Diving into Decentralized Storage
May 11, 2021
Data is the “new oil” of the digital economy, and increasingly so. It is easily collected in large amounts due to the ever-growing online presence of most people and the internet connectivity of many smart devices used in everyday life. As such, it seems logical that storage of this data becomes equally important – both from a technical, as well as a regulatory/privacy viewpoint.
Currently, much of the data is stored in centralized cloud servers. While this may be efficient, it comes with concerns typical for centralized systems, such as a single point of failure. Could blockchain technology improve upon this?
Inherently, blockchain technology is not particularly efficient in storing data, as data is stored by every node participating in the network (which creates its strong uptime and accessibility guarantees). However, innovation never sleeps – multiple projects are exploring the possibilities for data storage in the crypto space and developing efficient solutions. Recent market activity also suggests that investors are equally keen to dive into the decentralized storage space (Illustration 1).
Illustration 1: Of a selection of cryptocurrency related to storage projects, BTT performed best in 2021 so far at more than +2’000%. Bitcoin’s performance added for comparison.
Decentralized Storage: An Overview
There are many contenders in the race to providing the best storage network possible – ranging from Arweave to Swarm. Also, the following overview is not comprehensive, and even more approaches exist to enable decentralized storage or an application layer, such as dedicated data marketplaces, around it.
Filecoin had its ICO in 2017 (with $257 million raised one of the largest at the time), and released its mainnet in October 2020. It is currently the largest data storage project in the crypto space by market capitalization, sitting at no. 20 or a market cap of ca. $10 billion. Filecoin is built on top of the Interplanetary File System (IPFS), a file sharing system released in 2015, and adds an incentive layer on top of it. Storage providers get paid for their service by users looking to access files in the native currency of the network, FIL, and the rate for storage is set by the open market.
Filecoin uses “Proof-of-Spacetime”, in which storage providers create blocks. Additionally, “Proof-of-Replication” ensures that data has been stored as requested and paid for by the user. FIL miners need to deposit token collateral, similar to proof-of-stake protocols, which theoretically increases the security guarantees of the network.
There is large interest in mining FIL, as evidenced by a recently announced $1.3 billion project in China. This will further boost the network storage power, which currently sits at 5.12 EiB (roughly 6 million terabytes).
BitTorrent is one of the world’s biggest peer-to-peer filesharing protocols, initially released in 2001. After being acquired by Tron in 2018 for $140 million, it integrated with the Tron blockchain through forking the IPFS (called BTFS) as a base layer system. The BTT token, currently taking spot no. 44 by market cap at ca. $4.5 billion, is used to pay for the file storage services.
BitTorrent has strong user adoption, with 200 million wallets and 100 million monthly active users. However, the network already saw much of its growth in the early days of file sharing, and the BTT token was later added. Therefore, its economic viability seems less tied to the network’s success than for some of its competitors.
Swarm, a project which was originally incubated in the Ethereum Foundation, was described in detail already in 2015 at the Ethereum DEVCON1 in London. A full overview of the project is given in the 287 pages long Book of Swarm. Swarm 1.0, the initial mainnet release, is planned for Q2 of 2021.
In Swarm, data senders and receivers interact via “forwarders” – helper nodes that facilitate the connection between two parties through the network. Nodes record how much bandwidth is consumed and provided by each party. As soon as a threshold (called “trust limit”, which increases with the number of interactions between two nodes) is reached, the consumed bandwidth is settled and paid for in the network’s native BZZ token. This settlement takes place through cheques against a balance of BZZ tokens.
Additionally, there is a cost attached to upload data to the network, and nodes are compensated for storing it. Optimal connectivity between nodes is in the interest of all network participants, and thus indirectly incentivized.
The number of nodes participating in the Swarm testnet has been growing steadily over the past weeks and currently sits at ca. 80’000.
Announced in 2014, Sia (and its cryptocurrency Siacoin) is one of the older projects in the storage space, focusing on creating decentralized, trustless markets for cloud storage. Sia’s design principles are closely related to Bitcoin, with additional features that optimize for peer-to-peer storage contracts.
Sia aims to incentivize users to rent out spare hard drive space and, through this, challenge traditional cloud storage platforms – storing 1TB of files on the network for a month is priced in the single digit dollar range. As with other platforms, the native cryptocurrency of the blockchain, Siacoin, is the medium of exchange for renting and hosting cloud storage services.
The Arweave Lightpaper was published in 2018, describing a novel approach to data storage using a blockchain-like technology called “blockweave”. Arweave was specifically designed for efficient and cheap on-chain storage, with the idea of decentralized applications building on top of the blockweave in mind as well as enabling their incentivization. The cryptocurrency, AR, serves as a means of payment in the network.
Originally, Arweave used Proof-of-Access, which – in contrast to typical blockchains – required inclusion of data from a randomly selected previous block into new blocks (creating the “weave”). However, this was later switched to a new mechanism called Succinct Proofs of Random Access.
In Arweave, nodes are ranked in a system called “Wildfire”, based on their speed of transmitting new transactions and blocks to peer, and poorly performing nodes can theoretically even become blacklisted by their peers.
Despite having been researched for years already, the decentralized storage space is still in its early days, and currently live networks have so far only started to grab market share from their traditional cloud storage competitors. Technology that can scale to massive size seems to be a pre-requisite to building a successful decentralized storage network. The bootstrapping of such a network is tricky – however, once powerful network effects incentivized by their native cryptocurrencies are established, this might lead to exponential growth if the protocols are widely integrated and easily accessible also for non-tech-savvy users.
The pace of innovation in the decentralized storage space clearly shows that interest in the area is high, both from developers as well as from the community. Various approaches exist today, each with their own nuances – time will tell which ones are capable of achieving the vision of a “decentralized internet”.