Sharding
Our peer-reviewed Chainspace paper describes the Sharded Byzantine Atomic Commit (S-BAC) protocol for internet-scale blockchains. Presented at NDSS 2018.
Read the paperChainspace powers a community of makers.
We're helping everyone have more control over our digital world.
A distributed web of blockchains for scalability, speed and privacy.
Built for developers, designed for everybody.
Eco-system for peer-to-peer services and communities.
We are currently writing a whitepaper describing the overall functioning of the system. It links together all of our other works (below) into a coherent whole.
Chainspace is based on a number of peer-reviewed scientific works from our team and external collaborators.
Our peer-reviewed Chainspace paper describes the Sharded Byzantine Atomic Commit (S-BAC) protocol for internet-scale blockchains. Presented at NDSS 2018.
Read the paperOur new leaderless consensus protocol, Blockmania, achieves extremely fast and high-throughput consensus within each shard.
Read the paperFraud and data availability proofs are key to enabling on-chain scaling of blockchains while maintaining a strong assurance that on-chain data is available and valid.
Read the paperCoconut distributed threshold credential issuance integrates with blockchains to ensure confidentiality, authenticity and availability - even when a subset of credential issuing authorities are malicious or offline.
Read the paperSystemisation of Knowledge is a conceptual framework to evaluate the performance, security and design properties of different consensus frameworks.
Read the paperUses two-way stake wagers between entities, to build a social network of stake-weighted trust relationships.
Coming soonChainspace delivers on modern expectations of processing power, making it the platform of choice for complex real-world applications
Our new leaderless consensus algorithm, Blockmania is based on a DAG structure and is extremely fast, currently 350k-500k TPS on 16 nodes in a single shard distributed across the internet. It does consensus and transaction ordering within a single shard.
Our Sharded Byzantine Atomic Commit (SBAC) protocol allows us to run multiple chains at once. Incoming transactions can affect objects in more than one shard, allowing us to get horizontal scalability. Cross-shard commits are atomic.
Blockmania consensus functions as long as more than ⅔ nodes are honest. We can horizontally scale verification nodes, just as we can scale execution nodes. Verification nodes use fraud proofs to detect and communicate about shards that have gone rogue.
We use two-way stake wagers between entities, to build a social network of stake-weighted trust relationships. We use social network theory math to figure out whether an entity is a sybil based on the amount that it's connected to the main network.
How do you allow people to prove that they are able to vote in a petition, tally votes, and not allow anyone to vote more than once, without the user's actual identity to ever being recorded? Coconut, our threshold credential disclosure library, makes applications like this possible.
In order for the system to function, we need to have a DNS-like system mapping objects to nodes, nodes to shards, and contracts to nodes, shards & objects. We plan to implement this as a Chainspace smart contract itself.
Blockchain developers can use any language with a WASM output target, such as Solidity, Vyper, Hera, Rust, C++, and Go - and will get strong determinism guarantees. Alternately, as with Tendermint, developers who want to use traditional languages and commit to avoiding nondeterministic function calls, are free to do so.
We have both strong research and serious production software development skills. Our team has built hundreds of projects together in the past, and we are battle-tested when it comes to delivering software.
In a 20 year career, Dave has successfully delivered hundreds of software systems. From startups to the world's largest organizations, he's scaled systems to tens of millions of users. He also wrote the book on the open source Scalatra HTTP framework, used by the BBC, NetFlix and McLaren. Alongside Ramsey, he built and sold a 60-person software consultancy. At Chainspace, he works across business, code and research activities.
George is the former head of the Information Security Research Group at UCL. He co-designed and prototyped RsCoin, an experimental scalable cryptocurrency, and has a list of publications longer than a Leonard Cohen song. He did his doctorate at Cambridge under Ross Anderson, followed by a stint at Microsoft Research. George is also a Fellow at the Alan Turing Institute.
