Blockchain testing company, Whiteblock, published the results of a comprehensive research of the EOS blockchain titled EOS: An Architectural, Performance, and Economic Analysis.
The study concluded the following:
- EOS’ architecture is not fundamentally a blockchain-based network
- Network throughput is less than 50 TPS in a real-life setting
- Consensus issues
- No Byzantine fault tolerance
- Faulty voting mechanism
- EOS token is solely for purchasing computational power from the 21 block producers (BP)
- There is no transparency on the configuration and processing power of the BP
- EOS would need to redesign a ‘significant portion’ of its structure for it to be successful
Although very thorough and well researched, the report was heavily criticized by EOS community members due to the fact that several of the authors and contributors are indirectly involved with Ethereum (via a company called ConsenSys founded by Ethereum co-founder).
Centralized, slow, and flawed non-blockchain system
EOS is not a blockchain
The researchers explain that the network’s core structure does not cryptographically validate transactions but that it is only “implemented in proximity”. According to them, EOS is more like a cloud service where users use their tokens to gain access to services.
Less than 50 TPS
Furthermore, Whiteblock explains that in terms of transaction speed, EOS falls short of what the development team led by Dan Larimer promised in the whitepaper.
The performance section of the report reveals that in a real-life setting, the network can process less than 50 transactions per second (TPS), which is miles away from the thousands claimed by devs.
Centralized block producers
The level of decentralization on EOS network was also scrutinized, and called highly questionable, as the model where the power is concentrated in 21 block producers is said to be flawed and lacking transparency.
The supposed “Ethereum killer” was called insecure and the consensus did not demonstrate Byzantine Fault Tolerance.
Needless to say, the published research raised some eyebrows throughout the community of EOS supporters. Some of them claim that, since the research was funded partially by the EOS’ competitor, ConsenSys company, it represents a clear case of the conflict of interest.
Not a biased report
Whiteblock’s CTO Zak Cole gave an interview in order to clear up accusations of the report being biased and to provide explanations on the report.
As an answer to skeptics, Cole emphasized the neutrality of the research by saying that his research team doesn’t favor Etherum nor EOS as main competitors for the leading smart contract platform in the industry.
For those who don’t know, Bitshares was co-founded by Dan Larimer, and the technology behind it is also used partially by EOS.
“We’re blockchain people,” he declared, further explaining that the intent of the research wasn’t to compare the two in any way but to provide a scientifically based opinion of the EOS network.
When asked about the EOS’ TPS rate, Cole answered that the team noticed a substantial decrease in the number of successful transactions as the number of users on the network grows.
He concluded the interview by criticizing the EOS community for being doubtful of such tests and said that this kind of defensive stance may present a huge problem for the development of the EOS network.
The EOS community should focus their attention on validating the report rather than criticizing it blindly. In the end, numbers don’t lie.