Why go ASIC Resistant?

Image for post
Image for post

Since HCASH released its yellow paper in July 2018, the Foundation have been continually reviewing methods of preventing ASIC mining hardware from being deployed, as wide-spread ASIC mining use could effectively create computing power monopolies; a system whereby one or a small number of powerful miners dominate a blockchain network.

Today, after half a year of research and development, the HyperCash team are ready to deploy their new ASIC-resistant algorithm.

The origins of ASIC mining (and its resistance), and the threat of computing power monopolies

HyperCash is not the only project that has realised the potential harm that could be caused by ASIC mining to communities and network stability.

More and more projects are now developing ASIC resistant mining machines.

Take Monero for example. When Bitmain released an ASIC mining machine capable of mining Monero and other CryptoNight algorithm tokens, the Monero community became deeply concerned as it quickly became apparent that the computing power of these mining machines would be way higher than any other mining machine currently in use, thus enabling a monopoly to be formed. To counteract this, the Monero community swiftly decided to change their mining algorithm via a hard fork, to one that incorporates ASIC resistance. The hard fork outcome was widely welcomed by Monero miners and community members. Since the fork, the price of the Monero token has risen sharply (indicating its perceived value) and is widely known as a cryptocurrency whereby GPU mining is still attainable and profitable; something that is considered important to those still viewing cryptocurrencies as a ‘tool for the everyman’, and not another outlet for faceless companies to earn profits from. Further efforts to keep this status-quo have been implemented by Monero’s team. The project announced through social media channels that they will revise their current mining algorithm every 6 months to prevent against new ASIC hardware being developed and implemented, and ultimately, threatening the network.

Another example is Zcash. Zcash, another cryptocurrency that is resistant to ASIC mining machines have gone so far as to set up a technical advisory committee in charge of providing scientific advice on how to remain ASIC-resistant.

Lastly, Ethereum. Just earlier this year, Ethereum’s (ETH) core developers reached a preliminary consensus agreement to implement a new Proof-of-Work (PoW) algorithm that will improve GPU-based efficiency — instead of relying on ASIC-based network mining. It is believed that this approach will not only make ASIC mining more “difficult”, but will also stabilize the networks hash value.

From here, the question must be asked: why do so many blockchain projects have the desire to resist ASIC mining hardware? To understand this, we must understand exactly what an ASIC mining machine is.

When Satoshi Nakamoto was first developing Bitcoin, in his original vision he hoped that people could use home computers to mine tokens. Specifically, he believed that people would rely merely on their CPUs to validate transactions. However, with the increasing value of Bitcoin and other cryptocurrencies, mining has become an industry, leading to increase competitiveness and ultimately, difficulty. Therefore, miners started to purchase and use more advanced hardware, and soon after, companies began appearing with the sole intent of researching, developing and updating specialised hardware to mine tokens that utilise specific algorithms, thus the term ‘application-specific integrated circuit’ miner or ASIC miner slowly grew into the widespread phenomenon it is today.

The reason for this was that CPUs overall act as general-purpose processers. This means that while they are great at processing a multitude of different functions, they’re not effective at completing specific individual tasks. An example of this can be seen in graphics computing. In terms of graphics processing and 3D computing, it’s clear that CPUs aren’t as good at doing this as GPUs are, hence the creation and widespread adoption of GPUs.

An extension of this are ASIC miners. To mine Bitcoin all a computer needs to do is calculate SHA256 hash values; a singular function. And again, even though a general-purpose CPU can calculate these values, they’re not very effective when considering cost and efficiency. Therefore, mining companies have designed application-specific chips whose primary function is for SHA256 (and other algorithms, depending on the token) calculations. Because it is designed to process a specific algorithm, the design of the ASIC chip is much more simple and less expensive than a CPU; and most importantly, in terms of computing power, they are (generally) tens of thousands or more times higher than currently used CPUs and GPUs.

This makes it difficult to mine with normal CPUs and limits the market to those with large amounts of resources to spend on specialised hardware with a singular purpose, and access to cheap power.

This usually limited the participation pool to professional mining organizations. This phenomenon has led to the emergence of ‘mining tyrants’, leading to a higher and higher entry barrier for those looking to become a miner, which effectively blocks ordinary users from being involved. This, we believe, is not in line with what blockchain networks were meant to be, and thus, we’ve decided to take a stance against it.

The threat of centralization caused by a mining monopoly

Last year, the Bitcoin Cash (BCH) Hard Fork battle led to the price of Bitcoin falling to below $6,000 USD overnight, and once to nearly $3,000. This unpredictable price drop added to the cryptocurrency markets instability and volatility.

The reason that these fluctuations occurred was that during the hard fork, a battle was being waged by two major miners. In this battle, the computing power that was used to maintain normal operations of the BCH blockchain (via mining pools) were being used for other purposes, thus partially compromising the security of the chain, backlogging the validation and processing of transactions due to the now lack of computational power, and, ultimately sowing the seeds of doubt into users minds regarding the stability, longevity and independence of the network.

The most questionable thing that has occurred during this phenomenon is the behaviour of the so called “mining tyrants”, whose actions negatively affect the common interests of all blockchain users. Unless dealt with, blockchain will be no different than any other centralised mechanism.