A successful serial entrepreneur, Ramsey provides our business backbone. Previously, Ramsey founded, built up, and took a 60-person software consultancy through to a final sale. Subsequently an angel investor in multiple traditional startups, his interest in the potential of decentralised applications has been stoked over the past three years. He handles investor relations, business strategy, and a million details.
Alberto's research interests include privacy enhancing technologies, distributed ledgers, and secure enclave computations. He is also currently a doctoral researcher at UCL.
Andy is in charge of Brand & Product Design. Systematising UI component patterns for distributed apps and designing prototype applications.
Jeremy is the sort of coder who implements RAFT distributed consensus in Rust as a personal challenge, and whose Github repos have descriptions like "C++ reflection framework (for fun)". He's equally comfortable working in Go or Rust, and leads our experiments with Rust-to-WASM.
Lola is a product strategist who specialises in figuring out what new tech is actually useful for. She's in charge of research, prototyping, and business strategy to find what's useful (and ditch what's not) in what's still a very immature design space.
Mustafa is a researcher in computer security, scalable blockchain technologies and their applications. He made the Forbes 30 Under 30 in Techlist for 2016. He is currently a doctoral researcher in computer security at UCL.
Penny knows how to get teams working together. She has previously worked with organisations ranging from computer security and chat startups to major consumer brands and government organisations.
Bano is an expert on networking and distributed systems, with a PhD from Cambridge University, and a post-doctoral researcher at UCL. She is a member of IC3, the Centre for Blockchain Technologies,and helps run the London Infosec Group. Made MIT's 2018 Innovators under 35 list.
Stuart scaled a chat app to tens of millions of users. He builds developer-facing prototypes across multiple platforms.
A founder of the Apache Software Foundation, Ben wrote the SSL implementation in Apache HTTP server. He's also originator of Certificate Transparency, an internet-scale distributed ledger securing SSL against MITM attacks. Ben is Head of Security and Transparency at DeepMind.
Bryan Ford leads the Decentralized and Distributed Systems (DEDIS) lab at EPFL in Lausanne, Switzerland. He's one of the main brains behind OmniLedger,a sharded cryptocurrency. He also writes on Liquid Democracy, a set of governance structures designed for distributed democratic deliberation.
Klaus is an expert security consultant on critical infrastructure, including smart devices and energy grids across the globe. He has an encyclopedic knowledge of Byzantine distributed consensus and distributed trust,and has been publishing papers on these subjects since the turn of the millenium.
For investment opportunities email us at investors@chainspace.io
We have an alpha system running now in Golang.
It does sharding, consensus & executes smart contracts.
A global software consultancy is using Chainspace on the EU project Decode, creating civic services for the cities of Barcelona & Amsterdam.
We are currently achieving speeds of between 350-500k transactions per second on 16 nodes in a single shard distributed across the internet.
Our Sharded Byzantine Atomic Commit (SBAC) protocol allows you to run multiple chains at once. Incoming transactions can affect objects in more than one shard, allowing you to get horizontal scalability. Cross-shard commits are atomic.
Allows you to issue a re-randomizable credential from a smart contract call. Presents the unlinkable credential back to a contract anonymously to deposit a token or show an entitlement.
Blockmania consensus functions as long as more than ⅔ nodes are honest. What happens if all nodes in a shard go bad? Nodes will use fraud proofs to detect shards that are dishonest. Slashing their stake and rolling back their operations to a consistent state. We plan to horizontally scale verification, just as we can scale execution.
An open system needs strong Sybil defences. Proof of Stake protocols in most cases mean that attackers can simply borrow a lot of money, turn it into tokens, and subvert the system. To defend against this, we plan to use two-way stake wagers between entities, to build a social network of stake-weighted trust relationships.
We need to have a directory mapping objects to nodes, nodes to shards, and contracts to nodes, shards & objects. We plan to implement this as a Chainspace smart contract itself. It will support open governance about how shards are composed that contract creators can rely on.