We have been talking about it for a few months now: the technology described in the Burst Dymaxion white paper is going to be implemented through three consecutive hard forks. The first one, recently described by the PoC Consortium in a Reddit post, will enable PoC2, dynamic fees, more transactions per block and multi-out transactions in June. These changes will bring not only a huge raise in transactional capacity but also an incredible flexibility/dynamism to the Burst blockchain.

An increase in block space and efficiency

Each individual Burst transaction takes 176 bytes in a block. In its current state, the blockchain can support 255 transactions per block (44,880 bytes).

After the hard fork, each block will have a payload of 1020 * 176 bytes (for a total of 179,520 bytes). However, we can cram up to 80 times more transactions into these 4 times bigger blocks.

An elastical response to transactional load

If there is no load, the blockchain does not waste any space. The data simply isn’t there and the tx-payload in the block is 0 bytes.

This was true before and it is still true. We do not buy our bigger capacity with higher “baseline cost”.

If Burst was a boat: a realistic before/after comparison

How do we do that?

We are introducing Multi-Out transactions, allowing to send burstcoins to N recipients in only one operation.

64 ordinary payments take 64 * 176 bytes = 11264 bytes. If pools make their payout to miners, they nowadays have to issue e.g. 64 of these transactions, pay 64 Burst transaction fees and occupy 1/4th of the block for this.

With the new Multi-Out transactions, a pool can simply issue one transaction (176 bytes) and name up to 64 recipients and the amount they shall get (8 byte recipient-ID + 8 byte amount = 16 byte per recipient). A Multi-Out to 64 recipients takes 176 + 64 * 16 = 1200 bytes. So it costs 1/64th transaction fees, and takes up roughly 1/150th of the new block.

This payment is not exclusively for pools. Assume you would like to send money from your account to Bob and Alice. You could do 2 ordinary payments (2 * 176 byte = 352 byte) and pay 2 transaction fees or you could issue a Multi-Out to Bob and Alice at once (176 + 32 byte = 208 byte) and pay just one transaction fee. The more recipients you have, the more efficient a Multi-Out is for you (less tx fees) and the blockchain (less space taken).

With the raised block size and these Multi-Out transactions, we can effectively have 9600 transactions in a block – 40 transactions per second. A 40-fold to now.

We can do even better

Another type of Multi-Out transactions exists: the Multi-Out Same transactions. Instead of sending individual amounts to 64 recipients, it is also possible to send the same burstcoin amount to up to 128 recipients and take up the same space in the blockchain (176 bytes + 8 bytes per recipient). This is especially useful for faucets.

Together with the dynamic fee, after the hard fork a faucet can serve 128 people in one single transaction, can send each one 0.1 BURST and spend on that transaction itself 0.1 BURST. Total cost: 12.9 BURST, taking up 1/150th of the block, and the recipients can comfortably perform 3-4 transactions with that 0.1 Burst they got.

Right now, to serve 128 people, it would cost 256 BURST, would take 1/2 of the block and the recipients can do only 1 transaction. The faucets today would have to pay 19.76 times as much, would take 75 times as much space in the block and achieve only a result per recipient that is 0.25 times as good as the new system.

With this type of transaction, the theoretical limit becomes 19,200 transactions per block – 80 transactions per second. A 80-fold to now!

Nowadays, Burst can do a maximum of 91,800 transactions a day. After the upcoming HF, Burst will be able to perform at least 367,200 and a maximum of 6,912,000 on-chain transactions a day.

For comparison, Bitcoin can do a maximum of 604 800 transactions a day.

This also means our energy efficiency goes up by the same factor. After the hard fork, Burst will be a lot more energy efficient than it is now, and incredibly – by a factor of tens of thousands – more efficient than Bitcoin and other PoW cryptocurrencies.

Paving the way for the Dymaxion

The transactional capacity we described here happens only on the Burst blockchain. The Dymaxion will bring the real scalability that will make Burst a cryptocurrency for truly global use.

But this first hard fork will also improve the potential of the Dymaxion: since opening and closing a Dymaxion layer will be made through an Atomic Cross-Tangle Transaction (ACTT) taking some place in the Burst blockchain, improving its transactional capacity also improves the numbers of tangles that can be opened and closed.

When the Dymaxion will be enabled with the second hard fork, we are going to be able to open and close hundred of thousands of tangles per day – representing as much customizable and arbitrarily scalable payment channels for instant and fee-less transactions.

This infographic resumes the important transaction changes that the first hard fork will bring. Feel free to share it and use it to your liking.

Also published on Medium.



Subscribe now to our newsletter


    1. Hi Lee, you can check out the pre-Dymaxion hard fork checklist: https://burstwiki.org/wiki/500k_HF_Checklist

      Please use the latest BRS 2.2.0 wallet. Exchanges have been given the newest version, so you can keep your Burst there – even if it’s always preferable to withdraw them to your local wallet for security reasons.

  1. Hello
    May I know the date when the next #Burst hardfork would be happening…
    Big fan and enthusiasts. Your reply would really be appreciated.

    Thank you

Leave a Reply

Your email address will not be published. Required fields are marked *