Algorand, i.e. a platform for smart-contracts, based on the Pure Proof of Stake network security mechanism, enables the processing of 1k transactions per second. It would seem that these values are sufficient, many times exceeding the possibilities that at the moment are offered by, for example, Bitcoin or Ethereum. At the same time, the speed of the progressive adoption of the Algorand solution means that in the coming year the current Algorand network capacity may turn out to be insufficient. Let's take a look at how the Algorand network is being used today and what are the plans for the coming years.
Perhaps it is best to judge the level of viable Algorand adoption by looking at AlgoExplorer, which gives an overview of the main parameters of the Algorand platform and all transactions taking place on this network.
At the time of writing this article, Algorand is processing approx 6.8 real transactions per second,or about 590 thousand. transactions per day, which is 17M transactions per month. We are talking about the actual bandwidth of this network, which supports business transactions, from applications such as Props or PlanetWatch (I discuss them on the YouTube channel).
A year ago - Algorand processed only 100 transactions every month, which now gives 170x increase compared to 17M monthly transactions for December 2020. This is quite a gigantic increase, but what is most fascinating, the majority of these transactions are transactions generated by only 2 business applications.Consider that Silvio Micali announces that currently over 500 different projects are building their solutions based on Algorand, and some of them will see the light of day next year. So if 2 applications can generate such significant traffic on the Algorand network, it is shivering when we think about the number of not two, but five hundred applications running on Algorand in the coming time.
It is best to follow the level of adoption yourself (in line with the "don't trust, verify" assumption) and keep track of who and what transactions have already been admitted on the Algorand blockchain.
This is also seen in AlgoExploera.
To prepare for the massive influx of new applications (more on that in a moment), Algorand is planning a network upgrade in 2021 that will improve the main parameters of this decentralized network, namely:
- Time to propose the next block. This is the time it takes to build a block and wrap the current transactions into it. Such a block becomes a candidate to be added to the blockchain.
- Block finalization time.This is the time needed for the block to be validated by nodes in the network, and then permanently (irrevocably) adding such a block (and its transactions) to the block chain.
- Transactions processed per second (TPS). That is the average number of transactions that have been irrevocably recorded on the Algorand blockchain.
Of course, this is a rather simplified look at the topic, so I encourage you to read the details presented in the following infographic, or hearing my moviein which I describe in more detail the mechanisms of the Algorand network.
Infographics - consensus mechanism in the Algorand network
Currently, the parameters related to the speed of the Algorand network are:
- block proposed time = 0.5 sec
- PLAN for 2021:no change = 0.5 sec
- block finalization time: = 4.5 seconds
- PLAN for 2021 = 2.5 seconds
- number of transactions per second = 1000 TPS
- PLAN for 2021= 46,000 TPS
I propose to read the document by prof. Silvio Micali (founder of Algorand) who describes in detail Algorand's planned approach to changes in the network planned for 2021.
Before we move on to discussing Algorand's network bandwidth extensions, I would like to mention the main features of this platform that Algorand is instically developing. Over the past year, the team has made several major extensions to the platform, these extensions include:
- Smart Contracts on layer-1, i.e. a very safe, fast and at the same time cheap approach to the smart-contracts concept. More on this I am writing here.
- Algorand Standard Assets, i.e. the ability to generate tokens that are built on the basis of the Algorand protocol, being built into layer-1, and benefit from the security and speed of this layer. More on that here.
- Atomic Transfers, that is, an integral, atomic approach to transactions. The ability to transfer ownership of many assets, tokens, at once, in such a way that either all of them are transferred or none. I wrote about what these functions are and what opportunities they open up, among others In this article.
- TEAL language, the programming language that the Algorand team designed and built made it the native language for their platform. The language is incomplete in Turing's understanding, and this is his forte. Why? Here I am giving a number of arguments. Algorand is also working with the Blockstack team on another language (Clarity) that will debut on both platforms in 2021, providing a secure environment for creating predictable smart contracts.
