Scroll Top

Australian researchers claim a blockchain record of 440,000 transactions per second

WHY THIS MATTERS IN BRIEF

Blockchains relatively low transaction speeds could stop it from reaching its full potential, but now a research group from USYD claim to have smashed the performance barrier that could have held it back.

 

Researchers at Australia’s University of Sydney (USYD) announced this week that they’ve developed a new blockchain variant that can process over 440,000 transactions per second, and they say that they believe the system has the potential to revolutionise the global economy.

 

RELATED
Ethereum falls after rumours of new powerful ASIC emerge

 

The new blockchain variant, dubbed Red Belly Blockchain (RBB) after the country’s favourite snake would allow the almost instantaneous processing of huge volumes of cryptocurrency transactions, as well as other types, and when it’s compared to the Visa network’s processing speed of 56,000 transactions per second (TPS), PayPal’s 450 TPS, Ethereum’s 20 TPS, and Bitcoins measly 8 TPS it’s suddenly easy to see why the university think they’re onto a big deal.

As blockchain becomes the defacto technology touted by many to underpin everything from China’s future cryptocurrency, cloud computing and HPCdigital identities, the decentralised internet, Forex trading, supply chains, the entire US military, and everything in between, oh, and solve world hunger and world poverty… unless blockchains can process all of these transactions in a cost effective, fast and secure way then it’s likely that most of the initiatives, at some point, will cave, and that’s the problem that RBB looks like it could solve.

“The Red Belly Blockchain offers unprecedented throughput of more than 400 thousand transactions per second when it’s running on 100 machines and its safety aspect is of invaluable importance for critical industries, like banking, and unlike other blockchains its performance scales horizontally,” wrote the researchers.

 

RELATED
Adidas jumps into the world of NFTs with Bored Ape Yacht Club

 

The researchers also claim that the Red Belly Blockchain is the first blockchain variant to be built on both public and private networks which, if validated, would not only allow the peer-to-peer exchange of data but would also allow an unprecedented number of people and “nodes” to use and access the system simultaneously.

 

TPS Comparison Chart

 

“As opposed to mainstream public blockchains, [the RBB] is not subject to double spending, that’s when an individual successfully spends their money more than once, because its chain of blocks never forks,” said Dr. Vincent Gramoli, Head of Research at the Concurrent Systems Research Group at USYD, who headed up the program.

According to Gramoli by scaling horizontally RBB can avoid common issues that ail other blockchain variants, such as forking.

 

RELATED
Blockchain Autonomous Organisations, a new way to tackle corporate fuelled wealth inequality?

 

“The absence of forks means there is no need for any number of confirmations. A transaction in a block is there for good,” he said, “as opposed to consortium blockchains, it can process hundreds of thousands of transactions per second coming from a potentially unbounded number of clients, and it offers a performance that scales horizontally, which ensures the security of transactions.”

Now that the team have the first RBB system up and running their next step will be to develop a recommendation system that would automate the selection of participants of a consensus instance, so, as many people complain about the slow, and increasingly expensive, transaction processing speeds of today’s blockchains, who knows, tomorrow that problem might be just a thing of the past, and that’s only going to help boost blockchain’s adoption.

Related Posts

Comments (16)

“The Red Belly Blockchain offers unprecedented throughput of more than 400 thousand transactions per second when it’s running on 100 machines and its safety aspect is of invaluable importance for critical industries, like banking, and unlike other blockchains its performance scales horizontally,” wrote the researchers.

Charles M. I agree that outside the “blockchain” context, this sentence does not make sense. Think of a crash-tolerant model in which databases can outperform significantly the Red Belly Blockchain.

Plenty exits from ~ the 1980, these were called “non stop” systems and underpinned most critical resources, and still do today.?
To be viable any BC muts actually meet the requirements of the use case, if one is prepared to wait 24hrs for censorship resistant anarchy based cash fine, but to suggest it provides a solution to general purpose (non stop) distributed computing, is stretching reality into the world of delusion..
Compare thingy to say “cassandra” it will fail on every comparison point.. including tps? and yes all databases since ~ 1970 have run consensus algorithms for any distributed data, but have all relied upon techncial security not goodwill to power the result..
Asa above a simple response to the ability to match 2,000 messages per millisecond? quoting tps on what are “batch” processing is a con? why its always about “latency” not tps in any distributed transaction space?.

