Blockchain Scalability
During his presentation Dr. Jeff Welser commented on current innovation and potential impact of an exciting new technology, blockchain. In addition to a brief explanation and some example use cases (escrow, audits, general transaction processing) he conveyed that he recognizes that blockchain has a scaling issue, a popular concern in the space. I chose to investigate the cause and proposed solutions to this issue and while blockchain technology is a far reaching moniker two of the most popular use cases Bitcoin and Ethereum serve as a proving ground, or sandbox for emerging use cases.
The Problem
While in many cases the blockchain (a public distributed ledger),often using impressive cryptographic concepts, has proven a promising emerging technology the issue of scaling is (as Dr. Jeff Welser stated) currently a barrier to a number of spaces. Bitcoin, one of the most popular and proven blockchain based cryptocurrencies currently has a market cap fluctuating around forty two billion USD yet processes about seven transactions per second. This is in stark contrast to Visa’s system which “processes 2,000 transactoins per second on average and can handle up to 56,000 transactions per second” [1]. Clearly, bitcoin has a scaling problem and the cause can be traced down to the core technology. The source code provided by pseudonym Satoshi Nakamoto in 2009 is largely the same code running on the network today. The static nature of the underlying framework is part of what makes bitcoin secure, yet it also limits the size of a block to just one megabyte. The original intention of this block size limit was to prevent hackers from launching denial of service attacks[2] but other (likely) unforeseen consequences of this limit have limited bitcoin’s utility.
Bitcoin is not alone, the novel and increasingly popular cryptocurrency Ethereum has also made headlines recently due to scaling issues [3], giving investors and adopters pause. In subsequent interviews Vitalik Buterin, the co founder of Ethereum, admitted that scaling is an issue that must be addressed. Though considered less static in nature than bitcoin, the co founder cited spiking interest from regions like china and inefficient contracts bloating the network as reasons the cryptocurrencies throughput has failed to meet demand [4].
Proposed Solutions:
While private and targeted blockchain technologies can be designed / redesigned with relative ease Bitcoin and (to a somewhat lesser extent) Ethereum are slow moving due largely to fear of a hard fork compromising the currencies integrity. In response to the scaling issue a number of solutions have been proposed / are in the works.
Increasing Block Size
Some of the more obvious and less scalable proposals have been to simply increase the size of a block on the network [2], proposals BIP (“bitcoin improvement proposal”) 100 and BIP 101 were introduced in 2015 and aimed to accomplish this in slightly different ways. The proposals were met with a heated debate and have not yet been widely implemented. A more recent proposal Blockchain Unlimited would allow miners and nodes on the network to vote on increasing block size “as and when required” [5].
Reduce Block Content Size
In contrast, other proposals advocate for reducing the size of what is contained within blocks on the network to increase capacity. SegWit (segregated witness) moves son “non-critical data, called witness data, out of transactions and off the Blockchain”[2] allowing for an increased in block sizes.
Second Layer Scaling / Payment Channels
On the bitcoin network specifically the original intention of the SegWit change [6] opens the door for second layer scaling such as the “lightning network”. In a nutshell, the lightning network is a “mutual settlement system built on top of blockchain” [2]. Essentially sequences of transactions can be started on the blockchain, then continued outside of it without using network bandwidth until the transaction series end (such as a contract ending) or a dispute must be settled.
Sharding
A long proposed improvement to the ethereum network is sharding, which draws inspiration from older database concepts and separates the blockchain into separate pieces. In this way nodes processing one shard would not be required to store and process the full state of the network with every transaction, only a portion[7]. While promising this solution is not without its uncertainties and security concerns[8].
Proof of Stake
A more drastic (though not unforeseen) change to the Ethereum network referred to “proof of stake” has also been proposed. This change would strip power from the currency miners (whose power currently comes from the “proof of work” system) and require users to put up collateral if they desire to collect fees for validating transactions instead of awarding this right to those who demonstrate computing power[9]. Referred to as the “Casper Switch” as an acknowledgment of the GHOST protocol it uses this change would reduce the block time and therefore increase the transaction throughput of the network.
