up to 64% clock buffer/inverter usage, 12% clock power, decrease 99% hold time violating paths, and achieve 85% area saving for timing fixing. The proposed clock design methodology is general and applicable to blockchain and other ASICs with deep pipelines and strong data flow.
Article
Graphics processing units (GPUs) achieve high throughput by exploiting a high degree of thread-level parallelism (TLP). To support such high TLP, GPUs have a large-sized register file to store the context of all threads, consuming around 20% of total GPU energy. Several previous studies have attempted to minimize the energy consumption of the register file by implementing an emerging non-volatile memory (NVM), leveraging its higher density and lower leakage power over SRAMs. To amortize the cost of long access latency of NVM, prior work adopts a hierarchical register file consisting of an SRAM-based register cache and NVM-based registers where the register cache works as a write buffer. To get the register cache index, they use the partially selected bits of warp ID and register ID. This work observes that such an index calculation causes three types of contentions leading to the underutilization of the register cache: inter-warp, intra-warp, and false contentions. To minimize such contentions, this paper proposes a thread context-aware register cache (TEA-RC) in GPUs. In TEA-RC, the cache index is calculated considering the high correlation between the number of scheduled threads and the register usage of threads. The proposed design shows 28.5% higher performance and 9.1 percentage point lower energy consumption over the conventional register cache that concatenates three bits of warp ID and five bits of register ID to compute the cache index.
Chapter
Approximate computing enables construction of circuits that are faster and more compact and consume less power at the cost of accuracy of computation. Generally, it has been employed in a lot of error-tolerant applications such as image/multimedia signal processing, machine learning, etc., applications that allow accuracy degradation without quality degradation. But, approximation has also the potential being utilized to provide area and power efficient solutions in the domain of information security. This chapter will survey the practicality of deployment of approximate computing for the cryptographic primitives and applications along with the possible consequences on their correctness as well as security-level reduction.KeywordsApproximate computingInformation securityLearning with errors (LWE)Homomorphic encryption (HE)HashBitcoinRandom number generator (RNG)
Article
GPU- and FPGA-based clouds have been deployed to accelerate computationally intensive workloads. ASIC-based clouds are a natural evolution as cloud services expand across the planet. ASIC Clouds are purpose-built datacenters comprising large arrays of ASIC accelerators that optimize the total cost of ownership (TCO) of large, high-volume scale-out computations. On the surface, ASIC Clouds may seem improbable due to high nonrecurring engineering (NRE) costs and ASIC inflexibility, but large-scale ASIC Clouds have been deployed for the Bitcoin cryptocurrency system. This article distills lessons from these Bitcoin ASIC Clouds and applies them to other large-scale workloads, including YouTube-style video-transcoding and Deep Learning, showing superior TCO versus CPU and GPU. It derives Pareto-optimal ASIC Cloud servers based on accelerator properties, by jointly optimizing ASIC architecture, DRAM, motherboard, power delivery, cooling, and operating voltage. Finally, the authors examine the impact of ASIC NRE and when it makes sense to build an ASIC Cloud.
Article
Conference Paper
Article
GPU and FPGA-based clouds have already demonstrated the promise of accelerating computing-intensive workloads with greatly improved power and performance. In this paper, we examine the design of ASIC Clouds, which are purpose-built datacenters comprised of large arrays of ASIC accelerators, whose purpose is to optimize the total cost of ownership (TCO) of large, high-volume chronic computations, which are becoming increasingly common as more and more services are built around the Cloud model. On the surface, the creation of ASIC clouds may seem highly improbable due to high NREs and the inflexibility of ASICs. Surprisingly, however, large-scale ASIC Clouds have already been deployed by a large number of commercial entities, to implement the distributed Bitcoin cryptocurrency system. We begin with a case study of Bitcoin mining ASIC Clouds, which are perhaps the largest ASIC Clouds to date. From there, we design three more ASIC Clouds, including a YouTube-style video transcoding ASIC Cloud, a Litecoin ASIC Cloud, and a Convolutional Neural Network ASIC Cloud and show 2-3 orders of magnitude better TCO versus CPU and GPU. Among our contributions, we present a methodology that given an accelerator design, derives Pareto-optimal ASIC Cloud Servers, by extracting data from place-and-routed circuits and computational fluid dynamic simulations, and then employing clever but brute-force search to find the best jointly-optimized ASIC, DRAM subsystem, motherboard, power delivery system, cooling system, operating voltage, and case design. Moreover, we show how data center parameters determine which of the many Pareto-optimal points is TCO-optimal. Finally we examine when it makes sense to build an ASIC Cloud, and examine the impact of ASIC NRE.
Conference Paper
Bitcoin is a "crypto currency", a decentralized electronic payment scheme based on cryptography. It implements a particular type of peer-to-peer payment system. Bitcoin depends on well-known cryptographic standards such as SHA-256. In this paper we revisit the cryptographic process which allows one to make money by producing new bitcoins. We reformulate this problem as a specific sort of Constrained Input Small Output (CISO) hashing problem and reduce the problem to a pure block cipher problem, cf. Fig. 1. We estimate the speed of this process and we show that the amortized cost of this process is less than it seems and it depends on a certain cryptographic constant which is estimated to be at most 1.89. These optimizations enable bitcoin miners to save countless millions of dollars per year in electricity bills.
Article
There is an intense technological race underway to build the highest-performance and lowest-power custom Bitcoin mining appliances using custom ASIC processors. This article describes the architecture and implementation details of CoinTerra's first-generation Bitcoin mining processor, Goldstrike 1, and how this processor is used to design a complete Bitcoin mining machine called Terraminer IV. Because of high power density in the Bitcoin mining processor, delivering power and cooling the die posed enormous challenges. This article describes some of the solutions adopted to overcome these challenges.
Conference Paper
Recently, the Bitcoin cryptocurrency has been an international sensation. This paper tells the story of Bitcoin hardware: how a group of early-adopters self-organized and financed the creation of an entire new industry, leading to the development of machines, including ASICs, that had orders of magnitude better performance than what Dell, Intel, NVidia, AMD or Xilinx could provide. We examine this story for clues as to how we can foster greater innovation in the semiconductor industry and enable this phenomenon to occur more broadly for more application areas, spawning a new age of hardware innovation tailored to emerging application domains---an Age of Bespoke Silicon.
Article
A purely peer-to-peer version of electronic cash would allow online payments to be sent directly from one party to another without going through a financial institution. Digital signatures provide part of the solution, but the main benefits are lost if a trusted third party is still required to prevent double-spending. We propose a solution to the double-spending problem using a peer-to-peer network. The network timestamps transactions by hashing them into an ongoing chain of hash-based proof-of-work, forming a record that cannot be changed without redoing the proof-of-work. The longest chain not only serves as proof of the sequence of events witnessed, but proof that it came from the largest pool of CPU power. As long as a majority of CPU power is controlled by nodes that are not cooperating to attack the network, they'll generate the longest chain and outpace attackers. The network itself requires minimal structure. Messages are broadcast on a best effort basis, and nodes can leave and rejoin the network at will, accepting the longest proof-of-work chain as proof of what happened while they were gone.
Article
In this paper we describe the various scoring systems used to calculate rewards of participants in Bitcoin pooled mining, explain the problems each were designed to solve and analyze their respective advantages and disadvantages.
For Virtual Prospectors, Life in the Bitcoin Mines Gets Real
• J Light
Into the Bitcoin Mines
• N Popper
Google Built Its Very Own Chips to Power Its AI Bots
• C Metz