Is the cryptocurrency hyped on GetQuin $KAS (-4.22%) the $BTC (-1.71%) killer?
Has Kaspa really solved the blockchain trilemma as promised?
But first, what the hell is Kaspa anyway?
Since I have no idea about stock analysis, but have built up a certain amount of know-how about Bitcoin and crypto, I'm going to do a kind of altcoin analysis today for a change.
I'll start by explaining what Kaspa actually is and roughly how it works. I will then give you my personal assessment/opinion of Kaspa.
So - what is Kaspa?
Kaspa has been around since 2021 and is therefore a relatively young cryptocurrency that, like Bitcoin, relies on the proof-of-work consensus mechanism and pursues a novel BlockDAG (block-directed acyclic graph) approach. The Kaspa community claims that Kaspa has solved the blockchain trilemma of scalability, security and decentralization. This is why Kaspa is repeatedly praised by shitcoiners, er altcoin fans, as the "Bitcoin killer".
How Kaspa works
Kaspa uses a so-called BlockDAG instead of a classic blockchain. This means that new blocks do not have to be appended to a single predecessor blockchain in a strictly linear fashion; instead, several blocks can exist and be referenced in parallel. This concept is intended to overcome the limitations of conventional blockchains - which only process one block at a time - by processing several blocks simultaneously and later arranging them uniquely.
With Bitcoin, if two miners find a valid block at the same time, one of them is discarded. These are then so-called "orphaned blocks".
Kaspa, on the other hand, does not discard parallel blocks, but integrates them into a graph of blocks and subsequently arranges them using its consensus algorithm GhostDAG.
This results in a unique, valid sequence of all transactions from a large number of blocks found simultaneously, in which all blocks found are taken into account. Kaspa thus achieves block times of around 1 block/second, which is to be increased to more than 10 blocks/second in the future.
And here we come to the first misunderstanding, which the Kaspa community always misrepresents. It is always said that no computing power is "wasted" with Kaspa because there are no orphaned blocks that were "mined for nothing".
In reality, however, it looks like this:
(1) Several miners create parallel blocks
Several miners generate valid blocks in parallel and, of course, take some of the same transactions from the mempool.
(2) GhostDAG generates a unique sequence
After the blocks have been published in parallel, the GhostDAG algorithm sorts these blocks according to a criterion (weight). The decisive factor for the order is which block achieves the best (heaviest and best connected) position within the DAG.
(3) Allocation of fees
As soon as the final order has been determined, the block that contains the transaction first in the GhostDAG order is considered "final". This block receives the fees. All other parallel blocks containing the same transaction will not receive fees for these transactions, even if they are also valid and remain part of the DAG.
But is that a problem now? -> Yes and no.
- Miners regularly work for "free" if their parallel mined blocks are not given top priority in the DAG.
- This may reduce the long-term incentive for miners because they have to compete for (already very low) fees with higher risk. In a scenario with many parallel blocks, this could be a reason for miners to become more cautious or avoid transactions that have been in the mempool for some time. Or they try to avoid transactions with higher fees to prevent them from not receiving any transaction fees in the end.
Pruning
An obvious disadvantage of a BlockDAG with a high block rate is the flood of data. More blocks per time means massively more transaction data that nodes have to store and process. Without countermeasures, the storage requirements and sync times for new nodes would quickly explode.
Kaspa addresses this with a so-called pruning process. Pruning means that older blocks and transactions are removed from the active data set of a node, while the current state of the blockchain is retained.
In practice, this means
Normal "pruned" nodes only store a limited transaction history (perhaps a few weeks/months), while special archive nodes can optionally retain the complete history.
Pruning alleviates the storage space problem, but not the problem of high CPU load and network bandwidth, as explained below.
Hardware requirements for Kaspa nodes
With Bitcoin, anyone with a small Raspberry Pi and a hard disk can operate their own full node relatively inexpensively and thus contribute to the decentralization of the network. But what about Kaspa?
