Interest in cryptocurrencies is currently at fever pitch with banks, businesses, and governments racing to understand the technology and how they can exploit it. As a result, the cryptocurrency market has begun to rise exponentially, and last month reached an astonishing $378 billion in market capitalization. Whatever happens next, cryptocurrencies look certain to play an increasingly influential role in the global financial system. In today’s world, most of the blockchain coin prices are driven by scarcity model and hype which is not based on real world use case. None of the blockchain is capable enough to meet the demand of enterprise and governments. They need high transaction rate and security against future thread like quantum attack. Most of the banks are still keeping themselves away from existing blockchains because of high transaction fees, partial decentralism and future quantum attacks.
TIM is a two layered blockchain-
-First layer is root blockchain which only stores Merkle hash of second layer. First layer is very light weight (approx. 4 MB data size).
-Second layer (Multigraph) is the one which stores all the transaction data in form of blocks. First layer only references the blocks of second layer. Second layer blocks are partitioned using vertical shard.
Each node (mobile device / computer running TIM core engine) in TIM network contains entire root blockchain (first layer) and only a part of the second layer blocks. Assuming there are 10,000 nodes in a network, TIM first creates cluster of nodes based on geolocation using partition technology. So the nodes which belongs to ASIA will be part of ASIA cluster and nodes which belongs to North America will belong to North America cluster. TIM call this clusters as triangles / graphs.
The transaction which is originating in North America will participate in triangle of North America only. All blocks mined in North America will be stored in each of the nodes which belong to North America region only. However, all the nodes of North America will have complete root blockchain present but only part of second layer blocks which is mined in North America. Similarly, each Node of Asia Cluster (triangle) will contain entire root blockchain of TIM but only part of second layer blocks which is mined in ASIA. This is called as vertical partition of blocks.
If the data size of entire TIM blockchain is 1 GB, then it will split among these two regions (approx. 500 MB each). So, each node of North America will store root blockchain (4 MB size) and approx. 500 MB of block data. However, in ideal scenario of 1 GB data, TIM will have many triangles (approx. 100). And data will be partitioned (using dynamic vertical partition based of geolocation) to approx. 10 MB block data. So, each node in a given triangle will have 4 MB of root blockchain and 10 MB of partitioned second layer blocks data. This will enable even a mobile device with less memory to become a node of TIM blockchain.
Triangles (cluster of Nodes) / Multigraph – Each node in the triangles communicate to each other using DAG (Directed Acyclic graph) topology. Hence, we call these triangles as graphs also. And collection of these triangles is called as Multigraph (which is the second layer of TIM blockchain). Multigraph stores entire blocks split across graph/triangles. So, if we have 100 triangles, Network will look like as below.
It is because of the DAG topology of each triangle, nodes can communicate very fast, achieving speed of 10,000 tx/sec per triangle. And if the 100 partitions are created due to the demand of transactions, then total network capacity will be around 1 million tx/sec.
However, these triangular regions are created by TIM technology on demand of transaction load. If the load increases in a given region, the TIM technology can further split triangle into two parts and there by doubling the speed of that region. TIM network auto adjusts the load based on demand and also merges the region if the demand is less. This is driven by statistical model for load balancing and using geolocation of nodes to understand the transaction load in a given geographical area.