TWI874161B - A field-programmable logical gate array that aggregates financial product trading information - Google Patents
A field-programmable logical gate array that aggregates financial product trading information Download PDFInfo
- Publication number
- TWI874161B TWI874161B TW113108927A TW113108927A TWI874161B TW I874161 B TWI874161 B TW I874161B TW 113108927 A TW113108927 A TW 113108927A TW 113108927 A TW113108927 A TW 113108927A TW I874161 B TWI874161 B TW I874161B
- Authority
- TW
- Taiwan
- Prior art keywords
- message
- unit
- order
- gate array
- hash
- Prior art date
Links
Images
Landscapes
- Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)
Abstract
一種彙整金融商品交易行情的現場可程式化邏輯閘陣列,包含一收發單元,一與該收發單元資訊連接的運算單元,及一與該運算單元資訊連接的訂單簿記憶單元。該收發單元與一集中市場伺服器通訊,該收發單元接收該集中市場伺服器發佈的一訂單資料。該運算單元自該收發單元取得該訂單資料,該運算單元擷取該訂單資料中的一訂單號訊息,並對該訂單號訊息進行運算取得一雜湊訊息,該雜湊訊息的資料量小於該訂單號訊息的資料量,該運算單元以該雜湊訊息作為一邏輯位址將該訂單資料記錄於一訂單簿記憶單元中。A field programmable logic gate array for aggregating financial product trading market conditions includes a transceiver unit, an operation unit connected to the transceiver unit, and an order book storage unit connected to the operation unit. The transceiver unit communicates with a centralized market server and receives order data published by the centralized market server. The operation unit obtains the order data from the transceiver unit, extracts an order number message from the order data, and operates the order number message to obtain a hash message, the data volume of the hash message is less than the data volume of the order number message, and the operation unit uses the hash message as a logical address to record the order data in an order book memory unit.
Description
本發明涉及一種用於金融交易的現場可程式化邏輯閘陣列(FPGA),尤指一種可彙整金融商品交易行情的現場可程式化邏輯閘陣列。The present invention relates to a field programmable logic gate array (FPGA) used for financial transactions, and in particular to a field programmable logic gate array that can aggregate financial product transaction prices.
集中市場以組播(Multicast)或快照(Snapshot)即時向市場參與者發佈市場行情數據(Maket date),前述市場行情數據包含訂單資料。市場參與者可透過運算終端上執行的軟體來對市場行情數據進行解析,進而參與交易。The centralized market releases market data (Market date) to market participants in real time via multicast or snapshot. The aforementioned market data includes order data. Market participants can analyze the market data through the software running on the computing terminal and then participate in transactions.
高頻交易(High Frequency Trading,HFT)是透過高算力的終端設備來實施,透過高算力終端設備來解析市場行情數據,進而得以在快速變化的交易市場中尋求價格差實現套利。高頻交易對於市場行情數據運算延遲速度十分講究,延遲幾微秒即可能喪失先機。然,當前市場行情數據的解讀運算仍是以軟解方式實施,由終端設備中的中央處理器(CPU)執行,中央處理器運算前除需經過網路卡等設備的運算,運算時更需搭配相應的軟體實施,如此一來仍會產生一定運算延遲,不利高頻交易。High Frequency Trading (HFT) is implemented through high-computing terminal equipment, which analyzes market data and seeks price differences in the rapidly changing trading market to achieve arbitrage. High-frequency trading is very particular about the delay speed of market data calculation. A delay of a few microseconds may lose the opportunity. However, the current market data interpretation and calculation is still implemented in software, which is executed by the central processing unit (CPU) in the terminal device. Before the central processor calculates, it needs to go through the calculation of network cards and other devices, and the calculation needs to be implemented with corresponding software. This will still cause a certain calculation delay, which is not conducive to high-frequency trading.
本發明的主要目的,在於解決當前高頻交易仍以軟解方式實施仍有運算延遲較長的問題。The main purpose of the present invention is to solve the problem that the current high-frequency trading is still implemented in a soft solution manner and still has a long computational delay.
