WO2022124010A1 - 演算制御装置、演算制御方法、および記録媒体 - Google Patents
演算制御装置、演算制御方法、および記録媒体 Download PDFInfo
- Publication number
- WO2022124010A1 WO2022124010A1 PCT/JP2021/041830 JP2021041830W WO2022124010A1 WO 2022124010 A1 WO2022124010 A1 WO 2022124010A1 JP 2021041830 W JP2021041830 W JP 2021041830W WO 2022124010 A1 WO2022124010 A1 WO 2022124010A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- arithmetic
- sequence
- arithmetic processing
- quasi
- processing
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims description 37
- 230000014509 gene expression Effects 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 description 12
- 239000013598 vector Substances 0.000 description 9
- 238000002360 preparation method Methods 0.000 description 7
- 238000004364 calculation method Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 230000010365 information processing Effects 0.000 description 3
- 238000007792 addition Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 238000004088 simulation Methods 0.000 description 2
- 230000001771 impaired effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/30003—Arrangements for executing specific machine instructions
- G06F9/30007—Arrangements for executing specific machine instructions to perform operations on data operands
- G06F9/3001—Arithmetic instructions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F7/00—Methods or arrangements for processing data by operating upon the order or content of the data handled
- G06F7/58—Random or pseudo-random number generators
- G06F7/582—Pseudo-random number generators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F7/00—Methods or arrangements for processing data by operating upon the order or content of the data handled
- G06F7/58—Random or pseudo-random number generators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F9/00—Arrangements for program control, e.g. control units
- G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
- G06F9/30—Arrangements for executing machine instructions, e.g. instruction decode
- G06F9/38—Concurrent instruction execution, e.g. pipeline or look ahead
Definitions
- the present invention relates to an arithmetic control device, an arithmetic control method, and a recording medium, and more particularly to an arithmetic control device, an arithmetic control method, and a recording medium that generate a quasi-random number sequence based on a specific sequence.
- Quasi-random numbers are used in various technical fields such as computational physics.
- Related techniques that utilize quasi-random numbers include, for example, physics simulation, computer graphics generation, and pricing of financial derivative products.
- Non-Patent Document 1 describes a method of calculating a Sobol sequence using a Gray code. Further, Patent Document 1 describes a method of generating a quasi-random number sequence based on a generalized Niederreiter sequence. The generalized Niederreiter sequence includes the Faule sequence as well as the Sobol sequence.
- Patent Document 1 describes that a plurality of arithmetic units share and calculate a sequence. More specifically, in Patent Document 1, first, the section of the sequence in charge of each arithmetic unit is defined. Next, each arithmetic unit executes arithmetic processing for calculating a specific sequence in parallel from the starting point of the section of the sequence in charge. As a result, the time required for the arithmetic processing for calculating a specific sequence can be shortened.
- each arithmetic unit executes a matrix operation and a matrix vector product operation in order to obtain an initial value that is a starting point of a sequence section in charge of each arithmetic unit. Since the load of this matrix operation is large, the advantage of parallel arithmetic processing that the processing time for calculating a specific sequence can be shortened is impaired.
- the present invention has been made in view of the above problems, and an object thereof is to shorten the processing time for calculating a specific sequence by parallel arithmetic processing.
- the arithmetic control device is a preparatory processing means for determining the starting point of each arithmetic processing of a plurality of arithmetic units by using a recurrence formula, and a parallel arithmetic processing using the plurality of arithmetic units.
- a parallel arithmetic processing means that calculates a specific sequence by solving a given mathematical expression from the starting point of the arithmetic processing for each arithmetic unit, and a quasi-random sequence that generates a quasi-random sequence based on the specific sequence. It is equipped with a generation means.
- a recurrence formula is used to determine the starting point of each arithmetic processing of a plurality of arithmetic units, and in the parallel arithmetic processing using the plurality of arithmetic units, each arithmetic unit is used.
- a specific sequence is calculated by solving a given mathematical expression from the starting point of the arithmetic processing of the above, and a quasi-random sequence is generated based on the specific sequence.