In theory, Kaspa should be able to run on inexpensive hardware because of the pruning just explained. However, if you look at the recommended hardware requirements, a problem becomes apparent:
Recommended:
- 9th generation i7 8-core processor or AMD equivalent
- 16 GB memory
- 40 Mbit internet connection
docs/Getting Started/Full Node Installation.md at main · kaspanet/docs · GitHub
These specs are significantly higher than those of a Raspberry Pi or an average older PC. They reflect the fact that a Kaspa node needs a lot of computing power and internet bandwidth at the current throughput to process the many blocks and transactions arriving in parallel in real time. And at the moment, we are still a long way from where we want to be. Hardware requirements will therefore increase significantly in the future.
Has Kaspa really solved the blockchain trilemma?
The blockchain trilemma states that of the three properties of decentralization, security and scalability, only two can be maximized at any one time, while the third suffers. Many projects claim (and have claimed in the past) to have solved or circumvented this dilemma - Kaspa is no exception.
The official website states that GhostDAG is the only decentralized protocol that completely solves the trilemma. Kaspa combines a level of security and decentralization à la Bitcoin with performance comparable to state-of-the-art proof-of-stake networks. The team backs up this strong claim by pointing out that in Kaspa, blocks no longer need to be "discarded" and the block rate is only limited by hardware and the network, no longer by consensus rules (Features - Kaspa).
Of course, that sounds great and is a big promise. So it's no wonder that Kaspa is so hyped. But let's take a closer look at the three characteristics of the blockchain trilemma:
Decentralization
A high degree of decentralization means that as many independent participants as possible can fully support the network - whether as miners or validating nodes.
Kaspa is formally permissionless (open to all) and relies on PoW without a stake or master node, which is initially good for decentralization.
However, practical decentralization could suffer due to the aforementioned hardware requirements and the fast clock. If operating a full node requires a high-end computer with a stable broadband connection, many users will be put off and not operate their own node.
The more data a node has to consume and store, the more expensive it becomes to operate - and the more the number of nodes in the network decreases because only a few are willing or able to bear these costs. Kaspa tries to counteract this with pruning.
Although the storage load remains low, at the same time most nodes rely on not holding all historical data. This results in a certain division of labor: "light" full nodes vs. archive nodes. As long as there are enough archive nodes, this is relatively unproblematic. However, if there are only a few, trust could suffer - for example, if a normal node needs an older UTXO, it has to rely on other archival nodes and hope that they speak the truth. In my opinion, this could lead to a kind of oracle problem in the future, but it doesn't have to.
In addition, the high block frequency can be a hindrance for participants in regions with poorer internet connections (e.g. Germany) - one block per second (or even 10/sec.) means enormous data rates and hardly any time buffer during distribution.
In my opinion, Kaspa is making the same mistake as Roger Ver and the "big blockers" in the block size wars of 2015-2017.
Back then, there was a fierce debate about how Bitcoin should scale to enable more transactions per second. Some called for larger blocks - meaning more transactions directly on the blockchain - to scale quickly and in the short term. But the majority of the Bitcoin community rejected this. Why was that?
The key reason was that larger blocks drastically increase the drastically increase the hardware requirements for nodes. Larger blocks mean that nodes have to process and store more and more data, which increases their operating costs enormously. As a consequence, at some point only professional players - such as companies with specialized data centers - would have been able to operate fully-fledged nodes. This would have led to an increasing centralization of the network. Bitcoin therefore deliberately decided against larger blocks and in favor of scaling via layer 2 solutions, such as the Lightning Network. This kept hardware requirements low and decentralization intact.
This is precisely the critical point with Kaspa: Due to the extremely high block rate of one block per second (and much more in the long term), Kaspa follows exactly the approach of the "big blockers" of the time.
If only a few, well-equipped players are able to operate Kaspa nodes, control of the network falls into a few hands - exactly what Bitcoin wanted to prevent with all its might during the block size wars.
With its focus on scalability, Kaspa risks jeopardizing decentralization in the long term (my opinion).
Security
Like Bitcoin, Kaspa relies on proof-of-work and thus fundamentally inherits its security model:
As long as the majority of the hashrate is honest, the system is secure; however, if an attacker manages to provide >50% of the computing power, they can double-spend and reorganize, censor transactions, etc..