為達上述目的,本發明提供一種彙整金融商品交易行情的現場可程式化邏輯閘陣列,包含一收發單元,一與該收發單元資訊連接的運算單元,及一與該運算單元資訊連接的訂單簿記憶單元。該收發單元與一集中市場伺服器通訊,該收發單元接收該集中市場伺服器發佈的一訂單資料。該運算單元自該收發單元取得該訂單資料,該運算單元擷取該訂單資料中的一訂單號訊息,並對該訂單號訊息進行運算取得一雜湊訊息,該雜湊訊息的資料量小於該訂單號訊息的資料量。該運算單元以該雜湊訊息作為一邏輯位址將該訂單資料記錄於一訂單簿記憶單元中,該訂單資料包含該訂單號訊息,一價格訊息及一數量訊息。To achieve the above-mentioned purpose, the present invention provides a field programmable logic gate array for aggregating financial product trading market conditions, comprising a transceiver unit, an operation unit connected to the transceiver unit, and an order book storage unit connected to the operation unit. The transceiver unit communicates with a centralized market server, and the transceiver unit receives an order data issued by the centralized market server. The operation unit obtains the order data from the transceiver unit, extracts an order number message from the order data, and operates the order number message to obtain a hash message, and the data volume of the hash message is less than the data volume of the order number message. The operation unit uses the hash information as a logic address to record the order data in an order book memory unit, and the order data includes the order number information, a price information and a quantity information.
一實施例中,該現場可程式化邏輯閘陣列包含一行情記憶單元,該運算單元記錄該訂單資料時亦將該價格訊息及該數量訊息記錄於該行情記憶單元。In one embodiment, the field programmable logic gate array includes a market information storage unit, and the operation unit also records the price information and the quantity information in the market information storage unit when recording the order data.
一實施例中,該運算單元以至少二雜湊算式與至少二變數作為運算元,對該訂單號訊息運算,產生複數待選用雜湊訊息,該運算單元以該些待選用雜湊訊息查找該訂單簿記憶單元相應的記憶位址是否已使用,如該些待選用雜湊訊息其中一者所對應記憶位址未被使用,即以該待選用雜湊訊息作為該邏輯位址。In one embodiment, the operation unit uses at least two hash expressions and at least two variables as operators to operate on the order number information to generate a plurality of hash messages to be selected. The operation unit uses these hash messages to be selected to find out whether the corresponding memory address of the order book memory unit has been used. If the memory address corresponding to one of the hash messages to be selected is not used, the hash message to be selected is used as the logical address.
一實施例中,該運算單元於該些待選用雜湊訊息所對應記憶位址均已被使用時,讀取該些待選用雜湊訊息其中一者所對應記憶位址目前記錄的該訂單資料進行運算,變更其記憶位址。In one embodiment, when the memory addresses corresponding to the hash messages to be selected have all been used, the calculation unit reads the order data currently recorded in the memory address corresponding to one of the hash messages to be selected, performs calculations, and changes its memory address.
透過本發明前述實施,相較於習用具有以下特點:本發明令原本資料量不易被該現場可程式化邏輯閘陣列運算的該訂單資料,可快速被記錄於該訂單簿記憶單元,使該現場可程式化邏輯閘陣列可依此快速反應市場行情,解決過往資料處理延遲過長,不利高頻交易的問題。除前述外,該現場可程式化邏輯閘陣列本身的記憶空間有限,本發明透過前段描述程序可有效運用記憶空間。Through the aforementioned implementation of the present invention, compared with the conventional method, the present invention has the following characteristics: the present invention allows the order data, which is originally too large to be calculated by the field programmable logic gate array, to be quickly recorded in the order book memory unit, so that the field programmable logic gate array can quickly respond to market conditions, solving the problem of long data processing delays in the past, which is not conducive to high-frequency trading. In addition to the aforementioned, the field programmable logic gate array itself has limited memory space, and the present invention can effectively use the memory space through the process described in the previous paragraph.