- the recording medium uses a recurrence formula to determine the starting point of each arithmetic processing of a plurality of arithmetic units, and a parallel arithmetic processing using the plurality of arithmetic units.
- a recurrence formula to determine the starting point of each arithmetic processing of a plurality of arithmetic units, and a parallel arithmetic processing using the plurality of arithmetic units.
- the quasi-random number sequence includes a "lowdiscrepancy sequence”, a “low discrepancy sequence”, and a “low discrepancy sequence”.
- Equation 1 The generalized kneader lighter sequence can be expressed as in Equation 1. (Formula 1)
- b is an integer of 2 or more
- n is a positive integer
- T is an nth-order square matrix having a unique integer component
- ai is an n-dimensional vector having an integer component of 0 or more and less than b.
- G is defined as an nth-order square matrix in which all diagonal components are 1, all upper and sub-diagonal components are -1, and all other components are 0. (Formula 3)
- Equation 4 the vector ai described above has the properties shown in Equation 4.
- ⁇ is the Kronecker delta
- ki is the smallest j that satisfies ai , j ⁇ b-1.
- x'i is defined by the formula 5. (Formula 5)
- Equation 6 is obtained by applying the relational expressions shown in Equations 4 and 5 to Equation 1. (Formula 6)
- T is an nth-order square matrix with n eigenvalue components (n is a positive integer).
- Te ki is the ki column of the matrix T.
- k i is the minimum j ( ⁇ 1) that satisfies a i, j ⁇ b-1.
- x'i + 1 is calculated from the immediately preceding x'i according to Equation 6.
- Parallel arithmetic processing means that a plurality of arithmetic units share a certain arithmetic processing and execute arithmetic processing in parallel. In this case, the starting point of the arithmetic processing by each arithmetic unit is first determined.
- Formula 7 is obtained by applying the relational expressions shown in the above formulas 5 and 6 to the above formula 1. (Formula 7)
- each arithmetic unit uses mathematical expression 7 to determine the starting point of arithmetic processing.
- each arithmetic unit is in charge of d numbers in the sequence expressed by the formula 6 (d is a positive integer).
- FIG. 1 is a block diagram showing a configuration of an arithmetic control device 10 according to the present embodiment.
- the arithmetic control device 10 includes a preparation processing unit 11, a parallel arithmetic processing unit 12 including a plurality of arithmetic units, and a quasi-random number sequence generation unit 13.
- the preparatory processing unit 11 determines the starting point of each arithmetic processing of the plurality of arithmetic units by using a recurrence formula (as an example, the mathematical formula 8 described later).
- a recurrence formula as an example, the mathematical formula 8 described later.
- the preparatory process means a process of finding the starting point of a section of a sequence in which a plurality of arithmetic units are in charge of each. An example of the preparatory process will be described later.
- the preparation processing unit 11 outputs information indicating the starting point of the section of the sequence in charge of each of the plurality of arithmetic units to the parallel arithmetic processing unit 12.
- the parallel arithmetic processing unit 12 calculates a specific sequence by solving a given mathematical expression from the starting point of the arithmetic processing for each arithmetic unit in the parallel arithmetic processing using a plurality of arithmetic units.
- Parallel arithmetic processing means arithmetic processing executed in parallel by a plurality of arithmetic units. An example of parallel arithmetic processing will be described later.
- the parallel arithmetic processing unit 12 outputs the calculated specific sequence of data to the quasi-random number sequence generation unit 13.
- the quasi-random number sequence generation unit 13 generates a quasi-random number sequence based on a specific sequence.
- the specific sequence is the generalized kneader-writer sequence shown in Equation 1.
- the particular sequence may be a Sobol sequence or a Faule sequence.
- the specific sequence is not particularly limited as long as it is used to generate a quasi-random number sequence.
- the data of the quasi-random number sequence may be transmitted to a subsequent stage or an external device (not shown) for performing arithmetic processing (for example, physical simulation) using the quasi-random number sequence.