The Bitcoin network is the most secure computer network in the world. Not even states could oppose Bitcoin. Kaspa is nowhere near this level of security. A 51% attack would be much more likely with Kaspa than with Bitcoin.
Scalability
In terms of scalability, Kaspa has clear strengths on paper: many blocks per second, fast confirmations and therefore high transaction capacity.
However, the crucial question is: at what price?
As described, the price paid is primarily higher system complexity and hardware requirements. There is also the question of the sustainability of this scaling. High throughput is only an advantage if there is enough real usage to utilize it. Empty blocks every second bring fast confirmations, but do not fill up with fees on their own. If the demand for transactions lags behind the supply, Kaspa runs the risk of having a very inefficient utilization. This leads to an old problem that has also been frequently discussed with Bitcoin in the past - the security budget problem. But more on this in the next section.
Scalability is therefore not an end in itself. Kaspa must be able to generate sufficient adoption and transaction volume so that the high capacity is also utilized. Otherwise, you would have a sports car-fast protocol that chugs along at idle - and still consumes energy and resources.
So back to the original question: No, in my opinion Kaspa has not solved the blockchain trilemma. In my opinion, Kaspa has once again compromised on decentralization and security in order to be able to scale as well as possible.
Kaspa's security budget problem
The security budget is the sum of the incentives that a blockchain network provides to motivate miners (or validators) to secure the network. In the case of PoW, these are the block rewards (block subsidy + transaction fees). In the long term, the main question here is: What happens when the block subsidy expires? Bitcoin is discussing this problem intensively, as all 21 million BTC will be mined in 2140 - Kaspa, however, will reach this point much earlier due to an aggressive issuance plan - and will probably run into problems much sooner.
Kaspa was launched fairly on November 7, 2021 (no premine, no early allocations) and has a fixed issuance schedule from the beginning, which runs much faster than Bitcoin. In the first few months after launch, the block subsidy was 500 KAS per second. It then gradually decreases - the block subsidy is halved each year.
Now consider the following: How well does Kaspa actually have to perform for it to be worthwhile for miners to secure the network despite mining rewards halving every year? And in view of the fact that Kaspa confirms transactions every second and there is no competition for block space, as is the case with Bitcoin?
With Bitcoin, users outbid each other with transaction fees so that their transaction is included in a block and executed as quickly as possible. The effect is high transaction fees for miners - in the past, there have been some blocks where the transaction fees were higher than the block subsidy.
With Kaspa, you hardly pay any transaction fees and the transactions are executed almost in real time - cool, isn't it? But how is mining with Kaspa supposed to pay off in the long term? How is the network supposed to be secured in the long term? I don't see (and again, this is my own opinion, you're welcome to prove me wrong) how Kaspa is supposed to finance itself in the long term. And I believe that Kaspa will fail precisely because of this aspect. Long before decentralization even becomes an issue because of the high hardware requirements.
I actually wanted to write a section about why there will probably only be a proof-of-work network in the long term, but as the article would otherwise be completely out of scope, I'm postponing it. If there is interest, it will become a separate post :)
Conclusion
On paper, Kaspa impresses with its fast transaction processing and interesting approaches such as GhostDAG and pruning. However, a closer look reveals that Kaspa by no means completely solves the blockchain trilemma. Rather, it shifts the problem from scalability to decentralization and long-term security.
The high hardware requirements jeopardize decentralization in the medium term. Kaspa runs the risk of making the same mistake as the "big blockers" did during the Bitcoin Blocksize Wars. The danger of centralizing the nodes in professional data centers is very real.
But perhaps the biggest question mark is the security budget problem: how is a network that hardly charges any transaction fees and at the same time is rapidly reducing the block subsidy supposed to finance and secure itself in the long term?
I would say that Kaspa can be categorized as a "premium shitcoin" due to proof-of-work and no premine. But I have no idea how Kaspa will be able to establish itself in the future.
I hope I was able to help some of you a little with this article. As always, please feel free to post any questions, criticism etc. in the comments.
I would also be interested to know what you thought of this type of article and whether you would like me to write something about other altcoins in the future.
Have a nice weekend!