The problems and solutions in the realm of the popular cryptocurrencies Bitcoin and Ethereum have serious implications for the scalability and reliability of blockchain technologies in all of sectors moving forward. The current throughput limitations have not gone unnoticed by the community (or the media) and have inspired diverse and admirable innovations. Many of these proposals will soon come to a head, and the far reaching consequences of the outcomes may decide which use cases will be entrusted to blockchain technology in the future.
[1]https://www.technologyreview.com/s/600781/technical-roadblock-might-shatter-bitcoin-dreams/
[2]https://cointelegraph.com/explained/bitcoin-scaling-problem-explained
[3]http://www.cnbc.com/2017/06/22/buyers-beware-lessons-from-the-ethereum-flash-crash.html
[7]http://www.coindesk.com/information/will-ethereum-scale/
[8]https://themerkle.com/what-is-sharding/
7 comments on “Blockchain Scalability”
Comments are closed.
Thanks for posting- the article is very informative. I have invested in the past in cryptocurrency as there was (and still is) an immense hype about the field, pushing their valuations up. To most individual investors the limitations of ethereum are not well articulated. The most advertised information is about tax free appreciating currency (in 2017 went from 50 USD to 300 USD ) and central banks turning their attention to the possibility of using smart contract. The trend is so strong that even when the second largest exchange in crypto was hacked this year, ETH continue and Ripple continued to pick up. In a nutshell whilst I see your point about scalability , I still think the crypto trend is strong and growing.
Hey Lana,
Personally, I agree and don’t think that the crypto currency trend will be going anywhere, any time soon. A very interesting aspect of crypto currencies for me is, as you said, the immense hype pushing up valuations from individual investors with very different levels of understanding and who actually use it to varying degrees with many simply investing in it like a commodity.
Your comments about these currencies being “tax free” , which is something I am interested to see evolve. I have heard many people talk about investing in crypto as tax free, when in reality
“In March 2014, the IRS issued long-awaited guidance (IRS Notice 2014-21) labeling cryptocurrency, including Bitcoin, “intangible property.” Investors and traders hold Bitcoin as a capital asset, so it receives capital gain and loss treatment.”
https://www.forbes.com/sites/greatspeculations/2017/02/21/if-you-traded-bitcoin-you-should-report-capital-gains-to-the-irs/#1574a654e3d8
Given how few people seem to actually report any gains from these investments it seems only a matter of time before the government starts to take more heavy handed actions. There have actually been some recent moves from the IRS on this front.
http://fortune.com/2017/03/19/irs-bitcoin-lawsuit/
In short, I don’t think crypto currencies are a fad. However it is a relatively new technology / concept that is exploding in popularity and is not without its limitations at the moment (such as the scaling issues I outlined in my post). It is exciting to watch the crypto community tackle these problems and I have great faith that they will continue to do so with success, but certainty not overnight and not without some “healthy” discourse.
This is a great article Kyle, thanks for posting it.
It sounds like there are two issues at hand: scalability/processing speed of the ledger technology and the value of the underlying cryptocurrency.
What do you think about the value of the currency? Is there any relationship between the value of the currency and the success of the technology? Or are these two things de-coupled?
I’m not qualified or in the practice of given advice when it comes to money, but I think it is safe to say that the success of the technology and the value of the currency are intertwined. If the underlying tech breaks down and someone can create bitcoin out of thin air then I would expect the value of bitcoin to tank. An unlimited commodity isn’t worth much.
An interesting example will be the upcoming changes related to segwit and block size limit increases (which I talked about in my post). The consensus seems to be that the value of these currencies will be in flux while this is being determined and in the end will be heavily affected by the outcome
https://www.bloomberg.com/news/articles/2017-07-10/bitcoin-risks-splintering-as-civil-war-enters-critical-month
https://techcentral.co.za/looming-split-fracture-bitcoin/75596/
@ Lana Hampicke
I want to underline your comment and recommend everybody who didn’t start researches on Blockchain technology or crypto-currencies in general, to do so.