- Equation 8 is a recurrence formula showing the relationship between the two initial values ( x'i + b m and x'i ). (Formula 8)
- Tem + ki and Tem are the m + kith column and the mth column of the above-mentioned matrix T, respectively. Both em + ki and em are n -dimensional vectors.
- the parallel arithmetic processing unit 12 calculates a specific sequence from the starting point obtained by the previous preparatory processing according to the mathematical formula 6.
- Equation 6 is reprinted below. (Formula 6)
- FIG. 5 shows the flow of arithmetic processing in the related technique.
- the initial value which is the starting point of the arithmetic processing by the p arithmetic instruments is obtained.
- p arithmetic units calculate a specific sequence represented by the mathematical formula 6 according to the mathematical formula 6 from the starting point of each arithmetic processing.
- Equation 7 includes an operation (TG) ai of the matrix vector product.
- the matrix TG contains n ⁇ n components.
- FIG. 2 shows the flow of arithmetic processing in this embodiment.
- the initial value that is the starting point of the arithmetic processing by the p arithmetic units is obtained by the preparatory processing using the mathematical formula 8.
- p arithmetic units execute parallel arithmetic processing.
- the configuration according to the present embodiment is superior to the related technique described in Patent Document 1 in that 2n (p-1) ⁇ 2n 2- It is when n and when p ⁇ n + 1/2.
- the configuration according to the present embodiment is superior to the related technique described in Patent Document 1.
- the conditions described here are just examples. Under different conditions, the situation in which the configuration according to this embodiment becomes dominant may differ.
- FIG. 3 is a flowchart showing a flow of processing executed by each part of the arithmetic control device 10.
- the preparation processing unit 11 uses a recurrence formula to determine the starting point of each arithmetic processing of the plurality of arithmetic units (S1).
- the parallel arithmetic processing unit 12 calculates a specific sequence by solving a given mathematical expression from the starting point of the arithmetic processing for each arithmetic unit using a plurality of arithmetic units (S2).
- the quasi-random number sequence generation unit 13 After that, the quasi-random number sequence generation unit 13 generates a quasi-random number sequence based on a specific sequence (S3).
- the preparatory processing unit 11 uses a recurrence formula to determine the starting point of each arithmetic processing of the plurality of arithmetic units.
- the parallel arithmetic processing unit 12 calculates a specific sequence by solving a given mathematical expression from the starting point of arithmetic processing for each arithmetic unit using a plurality of arithmetic units.
- the quasi-random number sequence generation unit 13 generates a quasi-random number sequence based on a specific sequence.
- a plurality of arithmetic units calculate a specific sequence from each starting point. Perform parallel arithmetic processing to do so.
- the processing time for calculating a specific sequence by the parallel arithmetic processing can be shortened as compared with the configuration in which the plurality of arithmetic units independently calculate the starting points of the parallel arithmetic processing.
- Each component of the arithmetic control device 10 described in the above embodiment shows a block of functional units. Some or all of these components are realized by, for example, the information processing apparatus 900 as shown in FIG. FIG. 4 is a block diagram showing an example of the hardware configuration of the information processing apparatus 900.
- the information processing apparatus 900 includes the following configuration as an example.
- -CPU Central Processing Unit
- ROM Read Only Memory
- RAM Random Access Memory
- -Program 904 loaded into RAM 903
- a storage device 905 that stores the program 904.
- Drive device 907 that reads and writes the recording medium 906.
- -Communication interface 908 for connecting to the communication network 909.
- -I / O interface 910 for inputting / outputting data -Bus 911 connecting each component
- Each component of the arithmetic control device 10 described in the above embodiment is realized by the CPU 901 reading and executing the program 904 that realizes these functions.
- the program 904 that realizes the functions of each component is stored in, for example, a storage device 905 or ROM 902 in advance, and the CPU 901 is loaded into the RAM 903 and executed as needed.
- the program 904 may be supplied to the CPU 901 via the communication network 909, or may be stored in the recording medium 906 in advance, and the drive device 907 may read the program and supply the program to the CPU 901.
- the arithmetic control device 10 described in the above embodiment is realized as hardware. Therefore, it is possible to obtain the same effect as the effect described in the above embodiment.
