BNB Chain’s Large Block Strategy
Similar to other exchange-backed public chains like Solana and Heco, BNB Smart Chain (BSC) has long pursued high performance. Since its launch in 2020, BSC set each block’s gas limit at 30 million with a stable block time of 3 seconds. These parameters enabled BSC to achieve a TPS cap of 100+ (combining various transaction types). In June 2021, BSC increased its block gas limit to 60 million. However, when the blockchain game CryptoBlades exploded in popularity in July 2021, daily transactions exceeded 8 million, causing fees to spike. This revealed BSC’s efficiency bottlenecks.
(Source: BscScan)
To address network performance, BSC gradually raised the block gas limit, stabilizing around 80–85 million. By September 2022, it reached 120 million and peaked at 140 million by year-end—nearly 5x its 2020 capacity. Although BSC considered a 300 million gas limit, concerns over validator node load prevented implementation.
(Source: YCHARTS)
BNB Chain later shifted focus to modular/Layer 2 solutions rather than Layer 1 scaling. Projects like zkBNB (launched late 2022) and GreenField (early 2023) reflect this trend. This article explores opBNB, comparing its performance with Ethereum Layer 2 to uncover Rollup bottlenecks.
How BSC’s High Throughput Benefits opBNB’s DA Layer
Modular blockchains like Celestia categorize key components as:
- Execution: Processes smart contracts and state transitions.
- Settlement: Handles fraud proofs and L1-L2 bridging.
- Consensus: Orders transactions.
- Data Availability (DA): Publishes ledger data for validators.
DA and consensus layers are often coupled. For optimistic Rollups, DA data includes L2 transaction sequences, allowing full nodes to verify order. However, Ethereum’s low DA throughput (150 KB/s max) severely limits Rollup TPS, forcing L2s to throttle activity or use alternative DA layers like Celestia.
opBNB leverages BSC’s high-throughput DA:
- Ethereum: Processes ~150 KB/s (300M gas/block, 12s block time).
- BSC: Handles ~2,910 KB/s (140M gas/block, 3s block time)—18.6x Ethereum’s capacity.
Rollup Data Publishing Methods:
- Arbitrum: Sequencer EOAs post batched L2 transactions via contract calls (calldata).
- Optimism/opBNB: Sequencer EOAs send transfers with embedded L2 data.
DA Constraints:
- Ethereum’s 150 KB/s cap limits optimistic Rollups to ~2,000 TPS (after compression).
- Post-EIP1559, blocks average 50% capacity, further reducing TPS.
- EIP-4844 lowers fees but doesn’t significantly increase throughput.
BNB Chain Advantages:
- Faster finality: BSC blocks finalize in 45s (vs. ~3 minutes for Ethereum).
- Lower fees: BSC gas prices are 10–50x cheaper than Ethereum’s.
(Example: $0.004 fees on opBNB vs. $0.21 on Optimism for similar transactions.)
opBNB’s Execution Layer Boost: Cache Optimization
Beyond DA, execution layer speed is critical. Key bottlenecks:
- EVM’s single-threaded execution underutilizes CPU.
- Merkle Patricia Trie inefficiencies slow data retrieval.
opBNB’s Optimizations:
- Three-Tier Caching: Stores frequently accessed state data closer to the EVM.
- State Pre-Reading: Idle CPU cores prefetch future transaction data, reducing disk reads.
Performance Metrics:
- 1B gas/s processing capacity (EVM’s theoretical max).
- 4,761 simple transfers/s (vs. ~500–1,000 swaps/s).
- 40x Ethereum’s TPS, 6x Optimism’s.
FAQ
1. Why does Ethereum’s DA layer limit Rollup TPS?
Ethereum’s low throughput (150 KB/s) forces Rollups to compress data and batch transactions, capping TPS.
2. How does opBNB achieve higher TPS than Ethereum L2s?
By using BSC’s high-throughput DA (18.6x Ethereum’s capacity) and optimizing execution via caching/pre-reading.
3. What’s the trade-off for faster DA layers like BSC?
Lower fees and faster finality, but execution layer speed becomes the next bottleneck.
👉 Explore more about high-performance Rollups