Bitcoin is the most expensive crypto-currency at the moment. All of them are created and held electronically. The currency is decentralized, so no single institution controls the bitcoin network. This puts some people at ease, because it means that a large bank can not control their money. The price really depends on offer and demand.
It is the first example of a growing crypto-currency. But especially Etherium took his profit of Bitcoins improvement, but is still ways cheeper.
! But everybody should now, that the investments are still really risky, as you see on the Etherium chart this week → significant up and downs are predicted.
On the one hand many experts said, that most of the cryptocurrencys are completely overhyped. On the other hand for example Goldman Sachs wrote recently “Bitcoin could see a big drop then surge to almost $4,000“ (look at: http://www.businessinsider.com/bitcoin-price-goldman-sachs-2017-7). Get your own impression though.
Thanks for this eye opening article. This is a topic I was considering too ‘buzzy’ with so much excitement around it. I had to do further research to see how this technology can be applied. This article really sheds a good light on how the technology can be used in many other aspects (https://blockgeeks.com/guides/what-is-blockchain-technology/). I was actually impressed by the magnitude of the possible applications such as supply chain, governance, patent protection, property registration, copyright protection, peer to peer transactions, open markets….. Do you think such applications are also affected by the scalability problem of block chain technology or does this problem only affect financial transactions?
Another question I have to raise is how secure is this system? Its true that this technology is difficult to hack, mostly because one would need so much computational power to succeed in altering the ledgers that are distributed across the network. Now, with the emergence research around quantum computing; computational power will grow exponentially. Equipped with such computational power, would a good hacker be able to get around and hack the system?
I’m not an expert, so take all of this with a grain of salt and healthy skepticism…
My understanding is that scalability is more/less of an issue depending on the application. First, different applications would require different sized networks and different throughput requirement. Then there are different options to address scalability if it is an issue for your application (other than the solutions proposed for bitcoin and ethereum above) most notably using a blockchain that isn’t bitcoin or ethereum and customizing it to suit your needs. Though the most prominent examples bitcoin and ethereum do not represent the only blockchains, or the only ways to implement blockchains. Much of the scaling issue in bitcoin and ethereum is due to the inherent lack of trust and requirements of the decentralized community. These safeguards which prolong transaction processing may not be necessary if you are willing to allow an entity to control the network or have more power in the consensus protocol.
As for how secure it is, again it depends on the blockchain in question. Bitcoin and ethereum have proven to be relatively secure given the communities and amounts of money involved. The most prominent ethereum hacks were issues with individual smart contracts, not the blockchain itself. There have been questions about the security of the consensus protocols ever since mining pools were created, but so far all parties involved have shown a desire to play by the rules in order to keep their stake in the currency worthwhile (owning bitcoin isn’t much good if bitcoin crashes in price). That said in theory the consensus protocol is vulnerable if any one entity controls a large portion of the network
https://www.cs.cornell.edu/~ie53/publications/btcProcFC.pdf
https://arstechnica.com/security/2014/06/after-reaching-51-network-power-bitcoin-mining-pool-says-trust-us/
These protocols are still evolving and still a relatively new technology so there are definitely still unkowns. Again, some (including many financial institutions) would argue that these risks can be mitigated on more private, less decentralized blockchains with a trusted party holding the reigns
The quantum computing thing is more about cryptography. In the event that it does become viable (unlikely in the very near future, as the IBM presenter mentioned) then a malicious user can crack the key encryption used. This would be an enormous problem for bitcoin as it is today. Presumably once quantum computing is on the horizon technologies (and blockchains) will opt to use different crpyto techniques that quantum computing does not have such a significant advantage on cracking. As I understand it the government is already pushing for future looking methods of encryption to anticipate this shift.