Blockchain has always been viewed as a ‘decentralised’ technology. It’s a technology that was marketed as being more independent, fair and democratic. This technology has a large potential to change the world, especially considering its advancements in ledger technology and widespread appeal. However, if computing power becomes monopolised and ordinary users lose their ability to participate, then it begs the question: Is this new system any better than the last centralised system we moved from? Obviously, the potential of mining monopolies casts a shadow on the entire blockchain industry, and it will inevitably chip away at fundamentalist beliefs of the technology, thus, it must be addressed.

The rise of Blockchain monopolies heavily affect network security

Monopolies that may form within a network such as BTC, ETH and others not only have the effect of causing centralisation, but also damage the stability and security of the network.

These monopolies, that often take the form of mining pools sometimes employ dubious methods to increase their block rate higher than their computing power ratio. These methods include using ASIC boost, selfish mining, eclipse attacks and other means that enable these pools to conduct a 51% attack without needing to own 51% of the computational power on the network. An example can be seen on the XVG blockchain, where a mining tyrant achieved a 51% attack while only owning 10% of the total computational power on the network.

ASIC-resistant mining machines and technological progression

With what’s already been said above it seems fairly clear that ASIC-resistance should be something that all networks should strive to implement, however, there are some objections to these resistant technologies.

Some people consider ASIC machines as part of the natural course of progression relating to advancements in science and technology. Others believe that the advent of ASIC mining machines, which can mine faster and cheaper serve a vital purpose of reducing network costs. The fear within some communities is that as networks move away from mining machines and those machines spawn from technological progression, that means that the networks themselves are shying away from that same technological progression; they’re trying to live against technology.

In this regard, we must clarify that ASIC hardware is not technically considered to be a natural step in scientific and/or technological progression. Technological progression would imply that things on a network level are becoming more efficient, but they aren’t. If that were true then the network would be using less electricity, which isn’t the case when it comes to bitcoin for example. When Bitcoin first came out in 2009, Nakamoto used his own home computer to mine 7,200 BTC while using only several kWh’s of electricity per day. Now, tens of thousands of high-performance machines (including ASIC machines) around the world mine around 1,800 BTC while consuming millions of kWh’s of electricity per day. This hardware competition that is ultimately fuelled by ASIC mining machines has led to us consuming more energy than necessary. What’s more is that Bitcoin’s block rate is constant, and the blockchain network throughput is more dependent on the speed of the network, rather than the hardware. Thus, we don’t believe that implementing ASIC resistance means to rally against progression, rather, the opposite.

ASIC mining machines have increased barriers to entry, which is not conducive to the popularisation and development of blockchain technology

Previously, ordinary, household individuals were able to participate in the blockchain industry via mining if they had a computer to mine with, even if they didn’t know much about cryptocurrency or couldn’t afford an ASIC mining machine. What’s more, ASIC mining machines, whose target clients are large mining pools and large net worth individuals, are currently monopolised by a limited number of manufacturers.

The value of blockchain itself comes from consensus, and consensus comes from the participation of more and more people. The higher the participation rate, the more popular blockchain will be, and the more popular that blockchain is, the more stable and healthy the price of the cryptocurrency. In other words, the more decentralised the environment, the more fair and secure the industry.

It can then be derived that the advantages of high speed mining machines do not in-fact aid the network, but merely line the pockets of their owners with more tokens, as the block rate is unaffected. So, ASIC mining machines do not actually promote the development of a network from a technical point of view; on the contrary, it hinders the participation of ordinary people.

Thus, in order to prevent against problems caused by monopolies, many new blockchain projects have abandoned the PoW algorithm and have turned to PoS and DPoS algorithms, such as seen in EOS. However, the governance of EOS super nodes have also been questioned. For example, many people think that their super nodes and cloud servers makes it almost impossible for most ordinary users to participate. If blockchains can be centralised, are they still considered as blockchains in the most fundamental sense?

The HC team proposed in our yellow paper that the PoW algorithm has been recognised as the safest and most fair consensus algorithm after ten years of community testing, and as such, it should not be abandoned fully due to the potential of monopolies appearing. Thus, as a compromise, the HC team adopted a PoW+PoS hybrid consensus in its mainchain to maintain security, with the additional features and benefits of the PoS consensus mechanism.

The combination of the two creates an excellent consensus algorithm for fostering strengths and avoiding shortcomings. The HC team has been preparing solutions for hidden dangers such as mining tyrants and centralised problems by developing and deploying ASIC-resistant algorithms.

In conclusion, ASIC-resistant algorithms are necessary to increase community participation and secure network security. The ultimate goal of ASIC-resistant algorithms for each blockchain project, including HC, is not to resist ASIC mining machines, but to reduce barriers to entry, and to create a distributed PoW ecosystem that is as fair as accessible.

To stay up to date with information, or to join our community, check out the following channels:

Twitter, Facebook, Reddit, English Telegram, Chinese Telegram, Korean Telegram, Korean Announcements, KakaoTalk, Naver Blog

Written by

The New Standard of Value

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store