Topic: [ANN][XPU] The first stateless sharded blockchain | Instant Sync | PoW | CPU/GPU - page 5. (Read 2550 times)
December 14, 2022, 05:42:36 PM
Yes, Criptonight could be a good solution, too
December 14, 2022, 09:05:46 AM
When we can mine it?
December 13, 2022, 12:03:42 PM
CryptoNight should not be forgotten 
December 13, 2022, 08:31:31 AM
Update: I have just published the source code along with the technical roadmap. As such, the project is officially open to external contributors.
These can both be found on our Github page:
These can both be found on our Github page:
October 15, 2022, 03:49:10 PM
Update: Thanks to @noventrun#2672 from our Discord channel, we have chosen to change Whirlpool and Groestl to Fugue and JH as the former are vulnerable to Rebound attacks.
October 13, 2022, 01:44:15 PM
14222 transactions per second, that's more than half of VISA's TPS. Has this been tested and confirmed? Or is it just a prediction of what might be?
It is based on the calculations we have on the chosen parameters such as block size and shard numbers. We have indeed tested the architecture, but not on a live network. While the number can vary as custom transactions can be bigger, it gives a rough estimate when the bulk of transactions are the default format, which from what we can see with other networks seems to be the case 98% of the time.
Block size is actually smaller than Bitcoin, at 750kb. Sharding multiplies that by the number of shards which stands at 256. Nodes can listen to all shards or individual shards, depending on bandwidth. A full-node is also able to run as a cluster, with individual nodes listening to individual shards, this is useful for archival nodes, being able to scale bandwidth and storage. SPV nodes are able to sync all shards a few times per day, without requiring to download the whole backlog of transactions.
October 13, 2022, 01:29:53 PM
14222 transactions per second, that's more than half of VISA's TPS. Has this been tested and confirmed? Or is it just a prediction of what might be?
October 13, 2022, 01:06:09 PM
Purplecoin: The first stateless sharded cryptocurrency and blockchain
With the advent of Bitcoin, decentralised ledgers have sprung into existence. While revolutionary for its time, Bitcoin could not satisfy the needs of the masses due to its inability to scale to global needs. No other blockchain protocol has managed to achieve this, at least not without sacrificing the one thing that made Bitcoin as powerful as it is: being tamper-proof, even against state actors. We believe we have managed to achieve what Bitcoin couldn't. We will now present the first stateless cryptocurrency.
What is a stateless Blockchain/Ledger/Cryptocurrency?
To put it simply, a stateless blockchain allows full-nodes to validate transactions without storing past blocks or unspent transaction outputs. This is done by employing an incredible cryptographic construct known as a cryptographic accumulator. While Bitcoin can theoretically scale by increasing the block size, it would eliminate the ability of regular folks to run a node. This improvement basically allows big blocks sizes while still allowing regular folks to run a node. This also means several things:
Scalability through sharded Proof of Work
There are 16 chains, called sector chains, each representing a group of 16 shards, with a total of 256 shards. Each shard is an independent blockchain and can be processed separately along with the respective sector chain. Transactions are only possible within the same shard. The XPU token is mined therefore on different, independently running chains. As such the tps is equal to the sum of the tps of all chains. Our calculations at a block size of 380kb resulted in 14222 transactions per second assuming they are simple transfers and not complex transactions. As Purplecoin is stateless, the storage size of full-nodes doesn't increase and only relevant assets are kept.
More details about the consensus algorithm: https://github.com/purpleprotocol/purplecoin/blob/main/doc/consensus.md
Mining algorithm which consumes 50% less energy
We are using the Green PoW consensus model which elapses in epochs. Each epoch has 2 rounds. In the first round, every miner participates and has the potential to mine a block. The miner who finds the solution in the first round is not entitled to participate in the second round. Then, the rest of the miners continue mining the block in the first round. When at least 15 miners manage to mine the block, the second round begins. These miners are called runner-ups. Only runner-ups are entitled to mine the block in the second round. Once one of the runner-ups mines the second round block, the next epoch begins. This mechanism brings energy efficiency to mining (up to 50% compared to if it wasn't employed) and secures shard sectors against 51% attacks. More details around this can be found in the Whitepaper.
Now, every epoch will use a different algorithm. Half of the epochs will be mined exclusively with GhostRider, in order to prevent ASICs from mining a longer chain while allowing CPU/GPU mining. The other half of the epochs will be mined with an ASIC friendly algorithm, chosen randomly every epoch. This is in order to balance GhostRider, as fully ASIC resistant algorithms are subject to hashpower marketplace attacks, while ASIC algorithms are less likely to be subject to these attacks.
