GP Farm On Near

@GPfarm

GP Farm On Near

Việt Nam

Joined January 2023

84 Following

15 Followers

56 Posts

GPfarm retweeted

Justin Bons

@Justin_Bons

10 months ago

Let's cut through the BS with hard facts: 🧵 BTC= 7 TPS with 82 nodes ETH= 179 TPS with 8.1k nodes SOL= 20k+ TPS with 1.1k nodes ADA= 18 TPS with 3k nodes SUI= 20k+ TPS with 121 nodes APT= 20k+ TPS with 150 nodes NEAR= 100k+ TPS with 296 nodes ALGO= 10k+ TPS with 3.7k nodes (TPS = Capacity) (Nodes = Decentralization) Capacity + Decentralization = Scalability! This is all, a gross oversimplification, as there are far more metrics to take into account & many more devils in the details. However, these are still incredibly important metrics on their own & should not be underestimated! Node Count Methodology: When it comes to the number of "nodes", I was specifically referring to native-delegated & solo validators (block producers). Based on this definition, we can only count "pool nodes" in the case of BTC, as they serve the equivalent function of a "validator" in PoS. Considering that miners do not even run a node & full nodes are not involved in block production! In the case of ETH, we count the physical number of consensus clients. That has to be the upper bound for the number of "validators" on ETH, due to it being a requirement for block production. TPS Measure Methodology: When it comes to TPS, we are specifically referring to the maximum Theoretical Transactions Per Second for basic 1 to 1 value transfers. This gives us a fair & comparable scalability metric across all chains! The TPS figures were also rounded down in the case of the most advanced scaling technologies. In SOL's case, for example, the raw cu/gass figures give it a TPS exceeding 200k! However, in reality, bottlenecks are hit much sooner, in part because smart contract chains TX are far more complex. This leads to different bottlenecks getting hit much sooner in the real world. However, in the hypothetical scenario of running as many simple TXs as possible (for the purpose of comparison). We hit an EDDSA verification bottleneck at around 30k TPS & there are also other significant bottlenecks further up. This means we have to account for this with specific knowledge of these blockchains: That is how, while being aware of these specific bottlenecks for purely parallelized chains, we can safely estimate a TPS of 20k+ for SOL, SUI & APT. Chains utilizing sharding get around this bottleneck & are in the 100K+ TPS range (this is assuming more shards can be deployed to meet demand). For any chain that does not deploy advanced on-chain scaling techniques, we can give an exact TPS figure, as there are hard limits that can be easily calculated. Like in ETH (gas limit) & ADA & BTC (block size limit), respectively Need For Speed: Speed is distinct from capacity in the case of blockchain design; it is also one of the most important metrics, alongside more decentralization metrics. It did not fit in the opening, & I wanted to keep it more simple, so we are including it here instead: BTC= 7 TPS at 10min on 82 nodes ETH= 179 TPS at 12sec on 8.1k nodes SOL= 20k+ TPS at 0.4sec on 1.1k nodes ADA= 18 TPS at 20sec on 3k nodes SUI= 20k+ TPS at 0.5sec on 121 nodes APT= 20k+ TPS at 0.2sec on 150 nodes NEAR= 100k+ TPS at 0.6sec on 296 nodes ALGO= 10k+ TPS at 2.8sec on 3.7k nodes Speed, in this case, is simply measuring block time, not finality, which is another equally valuable metric. The real trade-off space is not between capacity & decentralization, but rather between capacity & speed. Another reason why the classical blockchain trilemma is so incredibly outdated today. My Selection, Deal With It: Instead of strictly following market capitalization as I usually do for such lists & other specific rules for inclusion, in order to avoid accusations of bias. I just included whatever I wanted to this time instead... I will be accused of being a paid shill, no matter what I do at this point, while also facing extreme toxicity for telling everyone about these basic truths... That is why I only included what I thought was interesting to compare & whatever happened to fit within the X-character limit, as most people do not read the body of the text, only the start anyway. It is likely that some of you will still think it is a grand conspiracy against your obscure chain because the tech is just so good... The tribalism, cope & delusion have become so ridiculous for some in this world... Honorable Exception: I usually exclude ALGO from such lists due to its centralized relay nodes, which make it technically permissioned (centralized). However, considering they are very close to fixing that issue, I thought I could make an exception here... So, this can be considered a shoutout for their positive efforts! Conclusion: There is little correlation between the number of nodes & TPS. That is because many aspects of the blockchain trilemma have been effectively solved & the trade-offs in the design space are now more focused on other levers such as capacity vs speed. Economic & consensus design also ends up having a far greater impact on node counts, at least when compared to the total capacity of these chains. Shoutout to ETH for having the most nodes & NEAR for having the highest capacity! 🏆 We should all strive to increase these figures & praise those who push the boundaries of innovation. The goal is to maximize freedom (decentralization) for as many people as possible (scalability). As cryptocurrency is useless without capacity & pointless without decentralization! 🔥

260

565

172

114K

GP Farm On Near

@GPfarm

Last Seen Users on Sotwe

Trends for you

Most Popular Users