# Algorand performance **URL:** \ **Category:** General\ **Created:** [January 3, 2022, 1:25pm UTC](https://forum.algorand.co/t/algorand-performance/5606 "2022-01-03T13:25:03Z")\ **Posts on this page:** 1\ **Showing post:** 2
**Author:** ![Maugli](https://sea2.discourse-cdn.com/flex016/user_avatar/forum.algorand.co/maugli/32/907_2.png) [@Maugli](https://forum.algorand.co/u/Maugli)\ **Post date:** [January 6, 2022, 10:47am UTC](https://forum.algorand.co/t/algorand-performance/5606/2 "2022-01-06T10:47:03Z")
In this [forum article](https://forum.algorand.co/t/question-about-the-future-of-inner-transactions-and-their-effect-on-the-network/5642) @cusma wrote: > The developing strategy for the AVM is: starting with very strict limitations and then removing or “relaxing” those limits step-by-step while being careful to preserve the equilibrium that ensures Pure PoS throughput and scalability. This is why the AVM has grow incrementally from purely stateless to stateful to even more powerful applications. So, what about the earlier expecations of 46000 tps by prof. Micali? Are these realistic for internal Algorand networks only? @cusma , please answer. --- _[View the full topic](https://forum.algorand.co/t/algorand-performance/5606)._