HPB — Update II


TL;DR. Ecology:3 (3). Technology:3 (4). Decentralization:3 (3). Valuation:7 (8). Rating:4/10 (5/10)

Six months have passed since the second BD Ratings article on HPB was published. Here is how the prior predictions have turned out.

Both are clear misses; BD Ratings overestimated the fundamentals. Let's go through the important events within the HPB ecosystem for the last 6 months to figure out where to go from here.

As of November 2018, the three parameters that allegedly affected a candidate node's chance of becoming a High-Performance node was bandwidth (50% weight), number of HPB coins (15% weight), and ranking of community voting (35%). This is a slightly slimmer approach from what was suggested earlier. The end of November also saw another round of node applications, followed by the free distribution of BOE hardware unit boards by parcel on December 3rd.

February 2019 saw some clarifications on the issuance schedule. The adopted model still seemed to aim for a dis-inflationary schedule where up to 6M tokens annually were released to validating nodes. A coin burn of 3M tokens from the reserve fund was put up for a burn plan, meaning effective inflation for the first year dropped. In March, the third node application process closed, with further hardware distribution as a result.

The year continued with various partnerships and other collaborations. On the 28th of April, a mainnet upgrade was announced by the team. The new version fixed a number of bugs and included different optimizations for many parts of the network infrastructure, like transactions, smart contracts and block routing.


The HPB subreddit is now almost completely dead, and most posts seems to come from team members. On-chain transactions are sparse as well.

It is evident that the project still conducts stupid kinds of marketing. This will of course only attract a certain kind of short sighted investors that already cares more about marketing than technology. Smart contract functionality is live on mainnet since long ago, but there are still not many entities building on HPB.

Grade: 3

Reasons: Bad marketing moves. Low on-chain activity as well as community activity.


The HPB GitHub repositories are continuously updated by a number of developers. A security audit by KnownSec.com completed on 1st of April, resulting in an acceptable verdict. Mainnet has more or less been confirmed to be able to process thousands of transactions per second, which of course is promising from a technical standpoint given that it consists of over a hundred nodes.

But there are still problems with how HPB is constructed. To have factors such as network bandwidth greatly influence the probability of node selection, again, seems complicated enough to be a gameable factor. If there is any chance of nodes being able to fake or temporarily adjust bandwidth to increase election chances, they will do this in the long run. BD Ratings is convinced that these kind of things must be kept simple.

It also appears that the election of candidate nodes is not incentivized in any way. This will obviously lead to voter apathy in the long run, resulting in a smaller hurdle to try to control candidate nodes and so the consensus process. It shows a rather weak understanding of what place incentives have in a blockchain network.

Grade: 3

Reasons: Still complicated consensus mechanism. Weird lack of incentives to vote for candidate nodes. Promising transaction throughput capacity.


The HPB team has shown indications of looking down a decentralization road going forward. The BOE hardware has been given away for free to hundreds of node applicants around the globe, which helps take away some power from the team itself. Another decentralization step is the upcoming expansion of the total number of BOE (candidate) nodes, from 150 to 300, and validating HP-nodes from 31 to 61.

But in an apparent lack of understanding what constitutes centralization, HPB founder Wang Xiaoming pondered the patent question in a recent AMA. He did not rule out trying to patent BOE components. He should ask himself why no core aspect of Bitcoin is patented and why bitcoiners still need not worry about projects copying code (which they do). The only thing patenting does is consolidating power to a certain group - something that any blockchain aspiring to be decentralized should try to keep at bay.

Grade: 3

Reasons: The project does still not value decentralization.


At time of writing, all circulating supply of HPB amounts to a USD 16M valuation. Considering total supply, this number grows to around USD 40M. The USD 16M valuation places HPB as #212 on CMC. All this means that valuation has fallen, but so has the fundamentals. It is sad that a quite promising project decided to mess things up, and BD Ratings can't do other than change the rating from 'hold' to 'sell'.

Grade: 7

Reasons: Fall in token price. Worse fundamentals.