Now do the same with 1000, 10000 and then 100000 nodes. 100 nodes is nothing.

Ivan Fernandez & Sean Lee – can you remember the scale we had with HPDA? Would be an interesting comparison given how long ago it was put together.

Call the bullshit here..the tps must be unlimited its always about latency not tps..
Nubes abound in this space.

Be great to run something here? David Tasker

Two important points:

1) their blockchain design is targeted for “consortium” blockchains, not permissionless (public) and not permissioned (private) blockchains, so keep this in mind when analyzing or critiquing
2) their consensus is very different than the current crop of consensus algorithms, which facilitates the throughput numbers; read their white paper, which describes the consensus and theory behind the design

(http://poseidon.it.usyd.edu.au/~concurrentsystems/doc/ConsensusRedBellyBlockchain.pdf)

Note, their work product will be commercialized, so it will not be open sourced, and therefore, will not become available to the public as a new universal norm for throughput, except as embodied in consensus-like clones–provided they reach some momentum and sizable install base.

Note, I am implementing a version of the consensus algorithm described in their paper with the desire to benchmark it against other competing designs and possible extend it or refactor it.

(1) our blockchain is not a consortium blockchain: think about a public consortium instead where representatives are picked among the public but simply to serve at one point of the execution then others are picked, so that eventually all nodes present for long enough can participate.
(2) the main difference is that we redefined the classic Byzantine consensus problem in the light of blockchain: instead of using BFT as a black box off the shelf, we came up with a blockchain consensus solution.

We will eventually open source it, but we first want an easy-to-use solution before releasing.

We would really like to hear about your experiments and performance results, let us know if you need some help.

Also, we are about to release in a paper an optimized variant of the algorithm linked above. To be informed, feel free to enter your email at the bottom of this page: http://poseidon.it.usyd.edu.au/~concurrentsystems/rbbc/

An form of consensus which doe not involve 100% of the involved parties is a centralised system pretending to be distributed… bitcoin is just one of these..
My thought is simple here, if in any consensus a single party disagreement is proven true, is there any consuss..
My anser is always no..
YOur model will lead to a false consus position being adopted by the chain, whcih is a vulnerability…
The bitcoin slight of hand with POW is very cleaver,and I haven’t seen any pretend consensus algorithms get close, including the one in red chain from the description in the white paper..
As one who has spent many years in the crypto space it is a simple fact one cannot have both speed and security there mist be a trade off..
One must state what are the trade offs in any design decision to allow full due diligence..
Enjoy..

PS: and this is with quantum resistant signatures on all transactions.. and true temporal bindings.
Enjoy..

Do you know you can earn 0.2btc every 9.6hours. We mine a block of bitcoin on the blockchain mining Network and cloud mining has been easier with the bitmain bitminer l3+ hardware.
Leave a message if interested and thank me later.
We have Bitcoin for sale too.

Does this also reduce the energy and computing consumption required for network to function?

If you refer to mainstream blockchains as the baseline, then yes. There is no need to prove any computational work.

Graham Hartland thought you’d like the read…

Anyway the actual requirement is with 7 billion devices the total transaction needs to be ~ 100ms the actual local marketplace latency we experience today in is ~ 30ms.. the order book matching should be ~ 250ns yep Nano to be commercial competitive..
TPS, no one cares?
Suspect the Red thingy does not even change state in 250ns in a 100% distributed system?
We can acheive a 250ns match in any local marketplace rail order book, with close to infinite tps ( only limited by the sharding).. The objective latency is 30ms, even though we claim ~ 100ms today..

Leave a comment

EXPLORE MORE!

1000's of articles about the exponential future, 1000's of pages of insights, 1000's of videos, and 100's of exponential technologies: Get The Email from 311, your no-nonsense briefing on all the biggest stories in exponential technology and science.

You have Successfully Subscribed!

Pin It on Pinterest

Share This