cross-posted from: https://infosec.pub/post/1264068

From all technical points of view, the network works perfectly.

I’ve attached some interesting graphs to prove it. More info: https://cexplorer.io/

block filling / capacity block filling / capacity

TPS progress and potential cap with same tx structure TPS progress and potential cap with same tx structure

x-posted from Reddit.

  • Cam
    link
    fedilink
    English
    211 months ago

    How many tps can cardano handle?

    • @demesisxOPM
      link
      English
      5
      edit-2
      11 months ago

      Currently 1-5 TPS

      That’s an answer that should be taken with a grain of salt for a eUTXO chain since transactions can be batched together in many ways that the accounts model cannot.

      The current TPS of Cardano is very slow since it is a single-threaded eUTXO. However, utxo can achieve all kinds of parallelism tricks that the accounts model (which depends on global state) cannot. For example, Ergo (a eUTXO sister protocol) holds a record for chained transactions with the execution of 10,000 outputs in a single transaction.

      Then, we have the Hydra heads on the Cardano side. The current estimate is that each Hydra head can process 1000 TPS…and there can be thousands of hydra heads running at the same time.

      Right now: Single-Threaded Mode 1-5 TPS

      Pessimistic Future Spec: 1,000 TPS

      Optimistic Future Spec: 1,000,000 TPS using batching, input-endorsers, parallelism, and all of the other tricks that eutxo enables

      Here’s a good article where they kind of lay out the scaling solutions that will bring about the scaling that I am talking about.

      How they’re scaling dApps with Plutus V2.

      Here’s Charles answering the question about scaling.