Frequently asked questions
Technical and commercial questions we hear most often. Don't see yours — ask us directly.
What makes MatchCore different from other matching engines?
MatchCore is the only matching engine that natively runs binary complementary matching, pari-mutuel pooling and CLOB stock & derivatives matching on one core, one cluster and one API surface — most vendors only offer the CLOB model.
What throughput can MatchCore handle?
A single matching segment sustains 150,000+ orders per second at sub-50 microsecond p99 latency. Sharded across a 16-node reference cluster, aggregate throughput reaches 1,000,000 transactions per second. See the full benchmark methodology.
Is MatchCore fully replicated?
Yes. Every state-changing command is committed to a replicated write-ahead log before acknowledgement, and every node — leader or standby — applies the identical deterministic state machine, so failover is a promotion, not a rebuild.
Can MatchCore run multiple matching models in one venue?
Yes. Binary, pari-mutuel and CLOB instruments can be listed side by side in the same cluster, sharing gateways, market data and the risk layer.
What protocols does MatchCore support?
FIX 4.2/4.4, FAST, WebSocket, REST, and a binary SBE protocol for the lowest-latency order entry and market-data path.
Where can MatchCore be deployed?
Bare-metal colocation, private cloud, or major public cloud providers. Topologies range from a single development node to a geo-distributed multi-region cluster.
Is there an audit trail suitable for regulators?
Yes — the write-ahead log is immutable, fully sequenced, and deterministically replayable, giving compliance and surveillance teams exact point-in-time reconstruction of the book.
How do we get pricing?
Pricing depends on throughput tier, deployment topology and matching models required. Contact us for a proposal scoped to your workload.