US 12,541,795 B1Granted

Enhanced ultra low-latency, high-throughput matching engine for electronic trading systems

OwnerJin Seok Yooninventor

Life of this patent

Estimating the term…

Loading patent…

Abstract

A high-speed matching-engine architecture is disclosed that sustains deterministic sub-microsecond latency while processing more than 10 million order messages per second per core on commodity multi-core processors. Orders reside in cache-aligned Data Holder Nodes whose occupancy and price-level boundaries are tracked with constant-time bitmask operations, eliminating pointer-chasing penalties. Per-core huge-page pools, SIMD copy kernels, and lock-free, cache-line-aligned queues further minimize TLB misses and coherence overheads. Overflow is handled by Push Back/Push Forward cascades that relocate the least- or most-prioritized orders between adjoining nodes without violating price-time priority. Node capacities vary monotonically with book depth and are re-tuned online by a lightweight machine-learning controller that maximizes cache-hit probability under changing market micro-structure. The design tightens spreads, raises match-rate revenue, and complies with stringent regulatory latency caps using standard x86-64, Arm, or other architectures.

Claims30

Dependent claims

A high-speed matching engine for processing electronic trade orders with ultra-low latency, implemented as computer-executable instructions stored in non-transitory memory and executed by one or more processors, the matching engine comprising:

(a) a memory-management system configured to allocate a plurality of Data Holder Nodes and at least one Price Level Descriptor;
(b) wherein each Data Holder Node:
(i) is assigned a contiguous memory region that is physically contiguous or virtually contiguous via page mapping and aligned to cache-line boundaries;
(ii) stores a fixed, predetermined maximum number of order objects in contiguously addressable slots within said region, each order object including at least a price value and a quantity value;
and
(iii) includes at least one priority-tracking bitmask that, for each price level represented in the node, indicates whether the node contains the lowest-priority order at that price level and, if present, the particular one of said contiguously addressable slots that holds that order;
(c) at least one ordered price-level index keyed by a price level value, the index comprising one or more Price Level Descriptors;
and, for each Price Level Descriptor, storing:
(i) metadata for the price level;
and
(ii) a reference to the Data Holder Node containing the lowest-priority order at that price level;
(d) at least one core-match unit, each operating in a single-threaded execution context for a financial instrument and configured to:
(i) receive incoming orders;
(ii) match each incoming order against opposite-side resting orders according to a predetermined priority rule;
(iii) execute trades;
(iv) insert any unfilled quantity of the incoming order into one of the Data Holder Nodes;
and
(v) update affected Data Holder Nodes, including the at least one priority-tracking bitmask therein, and the at least one ordered price-level index after each execution, insertion, cancellation, modification, or relocation.

15 more elements

A computer-implemented method for processing electronic trade orders with ultra-low latency in a high-speed matching engine, the method comprising:

(a) allocating, by a memory-management system executed by one or more processors, a plurality of Data Holder Nodes and at least one Price Level Descriptor;
(b) for each Data Holder Node:
(i) assigning a contiguous memory region, each said region physically contiguous or virtually contiguous via page mapping and aligned to cache-line boundaries;
(ii) storing, in contiguously addressable slots within said region, a fixed, predetermined maximum number of order objects, each order object including at least a price value and a quantity value;
and
(iii) maintaining at least one priority-tracking bitmask that, for each price level represented in the node, indicates whether the node contains the lowest-priority order at that price level and, if present, identifies the particular one of said contiguously addressable slots that holds that order;
(c) constructing at least one ordered price-level index keyed by a price level value, the index comprising one or more Price Level Descriptors;
and, for each Price Level Descriptor, storing:
(i) metadata for the price level;
and
(ii) a reference to the Data Holder Node containing the lowest-priority order at that price level;
(d) operating at least one single-threaded core-match unit to:
(i) receive incoming orders;
(ii) match each incoming order against opposite-side resting orders according to a predetermined priority rule;
(iii) execute trades;
(iv) insert any unfilled quantity of the incoming order into one of the Data Holder Nodes;
and
(v) update affected Data Holder Nodes, including the at least one priority-tracking bitmask therein, and the at least one ordered price-level index after each execution, insertion, cancellation, modification, or relocation.

15 more elements

A non-transitory computer-readable storage medium storing computer-executable instructions that, when executed by one or more processors, cause the processors to perform operations for processing electronic trade orders with ultra-low latency in a high-speed matching engine, the operations comprising:

(a) allocating, by a memory-management system executed by one or more processors, a plurality of Data Holder Nodes and at least one Price Level Descriptor;
(b) for each Data Holder Node:
(i) assigning a contiguous memory region, each said region physically contiguous or virtually contiguous via page mapping and aligned to cache-line boundaries;
(ii) storing, in contiguously addressable slots within said region, a fixed, predetermined maximum number of order objects, each order object including at least a price value and a quantity value;
and
(iii) maintaining at least one priority-tracking bitmask that, for each price level represented in the node, indicates whether the node contains the lowest-priority order at that price level and, if present, identifies the particular one of said contiguously addressable slots that holds that order;
(c) constructing at least one ordered price-level index keyed by a price level value, the index comprising one or more Price Level Descriptors;
and, for each Price Level Descriptor, storing:
(i) metadata for the price level;
and
(ii) a reference to the Data Holder Node containing the lowest-priority order at that price level;
and
(d) operating at least one single-threaded core-match unit to:
(i) receive incoming orders;
(ii) match each incoming order against opposite-side resting orders according to a predetermined priority rule;
(iii) execute trades;
(iv) insert any unfilled quantity of the incoming order into one of the Data Holder Nodes;
and
(v) update affected Data Holder Nodes, including the at least one priority-tracking bitmask therein, and the at least one ordered price-level index after each execution, insertion, cancellation, modification, or relocation.

16 more elements

Dependencies are inferred from the claim text.

Facts

Application: 19/260,613Filed 7 Jul 2025 · continuation in part
Priority: 16 Mar 2025
Granted: 3 Feb 2026
Term ends: Not estimatedThe source records don’t support an estimate.

Classification16 codes

G06Q 40/04
Finance; Insurance; Tax strategies; Processing of corporate or income taxes: trading; Exchange, e.g. stocks, commodities, derivatives or currency…
G06F 9/30038
Instructions to perform operations on packed data, e.g. vector, tile or matrix operations: using a mask
G06F 9/3887
Using a plurality of independent parallel functional units: controlled by a single instruction for multiple data lanes [SIMD]
G06F 9/4881
By program, e.g. task dispatcher, supervisor, operating system: scheduling strategies for dispatcher, e.g. round robin, multi-level priority queues