- Algorand Co-Chains, that is, private chains based on the Algorand protocol that can simultaneously communicate with other private chains as well as with the public Algorand chain. It is interoperability inside the Algorand ecosystem. Attention! that's not all, we will also soon be interoperable with platforms that have a different consensus protocol, such as Ethereum. Algorand is working on its approach to bridging networks based on different consensus (similar to Polkadot or Cosmos).
In mid-2020, Algorand additionally extended the platform with:
- ReKeying, i.e. a functionality that allows users to replace the private key without having to change the public address. Changing the key provides greater flexibility in the use of the keys and opens the door to use cases such as mass migration of users from other platforms. Thanks to ReKeying, large, centralized platforms can migrate their users' accounts to the Algorand network, create accounts automatically on their behalf, and then give them the opportunity to take full control over such an account (by replacing the private key)
- Stateful Smart Contracts, i.e. a kind of smart contracts with global memory. So far, ASC1, or Algorand Smart Contracts, on the first layer, provided the benefits related to safety, speed and mega-low costs that layer-1 offers for Algorand (they are built into this layer). However, these are smart contracts with limited functionality. In August 2020 Algorand expanded smart contracts with the possibility of storing a global state, which can also be available from other smart contracts. This may sound quite mysterious (the following will tell you about it best film), but in general it greatly extends the possibilities of what can be built on the basis of the Algorand platform, importantly = the possibilities of what can be built in a decentralized way. In the previous version of the protocol, for example, it would not be possible to make a decentralized exchange, or a UniSwap liquidity pool, or even a voting system. Such a voting system requires that the results of voting are stored in an unmanipulated, transparent and undeniable manner - i.e. ON CHAIN. It is now possible and easily accessible.
Generally like this functional abundance is attracting more and more use cases to the Algorand network.
On the other hand, more and more applications based on Algorand make the need to change the TPS parameters (speed) and the time of transaction finalization inevitable.
So let's discuss briefly how Algorand wants to achieve this leap in rock.
- TPS is currently = 1k, TPS in 2021 = 46 k
- checkout is currently = 4.5sec, finalized in 2021 = 2.5sek
Algorand 46k / second
Algorand as we know it today, it is able to "pack" about 1 transactions into 5000 block. Since the block is finalized in 4.5 seconds, this gives us an approximate number of 1k transactions per second. Jump from 1k per second, to 46x more = 46k transactions per second brings Algorand closer to giants such as Visa.Check for yourself where Algorand would rank with such bandwidth in relation to otherplatform.
Okay, but what are the challenges to achieving such a goal?
- with more transactions packed in a block, the block size (measured in MB) also increases.
- increasing the block size makes the propagation of such a block slower across the network.
- increasing the number of transactions in a block makes consensus on block correctness more difficult among network securing nodes.
These are just some of the consequences that arise from pushing more transactions through the network. Such improvements are especially difficult to achieve in decentralized networks, where we have thousands of nodes that have to come to the only right agreement with each other on a regular basis which blocks / transactions are correct and which are not.
How to do it without losing the spirit of decentralization?
Algorand, when designing increased bandwidth, must therefore somehow cope with the above challenges. Details of how it will be done are described by Silvio Micaliin the attached documentI strongly encourage you to read it. The abbreviated interpretation of this document looks something like this:
- The block will contain 25k transactions vs the 5k transactions the block does today. So the number of transactions in the block increases 5 times.
- Blocks will be built every 0.5 seconds, processed in parallel, and finalized after 2.5 seconds.
- Thanks to this, after 2.5 seconds, we will have 5 blocks built, each of them will contain 25 transactions, so in 2.5 seconds we have 5 blocks and 125 transactions in the network. This is about 50 transactions per second (125k / 2.5 = 50)
- Block transactions will be compressed thanks to a smarter approach to saving data and thanks to new cryptographic tricks (let's not forget that Silvio Micali is a professor at MIT specializing in cryptography and winner of the Alan Turing Award for his achievements in this field).
- Algorand also assumes that by parallelizing the processing, some blocks will be incorrect (about 1%), so effectively processed transactions per second will be, less. Within 46 and not the above-mentioned 50 transactions per second.
