Polymarket is predicting that blobs (~125 kB) will cost ~0.001 ETH.
Today, 125 kB calldata costs ~30 gwei per gas * 16 gas per byte * 125000 gas ~= 0.06 ETH
And if you think polymarket's guess that blobs will be 60x cheaper is over-optimistic, you can use the market to hedge!
Thread from 1.5 years ago on ERC-4337. Still very relevant today, important to understand why it works and how the different parts of the ERC flow directly from the problems that it solves.
@0xMantle has been consistently one of the cheapest ecosystem for TXs
$ETH EIP-4844 upgrade will push L2 gas fees even lower
The lowered gas fees will reduce friction for micro transactions on-chain and chains would be able to pass on cost savings to users.
The low cost environment will greatly benefit dapps like games or prediction markets that rely heavily on micro TXs to function.
Expecting L2s to pivot focus towards TX heavy Dapps soon. Exciting times ahead.
📣 The 11th Ethereum Foundation Research Reddit AMA is starting in 30 minutes, at 1PM UTC today!
Join the conversation and get your questions answered!
🐼🌊🧟✅🧹🎩
#Ethereum#EFResearch#AMA
https://t.co/CI6BO4zN19
📣 The 11th Ethereum Foundation Research Reddit AMA is starting in 30 minutes, at 1PM UTC today!
Join the conversation and get your questions answered!
🐼🌊🧟✅🧹🎩
#Ethereum#EFResearch#AMA
https://t.co/CI6BO4zN19
3.For more blockchain adoption, the Ethereum community would love to come to VISA scales of 2000 Tps. Which options do we have to come closer to this target, besides waiting for full danksharding - which seems far out?
2.(See spec here: https://t.co/bBgUHBpaPw). This would only correspond to 440 tps of simple transfers on L2s, assuming one simple transfer consumes only 71 Byte in verification data.