Justin Bons, founder and chief investment officer of Cyber Capital, has reopened an old wound in Cardano’s scaling narrative, telling his followers on X that the network’s “arbitrary” block-size ceiling and real-world throughput of roughly 18 transactions per second (TPS) leave “no guarantee of future scale” for builders—or for institutional allocators.
Cardano Faces Backlash Over 18 TPS Limit
“I want to like ADA; I really do,” Bons wrote June 10. “Can someone please convince me to take their scaling efforts seriously? ADA’s current capacity is 18 TPS, which is pathetic! Input Endorsers only increases that to 400 TPS… L2s such as Hydra do not solve for L1 scaling. So what am I missing?”
That salvo drew an immediate technical rebuttal from Cardano advocates. Stake-pool operator Jaromír “Cardano YOD₳” Tesař replied with a screenshot from IOG’s Ouroboros Leios paper, saying the new consensus layer “can have a throughput of around 10,000 TPS, and maybe higher, while maintaining a single global state.”
Minutes later, Phil UPLC, chief executive of Anastasia Labs, added that internal simulations “with 30 input-blocks per second hit 11,790 TPS,” insisting there is “nothing preventing higher IB/s outside of increasing validator hardware requirements.” Bons was unmoved.
“This is all great in theory,” he conceded. “If this were implemented in the next year, along with a block-size increase, it would flip my support. However, promises alone are not enough, as this has been promised for many years now. … Without stronger guarantees, the roadmap is too speculative.”
Leios, first described in a November 2023 research pre-print and now tracked via weekly progress notes from IOG’s dedicated scaling team, proposes a three-tier block structure—input, endorsement and ranking blocks—that allows nodes to process transactions in parallel before final ordering. The design echoes the “local fee market” strategy popularised by Solana, yet aims to retain Cardano’s stake-weighted Nakamoto consensus for settlement.
For critics, the sticking point is not design elegance but time to mainnet. Bons pressed the issue again when Anastasia Labs’ Phil argued that Input Endorsers (IE) without Leios should already reach “much more than” 400 TPS if governance lifts the hard cap. “The 400 TPS figure was from running IE without Leios,” Bons replied. “I was expecting IE to launch before Leios. Is that what I am missing? I understood there is no release date for Leios? I am reading the Lanningham paper now, which is brilliant—but still purely theoretical.”
The debate then shifted to Cardano’s conservative block parameters. Delegate representative @Cerkoryn defended the status quo, saying the 100 kB body limit “makes the chain more egalitarian by making it so that almost anyone can run a node with a basic desktop computer” and arguing that the “bigger issue is having enough meaningful transactions to fill block space to begin with.”
Bons countered with a flashback to the Bitcoin scaling wars: “You’re defending the block-size limit, then… Based on this, I should have ZERO confidence that the limit will actually be increased. Understand where I am coming from; I saw how this promise was pushed back forever in BTC. I am afraid the same is happening to ADA right now.”
Theory Must Become Practice
When @Cerkoryn retorted that Leios will “fully parallelise consensus” so Cardano can “1000× the number of blocks flowing across the network,” Bons said the mechanism scarcely mattered: “I honestly do not care how you increase TPS—through size, number, or speed of blocks. The bottom line is capacity. If that is achieved through a multi-leader structure, great. Just do it.”
Still, even the sceptic sees a path to conversion. “Sounds great in theory,” he told fellow delegate Dave ADA, who linked to Leios sprint updates. “If it is ever implemented in practice, I will become an ADA supporter overnight! If I believed all of ADA’s promises, I would have been a supporter this entire time. However, I was correct not to believe them, as none of it has happened yet.”
At press time, ADA traded at $0.7173.