The diagram below is only a small excerpt from what is happening on the Algorand network. There we see use cases in areas such as:
- Securities, specifically their tokenization,
- Decentralized Finance,
- Core infrastructure,
- Blockchain insurance,
- Digital Identity,
- Digital Assets of various types,
- Governmental and Public Sector Projects,
- Stablecoins of various types,
- Integration of Traditional Financial Institutions,
- Improvements in supply chains.
It's best to go directly to the website Algorand.com/use-caseand choose the use cases that interest you in the area you want to learn - because there is a lot of it.
Alternatively, on my YouTube channel I have about 40 videos on Algorand where I discuss different use cases and the technology behind Algorand itself. By clicking on the link below, you will be taken to a playlist about Algorand.
I discuss various business applications there, including perhaps the world's first CBDC creation (digital currency of central banks) - where, on the basis of the Republic of the Marshall Islands, I discuss what the implementation of the CBDC called SOV looks like.
From other sources, we also learn that SOV for Algorand in the context of CBDC is not everything => currently there are 16 different CBDCs that are experimenting in private, with the Algorand blockchain, so we, as the "audience" of this phenomenon, see only the tip of the iceberg.
Going back to the beginning of my argument - Algorand currently has over 500 different projectsthat build on this platform. On their website, you can see representative cases in different areas, but that's just 10% of everything that's built there. Considering the fact that Props and PlanetWatch, these two larger applications that generate traffic on the public Algorand network today, bring this platform to 17M transactions per month, I have the impression that in 2021 we will see a dramatic increase in adoption (new use cases entering the mainnet Algorand ) and possibly hundreds of millions of transactions a month .. who knows - maybe billions of transactions a month. Will the network stand it?
Algorand in 2021
Exactly - I think this is the problem most decentralized networks, which today are building their infrastructure and fighting for adoption, fighting to attract real businesses to their ecosystem, would like to have. I write this with particularly hard pressure on the keys of the letter
Most of the businesses that think about using blockchain came to this phenomenon rather cautiously. Usually, these were prototypes and test solutions that never surfaced, and, in principle, large organizations have not yet replaced their core business processes, which are now implemented on the basis of the existing IT infrastructure, by a decentralized counterpart. They did not migrate key processes to blockchain and did not overwrite the counterparts of business processes by implementations on a decentralized infrastructure, i.e. based on a public network and tokenization.
The question, of course, is whether they should do such a "migration" and what would they benefit from? I believe that the most fascinating use cases for this technology will be based on new business models that were not possible before without the participation of a decentralized public network based on cryptocurrencies.
There are many indications that in 2021 we can see the first bolder business applications announcing their achievements in the context of "blockchainization" ;-). In my eyes, Algorand is a definite leader in which this type of ecosystem is just maturing.
Of course, we do not want to have a situation where, for example, for Ethereum (limit of 20 transactions / second) or Bitcoin (limit of 7 transactions per second) - the increase in popularity of the platform simultaneously exposes its readiness to accept additional traffic / users. In such cases, either the ability to process new transactions drops dramatically, or the transaction costs increase significantly, or both - as a result, current clients withdraw their solutions from a given platform (for example, Props escaped from Ethereum to Algorand) and new clients businesses postpone the transition to blockchain "for later".
On the one hand, Algorand is working very intensively on attracting hundreds of use cases to its ecosystem, on the other hand it is developing migration bridges (such as the aforementioned re-keying function) and on the third hand it has to prepare network capacity so as not to repeat the fate of Ethereum, which has become a victim your success.
2021 = it will be an amazing year for Algorand, a year of real stress tests of their production environment.
What will be the actual transaction volume on the Algorand network in December 2021?
Millions? Hundreds of millions?
How many of these 500 projects will actually enter mainnet 2021?
Will the Algorand blockchain stand it?
Will it be possible to effectively adjust the platform's capacity and maintain a spirit of decentralization?
Everything before us.
Visual material used: