Morpho runs SQD Portal in production
Morpho, one of DeFi's largest lending protocols (~$16.5B in deposits and active loans), runs a dedicated SQD Portal in production. Here is how it performs, measured on their own production traffic.
"Morpho holds its infrastructure to a high standard, and supporting a team of this calibre is exactly what SQD Portal was built for."
The most correct data, by construction
Every block is validated at ingestion before it ever reaches an indexer. RPC endpoints pass through whatever the node returns.
SQD applies 6 cryptographic checks per block: multi-source RPC cross-checking, transaction-root reconstruction, and log-bloom verification among them. In Morpho's own production benchmark, SQD returned ≈100% correct data. RPC tiers returned 85 to 93%, as many as 1 in 7 requests wrong.
SQD returns 100% correct data. The major RPC provider fails 1 in 14 requests. The edge gateway fails 1 in 7.
Premium RPCs match our speed but fail 5–15% of requests. SQD delivers both, under 50ms at P90 with zero errors.
SQD degrades 1.8x from median to P90. Others degrade 9-10x, that's the difference between consistent and unreliable.
A dedicated, validated data layer
Private to Morpho, fully managed, and validated at ingestion. No nodes to run, no providers to babysit.
Ethereum, Base, Arbitrum, Hyperliquid and more, kept at the head.
Dedicated Portal Every block passes 6 cryptographic checks at ingestion before anything downstream sees it.
Morpho's indexers consume validated onchain data, at production scale.
The same Portal any team can run: dedicated infrastructure, validated data, fully managed by SQD.
Production-grade, measured in production
Measured on Morpho's dedicated Portal over a rolling 90-day window.
Consistency is reliability. SQD degrades 1.8× from median to P90, where RPC tiers degrade 9 to 10×. That is the difference between predictable and unpredictable under load.
Built for historical and batch workloads
Where indexers actually spend their time: backfills and large batched reads.
Base mainnet. Single small requests are roughly at parity.
Internal benchmark, 2026-06-02.
Portal vs raw RPC, for indexing
Scoped to the workload Morpho runs: validated historical and batched reads.
| Dimension | SQD | Raw RPC |
|---|---|---|
| Correctness | ||
| Data validation | 6-step cryptographic checks at ingestion | Raw, unchecked node output |
| Accuracy | ≈100% | 85 to 93% |
| Performance | ||
| Large batched reads Base, 10k request batches. | Up to 41× faster | Baseline |
| Historical backfill | Beats a dedicated node | Node-bound |
| Decoded data | ✓ | ✗ |
| Coverage | ||
| Networks served | 120+ networks | Per-provider |
| Operations | ||
| Node operations | None (managed) | You run or rent nodes |
Want this for your protocol?
A dedicated Portal, validated data, and enterprise support: the same setup Morpho runs, deployed for your team.
in Morpho's benchmark
vs premium RPC
rolling 90 days
Sources: Morpho production correctness and latency benchmark (2026); dedicated Portal SLO telemetry, rolling 90 days as of 2026-06-30; public batch benchmark (Base, 2026-01-25); internal historical sync benchmark (2026-06-02). Figures are point-in-time snapshots of production telemetry.