本發明詳細說明及技術內容,茲配合圖式說明如下:The detailed description and technical contents of the present invention are as follows with reference to the accompanying drawings:
請參閱圖1,本發明提供一種現場可程式化邏輯閘陣列100(Field Programmable Gate Array,縮寫為FPGA),該現場可程式化邏輯閘陣列用於金融交易領域,且可彙整金融商品交易行情。該現場可程式化邏輯閘陣列100可配置於一伺服設備中,該現場可程式化邏輯閘陣列100實現硬體加速,令資訊可於資料鏈路層內完成處理,以因應高頻金融交易需求。Please refer to FIG. 1 . The present invention provides a field programmable gate array 100 (Field Programmable Gate Array, abbreviated as FPGA), which is used in the field of financial transactions and can summarize financial product transaction trends. The field
承上,該現場可程式化邏輯閘陣列100包含一收發單元11,一與該收發單元11資訊連接的運算單元12,及一與該運算單元12連接的訂單簿記憶單元13。其中,該收發單元11為該現場可程式化邏輯閘陣列100與外部聯絡的窗口。進一步來說,該收發單元11透過一網際網路20(圖中未示)與一終端設備資訊連接,該終端設備可為一集中市場伺服器21,該收發單元11與該集中市場伺服器21建立通訊,以接收該集中市場伺服器21發佈的一訂單資料211。前述該集中市場伺服器21可為那斯達克股票交易所(NASDAQ)、日本交易所集團(JPX)、新加坡交易所(SGX),芝加哥商品交易所(CME)、香港證卷交易所(HKEX)或歐洲期貨交易所(EUREX)所有。As mentioned above, the FPLG 100 includes a
又,該運算單元12為該現場可程式化邏輯閘陣列100經編程執行運算的部分,該運算單元12可基於一控制程序記憶單元14儲存的複數程序運算。前述該控制程序記憶單元14是包含於該現場可程式化邏輯閘陣列100內,該控制程序記憶單元14資訊連接該運算單元12。Furthermore, the
併請參閱圖1與圖2,本發明該現場可程式化邏輯閘陣列100的該運算單元12自該收發單元11取得解碼後的該訂單資料211,該訂單資料211是針對特定金融商品所發佈。該訂單資料211至少包含該一訂單號訊息212、一價格訊息213、及一數量訊息214。該運算單元12擷取該訂單資料211中的該訂單號訊息212,並對該訂單號訊息212進行運算取得一雜湊訊息121。該雜湊訊息121的資料量小於該訂單號訊息212的資料量,具體來說,於那斯達克股票交易所(NASDAQ)所規範的資料格式中,該訂單號訊息212為32位元(Bit),本發明該雜湊訊息121於一實施例中為11位元(Bit)。該運算單元12以該雜湊訊息121作為一邏輯位址,將該訂單資料211記錄於該訂單簿記憶單元13。Referring to FIG. 1 and FIG. 2 , the
本發明該現場可程式化邏輯閘陣列100透過前段描述程序,令原本資料量不易被該現場可程式化邏輯閘陣列100運算的該訂單資料211,快速被記錄於該訂單簿記憶單元13,使該現場可程式化邏輯閘陣列100可依此快速反應市場行情,解決過往資料處理延遲過長,不利高頻交易的問題。除前述外,該現場可程式化邏輯閘陣列100本身的記憶空間有限,本發明透過前段描述程序可有效運用記憶空間。The field programmable
