The developer and chief technology officer of the startup Casa, Jameson Lopp, after testing seven different clients of a full Bitcoin node, once again proved, as in 2018, that official Bitcoin Core (0.19.0) is the most synchronization software Fast and efficient
Lopp published on his blog, on November 10, the result of his study by which he tested the performance of the different implementations to run full nodes. In that way, he concluded that Bitcoin Core is the most efficient software to verify all transaction data, taking into account that it needed a synchronization time of 6 hours and 39 minutes to complete the evaluation.
It is convenient to note that Bitcoin Core is the reference software of Bitcoin and the most popular implementation, taking into account that 96% of the reachable nodes of the network run some version of Bitcoin Core. It was originally released as Bitcoin 0.1 by Satoshi Nakamoto on January 9, 2009, but changed its name to Bitcoin Core in 2014. It is open source and operates a full node.
Jameson further explained that the other six software tests he used were Bitcoin 77d8804, which took 18 hours and 29 minutes. While Gocoin 1.9.7 took 19 hours and 56 minutes. The Libbitcoin 3.2.0 Node completed the test in 1 day, 3 hours and 37 minutes, while Parity Bitcoin 7fb158d did it in 2 days, 2 hours and 10 minutes. BTCD v0.20.0-beta, culminated in 3 days, 3 hours and 12 minutes, while Stratis 18.104.22.168 did not complete the test.
In a tweet Jameson wrote: “It took 399 minutes to synchronize Bitcoin Core 0.19 from Genesis to block 601,300 with assumevalid = 0 on my reference machine. The bottleneck is still the CPU ». In a similar message from a year ago, the developer needed 88 minutes less to perform the test with Bitcoin Core 0.17.
As the name implies, full nodes are responsible for maintaining and distributing copies of the entire blockchain record. As such, they play a vital role in networking, as they are the main point in the validation of transactions.
When comparing his recent test with that of last year, Lopp noted that the implementations took longer to synchronize. In that regard, he said that the total size of the blockchain has grown by 30% since its last round of tests, so he would expect that an implementation without further performance improvements or bottlenecks would require more or less time to complete his synchronization.
Lopp concludes that due to the nature of blockchains, the amount of data that must be validated for a new node that synchronizes from scratch will continue to increase steadily over time. As a positive point, he pointed out that the tests he has been carrying out are on the same hardware, however, he considers that the hardware performance will also increase every year.