Benchmarking BSC Block Discovery: bloXroute vs. BlockRazor (Frankfurt)

Trading on BNB Smart Chain is brutally competitive. If your infrastructure lags even a few milliseconds, you lose edge.

Most teams obsess over mempool speed - but block discovery is just as important. New blocks are the canonical truth: state changes, newly confirmed opportunities, and the final outcome of everything that happened in the public mempool, along with all private transactions and bundles. If you see the next block late, you’re reacting late.

We benchmarked bloXroute’s bdnBlocks stream against BlockRazor’s NewBlocks stream in Frankfurt. The results show that bloXroute consistently outperforms on block discovery and propagation in Frankfurt. Below is our testing methodology and what we observed.

Why block discovery speed matters for searchers

On BSC, blocks arrive roughly every ~450ms. That’s fast enough that even “small” delays are meaningful.

Seeing a new block first helps you:

• Update internal state faster (balances, pool reserves, nonce expectations)
• React sooner to newly confirmed arbitrage / liquidity conditions
• Reduce stale assumptions in your simulation pipeline
• Build and submit follow-up actions earlier in the next block window

In practice: late block awareness becomes late decision-making, and late decision-making becomes lost edge.

Benchmark setup

We designed this benchmark to mirror what a real searcher cares about: which feed delivers the full block first to a consumer in-region.

Streams compared in Frankfort region

• bloXroute: bdnBlocks 
• BlockRazor: NewBlocks 

Fairness controls

• Both streams consumed from the same server / same machine
• Same region (Frankfurt)
• Same timestamping and comparison logic
• Same observation window: 7 days
• All blocks in the period included (no sampling)

Results: Overall win rate in Frankfurt

Across the 7-day window between 2026-02-18 00:00:00 and  2026-02-25 00:00:00:

bloXroute delivered 76.6% of blocks first.

We measured the arrival-time difference between streams (in milliseconds) for blocks where one side won.

When bloXroute wins (arrival-time advantage)

• P10: 0.73 ms
• P50: 4.01 ms
• P90: 11.50 ms

Interpretation:

• In a typical win (median), bloXroute is ahead by ~4ms
• In the faster-tail wins, bloXroute is ahead by 10ms+
• Those margins are large enough to change ordering and outcomes for time-sensitive strategies

When BlockRazor wins (arrival-time advantage)

• P10: 0.02 ms
• P50: 2.38 ms
• P90: 5.94 ms

Interpretation:

• Even when BlockRazor wins, the margin is generally smaller
• bloXroute’s advantage is both more frequent and more decisive

What this means in practice for BSC searchers

This benchmark points to two practical implications:

1) You see more blocks first

Seeing ~3 out of every 4 blocks first compounds quickly over a trading day. Even if your strategy only depends on “being earliest” some of the time, more frequent first delivery translates into more first-mover windows.

2) When you win, you win by enough to matter

A 1-2ms lead is nice. A 4-12ms lead is decisive - especially on BSC where downstream processing (decode → simulate → decide → submit) already consumes time. A consistent multi-millisecond head start gives your pipeline more room to operate before the next block cutoff.

Upgrade your block discovery

If your strategy depends on being first, your infrastructure should too.

Upgrade your BSC block discovery with bloXroute and give your simulation and submission pipeline the head start it needs.

To chat with our team, feel free to book a meeting.