復請參閱圖1與圖2,一實施例中,該現場可程式化邏輯閘陣列100包含一行情記憶單元15,該行情記憶單元15資訊連接該運算單元12。該運算單元12以該雜湊訊息121作為該邏輯位址記憶該訂單資料211時,將該價格訊息213與該數量訊息214記錄於該行情記憶單元15。由於,該行情記憶單元15僅記錄價格與數量,簡化了運算資訊量,使該現場可程式化邏輯閘陣列100於實現高頻交易過程中,該運算單元12可基於該行情記憶單元15內資料運算,快速反應當前市場情緒。Referring again to FIG. 1 and FIG. 2 , in one embodiment, the field programmable
請搭配參閱圖3,一實施例中,該運算單元12以至少二雜湊算式40及至少二變數41作為運算元,對該訂單號訊息212運算,產生複數待選用雜湊訊息122。該運算單元12以該些待選用雜湊訊息122查找該訂單簿記憶單元13相應的記憶位址是否已使用,如該些待選用雜湊訊息122其中一者所對應記憶位址未被使用,即以該待選用雜湊訊122息作為該邏輯位址。進一步地,該運算單元12於該些待選用雜湊訊息122所對應記憶位址均已被使用時,讀取該些待選用雜湊訊息122其中一者所對應記憶位址目前記錄的該訂單資料211進行運算,變更其記憶位址。如變更計算後仍有記憶位址上的衝突,則再次執行前述運算,直至無衝突為止。Please refer to FIG. 3 , in one embodiment, the
請搭配參閱圖1,一實施例中,該運算單元12執行有一策略交易程序,該策略交易程序相關參數記憶於該控制程序記憶單元14。該運算單元12執行該策略交易程序時,可經該收發單元11向該集中市場伺服器發佈訂單,該運算單元12可基於該行情記憶單元15內資料判斷價量分佈及市場情緒,進行策略投資。此舉相較於過往,資料均須經中央處理器(CPU)運算後才進行策略投資的方式,運算延遲明顯減少,能更有效地增加策略投資的成功率。Please refer to FIG. 1 . In one embodiment, the
100:現場可程式化邏輯閘陣列 11:收發單元 12:運算單元 121:雜湊訊息 122:待選用雜湊訊息 13:訂單簿記憶單元 14:控制程序記憶單元 15:行情記憶單元 20:網際網路(圖中未示) 21:集中市場伺服器 211:訂單資料 212:訂單號訊息 213:價格訊息 214:數量訊息 40:雜湊算式 41:變數 100: Field programmable logic gate array 11: Transceiver unit 12: Operation unit 121: Hash message 122: Hash message to be selected 13: Order book memory unit 14: Control program memory unit 15: Market memory unit 20: Internet (not shown) 21: Centralized market server 211: Order data 212: Order number message 213: Price message 214: Quantity message 40: Hash formula 41: Variable
圖1,為本發明現場可程式化邏輯閘陣列一實施例的單元組成示意圖。 圖2,為本發明訂單資料一實施例的組成示意圖。 圖3,為本發明現場可程式化邏輯閘陣列一實施例的實施示意圖。 Figure 1 is a schematic diagram of the unit composition of an embodiment of the field programmable logic gate array of the present invention. Figure 2 is a schematic diagram of the composition of an embodiment of the order data of the present invention. Figure 3 is a schematic diagram of the implementation of an embodiment of the field programmable logic gate array of the present invention.