Example:
The following are the current proposed ASIC friendly algorithms:
Find out more about mining: https://github.com/purpleprotocol/purplecoin/blob/main/doc/mining.md
Tokenomics and crowdsale
Find more info here:
Crowdsale info: https://github.com/purpleprotocol/purplecoin/blob/main/doc/crowdsale.md
Tokenomics: https://github.com/purpleprotocol/purplecoin/blob/main/doc/tokenomics.md
Links
With the advent of Bitcoin, decentralised ledgers have sprung into existence. While revolutionary for its time, Bitcoin could not satisfy the needs of the masses due to its inability to scale to global needs. No other blockchain protocol has managed to achieve this, at least not without sacrificing the one thing that made Bitcoin as powerful as it is: being tamper-proof, even against state actors. We believe we have managed to achieve what Bitcoin couldn't. We will now present the first stateless cryptocurrency.
What is a stateless Blockchain/Ledger/Cryptocurrency?
To put it simply, a stateless blockchain allows full-nodes to validate transactions without storing past blocks or unspent transaction outputs. This is done by employing an incredible cryptographic construct known as a cryptographic accumulator. While Bitcoin can theoretically scale by increasing the block size, it would eliminate the ability of regular folks to run a node. This improvement basically allows big blocks sizes while still allowing regular folks to run a node. This also means several things:
- The initial sync to the network is more or less instant, regardless of the chain size. Further syncs are an order of magnitude faster, taking possibly seconds at most. We still have to test this but it seems really fast.
- Transaction validation first and foremost becomes more efficient because we don't need to search the disk for UTXOs as we only check the accumulator, which lives in memory. Memory is orders of magnitude faster than disk, which is the main bottleneck in Bitcoin.
- Users are responsible for storing their own UTXOs in their wallet. Besides these, a full validating node can run while storing only a few kilobytes of data. This means a full-node does not in fact need a disk to run at all.
Scalability through sharded Proof of Work
There are 16 chains, called sector chains, each representing a group of 16 shards, with a total of 256 shards. Each shard is an independent blockchain and can be processed separately along with the respective sector chain. Transactions are only possible within the same shard. The XPU token is mined therefore on different, independently running chains. As such the tps is equal to the sum of the tps of all chains. Our calculations at a block size of 380kb resulted in 14222 transactions per second assuming they are simple transfers and not complex transactions. As Purplecoin is stateless, the storage size of full-nodes doesn't increase and only relevant assets are kept.
More details about the consensus algorithm: https://github.com/purpleprotocol/purplecoin/blob/main/doc/consensus.md
Mining algorithm which consumes 50% less energy
We are using the Green PoW consensus model which elapses in epochs. Each epoch has 2 rounds. In the first round, every miner participates and has the potential to mine a block. The miner who finds the solution in the first round is not entitled to participate in the second round. Then, the rest of the miners continue mining the block in the first round. When at least 15 miners manage to mine the block, the second round begins. These miners are called runner-ups. Only runner-ups are entitled to mine the block in the second round. Once one of the runner-ups mines the second round block, the next epoch begins. This mechanism brings energy efficiency to mining (up to 50% compared to if it wasn't employed) and secures shard sectors against 51% attacks. More details around this can be found in the Whitepaper.
Now, every epoch will use a different algorithm. Half of the epochs will be mined exclusively with GhostRider, in order to prevent ASICs from mining a longer chain while allowing CPU/GPU mining. The other half of the epochs will be mined with an ASIC friendly algorithm, chosen randomly every epoch. This is in order to balance GhostRider, as fully ASIC resistant algorithms are subject to hashpower marketplace attacks, while ASIC algorithms are less likely to be subject to these attacks.
Example:
- Epoch 1: GhostRider
- Epoch 2: Sha256
- Epoch 3: GhostRider
- Epoch 4: Fugue
The following are the current proposed ASIC friendly algorithms:
- Blake2s256
- Blake2b512
- Sha256
- Keccak256
- JH
- Fugue
Find out more about mining: https://github.com/purpleprotocol/purplecoin/blob/main/doc/mining.md
Tokenomics and crowdsale
Find more info here:
Crowdsale info: https://github.com/purpleprotocol/purplecoin/blob/main/doc/crowdsale.md
Tokenomics: https://github.com/purpleprotocol/purplecoin/blob/main/doc/tokenomics.md
Links
- Whitepaper: https://github.com/purpleprotocol/purplecoin_whitepaper/blob/main/whitepaper.pdf
- Vision: https://github.com/purpleprotocol/purplecoin/blob/main/doc/vision.md
- Crowdsale website: https://sale.purplecoin.io
- README: https://github.com/purpleprotocol/purplecoin/blob/main/README.md
- Technical Roadmap: https://github.com/purpleprotocol/purplecoin-core/blob/main/ROADMAP.md
- Contributing: https://github.com/purpleprotocol/purplecoin/blob/main/doc/contributing.md
- Website: https://purplecoin.io
- Sale: https://sale.purplecoin.io
- Github: https://github.com/purpleprotocol
- Discord: https://discord.gg/5ZVZnKd
- Reddit: https://www.reddit.com/r/purplecoin
- Twitter: https://twitter.com/xpu_official
- Medium: https://medium.com/@octavianonce
- Telegram: https://t.me/purplecoin_io
- Merch: https://merch.purplecoin.io/
Jump to: