Bytom — On a Bitmain Leash


TL;DR. Ecology:2. Technology:3. Decentralization:2. Valuation:3. Rating:2/10


(Update 2019-02-08: Please see the semi-annual ratings review of Bytom here.)

Bytom was announced by user hl5460 on 20th of June 2017 to the English speaking part of the global cryptocurrency community. The main stated focus of the project, according to and Bytom founder Chang Jia, is to connect the digital world with the physical by featuring for example digital and physical asset registration. This is allegedly possible through the use of standardized smart contracts.

In August 2017, the Bytom team - itself located in Hangzhou - set up a non-profit foundation in Singapore. Later that year the price of the BTM token tumbled after Peoples' Bank of China ordered multiple exchanges to delist Bytom (among other coins) and to refund investors. Bytom developers continued to code however, and soon a testnet beta client written in Go was released, with a corresponding bug bounty program.

The BTM token was initially issued on Ethereum under the ERC20 standard so that investors could trade it freely before mainnet was released. Mainnet, arrived on 24th of April 2018, utilizing the newly developed consensus algorithm Tensority which is dedicated to Bitmain ASICs that are specialized for AI research computation. The algorithm seems to have been developed by Bytom and researchers from Southwest Jiaotong University, while a partnership with Bitmain provided the ASIC hardware. In the end of July 2018, smart contract functionality was released, as well as some useful templates.

According to the Economic Whitepaper, a total of 2.1 billion BTM will ever exist. 30% were distributed after the initial ICO. 33% were reserved for mining. 7% were allocated to Private Equity investors. 20% were allocated to the Bytom Foundation (allegedly linearly vested for 4 years) while the last 10% were reserved for Business Development. The fact that now states a total supply of 1 407 000 000 may have to do with the reimbursement to ICO investors, but it is not totally clear.


When researching Bytom, one is quickly met by the concept of AI-ASIC mining. BD Ratings understands this to be clever marketing trick in order to throw the Artificial Intelligence buzzword in to the mix. ASIC mining farms will mine Bytom, but they should according to the Bytom team also be able to allocate resources to calculations that help advance AI research. BD Ratings' conclusion may seem cynical but it is only natural in this space. Bytom seems to also publish PR-articles as 'news', where focus is on the blockchain-AI connection as well.

The team is good at communicating updates regarding the network as a whole. They publish weekly reports on and link to them on the subreddit which is just sporadically active. Sadly, the team lacks professionalism in other areas as shown here where the official user pushes people to invest and ends with a tragic 'too the moon' comment. In another post, this official BCT handle doesn't even understand the concept of swapping ERC20 tokens to Bytom mainnet tokens. Looking through the official subreddit, there are instances where official team members encourages BTM holders to participate in a Chandler Gou led Bitcoin fork called Bitcoin God (not to be confused with Bitcoin Gold). This is further proof of a low level of professionalism, as some users may very well make mistakes and hurt their financial position.

Bytom has multiple times offered people money to follow and retweet the official Twitter handle. This is a hurtful practice that encourages spam, fake accounts and short sighted bought loyalty. BD Ratings also strongly discourages buy-to-win competitions like this one where those who buy most Bytom tokens win prices. It is not entirely clear that this competition was initiated by the official team, but no team member spoke up against it.

As mainnet was launched, it became evident that that the first few thousand blocks were mined by insiders, meaning by either the team or Bitmain who manufactured the ASICs. On top of this, Bitmain also earned a considerable amount of money from selling these ASICs that was not delivered until after mainnet launch. Lastly, as the coin now has launched, the number of on-chain transactions per day can be measured. At time of writing it is just a couple of hundreds.

Grade: 2

Reasons: Unprofessional communication. Low visible community activity. Low on-chain activity. Scammy promotion offers.


One of the biggest issues with Bytom according to BD Ratings is the development and implementation of a new consensus algorithm designed for AI ASIC hardware. Most of these concerns have to do with centralization, but there are some technical ones as well. Some further details regarding this new algorithm is outlined here, and on the whole it seems there is no discussion about what can go wrong. Bitcoin's mining process is extremely simple and can be summarized as iterative SHA256 hashing. Bytom, with the Tensority whitepaper, claims to have achieved the same level of fairness in the sense that iteration outcomes are random yet has AI research utility. More peer review of this is needed and BD Ratings would like to hear what other cryptographers in the space have to say. Intuitively any consensus algorithm dedicated to a process that today can advance AI research, may find itself useless in the future if another type of computing is needed.

One interesting aspect of Bytom's choice to adapt to AI ASICs is the fact that they also have brought up China's massive surveillance program and its thirst for pattern seeking facial recognition algorithms that could be further improved with AI. In other word, Bytom is doing its best to advance this totalitarian dystopia by trying to tie advancement of AI research to its block rewards.

Bytom applies technical bug bounties, as seen in this testnet beta launch post. This is a great way of using pre-allocated tokens in order to strengthen the quality of the network.

When inspecting the GitHub repository, we can conclude that there is continuous activity there with multiple contributors. According to official information, there were 12 core developers in the beginning of 2018.

Grade: 3

Reasons: Technical bug bounties. Active GitHub with multiple developers. New, untested consensus algorithm that is seemingly constructed for a purpose other than just securing the blockchain.


In an interview, team member Duan Xinxing indicates that they will conduct a paid due diligence investigation of all projects listed on Bytom. What this means in practice remains to be seen, but an important aspect of a public, decentralized blockchain is its permissionlessness. Anyone must be able to issue assets on the blockchain. It could be that they just mean optional due diligence from an investor viewpoint.

One worrying aspect of the consensus algorithm is that it is modelled to fit Bitmain ASIC miners. What this results in is an imminent threat of control of the chain by Bitmain, which was evident as mainnet launched and a majority of block rewards went to one single Bytom address. Parts of the community were concerned about this centralization, but the Bytom team did not want to admit any responsibility. When launching with a chain that utilizes an ASIC friendly consensus algorithm, it is obviously a good idea to think about how the ASIC manufacturer behaves and if ASICs had been distributed in time or not. The fact that Bitmain was a part of the Bytom blockchain construction efforts is just an extremely centralizing aspect.

Since the blockchain runs on Proof of Work and not some Proof of Stake variant, it is not a threat when the team allocated a quite substantial number of BTM tokens to themselves or the foundation during the ICO process. The tokens can not directly be used to threaten the integrity of the blockchain, as would be the case with some of the common PoS mechanism.

And lastly, in the Economic Whitepaper a three-pillar governance structure is discussed, with focus on coin holders. This effort may very well turn out to decentralize the blockchain in the future. BD Ratings look forward to see it works in practice, with new voting infrastructure, with ousting of bad committee members etc.

Grade: 2

Reasons: Consensus algorithm created to narrowly fit Bitmain ASICs. Untested governance structure.


Going with CMC numbers with regards to maximum supply, all Bytom tokens are valued at around USD 318M. This is too high for various reasons: Despite what is possibly a technically strong team (with reservations to the consensus algorithm), there are still too much unprofessional conduct as outlined above. On top of this, the blockchain is probably very centralized by now.

Grade: 3

Reasons: Low community activity. Scammy team practices. Centralization.