100:現場可程式化邏輯閘陣列 100: Field programmable logic gate array
11:收發單元 11: Transceiver unit
12:運算單元 12: Arithmetic unit
13:訂單簿記憶單元 13: Order book memory unit
14:控制程序記憶單元 14: Control program memory unit
15:行情記憶單元 15: Market memory unit
21:集中市場伺服器 21: Centralized Market Server
211:訂單資料 211: Order information
Claims (4)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW113108927A TWI874161B (en) | 2024-03-12 | 2024-03-12 | A field-programmable logical gate array that aggregates financial product trading information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| TW113108927A TWI874161B (en) | 2024-03-12 | 2024-03-12 | A field-programmable logical gate array that aggregates financial product trading information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| TWI874161B true TWI874161B (en) | 2025-02-21 |
| TW202536709A TW202536709A (en) | 2025-09-16 |
Family
ID=95557469
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| TW113108927A TWI874161B (en) | 2024-03-12 | 2024-03-12 | A field-programmable logical gate array that aggregates financial product trading information |
Country Status (1)
| Country | Link |
|---|---|
| TW (1) | TWI874161B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110781511A (en) * | 2019-09-10 | 2020-02-11 | 深圳壹账通智能科技有限公司 | Transaction information storage method, device, computer equipment and storage medium |
| TWI729682B (en) * | 2020-01-21 | 2021-06-01 | 元大期貨股份有限公司 | Network communication device for realizing high-frequency issuance of transaction data |
| CN115187403A (en) * | 2022-07-26 | 2022-10-14 | 上海全创信息科技有限公司 | Securities Quote Forwarding Method and System, Computer-readable Storage Medium and Device |
| TWM634399U (en) * | 2022-05-19 | 2022-11-21 | 富邦期貨股份有限公司 | Features immediately trading system |
| CN117520260A (en) * | 2023-11-20 | 2024-02-06 | 华东师范大学 | FPGA-based template securities trading quotation analysis system |
-
2024
- 2024-03-12 TW TW113108927A patent/TWI874161B/en active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110781511A (en) * | 2019-09-10 | 2020-02-11 | 深圳壹账通智能科技有限公司 | Transaction information storage method, device, computer equipment and storage medium |
| TWI729682B (en) * | 2020-01-21 | 2021-06-01 | 元大期貨股份有限公司 | Network communication device for realizing high-frequency issuance of transaction data |
| TWM634399U (en) * | 2022-05-19 | 2022-11-21 | 富邦期貨股份有限公司 | Features immediately trading system |
| CN115187403A (en) * | 2022-07-26 | 2022-10-14 | 上海全创信息科技有限公司 | Securities Quote Forwarding Method and System, Computer-readable Storage Medium and Device |
| CN117520260A (en) * | 2023-11-20 | 2024-02-06 | 华东师范大学 | FPGA-based template securities trading quotation analysis system |
Also Published As
| Publication number | Publication date |
|---|---|
| TW202536709A (en) | 2025-09-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10614523B2 (en) | Method and apparatus for order entry in an electronic trading system | |
| US7908213B2 (en) | System and method for improving electronic trading | |
| KR101667697B1 (en) | Synchronized Processing of Data By Networked Computing Resources | |
| JP6045505B2 (en) | Method and apparatus for managing orders in a financial market | |
| US20110264578A1 (en) | Automated trading system | |
| US11887191B2 (en) | On-demand auction | |
| US12198194B1 (en) | Message randomization and delay based on order type in an electronic trading system | |
| WO2019227609A1 (en) | Blockchain-based smart stock purchase method, device, apparatus and storage medium | |
| TWI874161B (en) | A field-programmable logical gate array that aggregates financial product trading information | |
| WO2023017476A1 (en) | Pipelined credit checking | |
| TWI841061B (en) | Transaction risk control device | |
| CN117853110A (en) | Wind control transaction realization method, device and storage medium | |
| TWI761236B (en) | Transaction risk control device | |
| JP2002109226A (en) | Derivative for enhancing operation efficiency in tokyo stock price index (topix) futures and its commercialization method | |
| US20250112881A1 (en) | Optimizations for non-blocking messages | |
| CN115456787A (en) | A transaction data processing method, device, equipment and medium | |
| TWI867649B (en) | Low-latency financial transaction publishing system | |
| CN111798318B (en) | An extremely fast position management method and system | |
| CN111489258A (en) | Data processing system, method, electronic device, and storage medium | |
| WO2025126138A1 (en) | Optimizations for non-blocking messages | |
| JP2026504385A (en) | Method, apparatus, and system for implementing orders with minimal latency | |
| EP4612637A1 (en) | Systems and methods for deriving synthetic assets and atomic transaction processing for the same | |
| CN118505396A (en) | Investment management method and related equipment |