- the quasi-random number sequence is used, for example, in physics calculations in various science and technology such as computational physics, as well as in the generation of computer graphics and the pricing of financial derivative products.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
- Complex Calculations (AREA)
- Advance Control (AREA)
Abstract
Description
まず、準乱数列を生成するための数列について、説明する。準乱数列を生成するために用いられる数列として、非特許文献1に記載されたソボル列のほかに、ファウレ列(Faure sequence)、また、一般化されたニーダーライター列(generalized Niederreiter sequence)が知られている。
(数式2)
図1は、本実施形態に係わる演算制御装置10の構成を示すブロック図である。図1に示すように、演算制御装置10は、準備処理部11、複数の演算器を含む並列演算処理部12、及び準乱数列生成部13を備えている。
上述した準備処理および並列演算処理の一例について、補足して説明する。
(数式8)
図3を参照して、本実施形態に係わる演算制御装置10の動作を説明する。図3は、演算制御装置10の各部が実行する処理の流れを示すフローチャートである。
本実施形態では、準備処理部11は、漸化式を用いて、複数の演算器のそれぞれの演算処理の起点を決定する。並列演算処理部12は、複数の演算器を用いて、演算器ごとの演算処理の起点から、所与の数式を解くことによって、特定の数列を算出する。準乱数列生成部13は、特定の数列に基づいて、準乱数列を生成する。
前記実施形態で説明した演算制御装置10の各構成要素は、機能単位のブロックを示している。これらの構成要素の一部又は全部は、例えば図4に示すような情報処理装置900により実現される。図4は、情報処理装置900のハードウェア構成の一例を示すブロック図である。
・ROM(Read Only Memory)902
・RAM(Random Access Memory)903
・RAM903にロードされるプログラム904
・プログラム904を格納する記憶装置905
・記録媒体906の読み書きを行うドライブ装置907
・通信ネットワーク909と接続する通信インタフェース908
・データの入出力を行う入出力インタフェース910
・各構成要素を接続するバス911
前記実施形態で説明した演算制御装置10の各構成要素は、これらの機能を実現するプログラム904をCPU901が読み込んで実行することで実現される。各構成要素の機能を実現するプログラム904は、例えば、予め記憶装置905やROM902に格納されており、必要に応じてCPU901がRAM903にロードして実行される。なお、プログラム904は、通信ネットワーク909を介してCPU901に供給されてもよいし、予め記録媒体906に格納されており、ドライブ装置907が当該プログラムを読み出してCPU901に供給してもよい。
11 準備処理部
12 並列演算処理部
13 準乱数列生成部
Claims (6)
- 漸化式を用いて、複数の演算器のそれぞれの演算処理の起点を決定する準備処理手段と、
前記複数の演算器を用いた並列演算処理において、演算器ごとの前記演算処理の起点から、所与の数式を解くことによって、特定の数列を算出する並列演算処理手段と、
前記特定の数列に基づいて、準乱数列を生成する準乱数列生成手段と
を備えた演算制御装置。 - 前記漸化式は、前記複数の演算器のそれぞれの前記演算処理の起点である初期値の間の関係を示す
ことを特徴とする請求項1に記載の演算制御装置。 - 前記特定の数列は、一般化されたニーダーライター列である
ことを特徴とする請求項1または2に記載の演算制御装置。 - 前記準乱数列は、超一様分布列、低食い違い量列、および低ディスクレパンシ列を含む
ことを特徴とする請求項1から3のいずれか1項に記載の演算制御装置。 - 漸化式を用いて、複数の演算器のそれぞれの演算処理の起点を決定し、
前記複数の演算器を用いた並列演算処理において、演算器ごとの前記演算処理の起点から、所与の数式を解くことによって、特定の数列を算出し、
前記特定の数列に基づいて、準乱数列を生成する
演算方法。 - 漸化式を用いて、複数の演算器のそれぞれの演算処理の起点を決定する処理と、
前記複数の演算器を用いた並列演算処理において、演算器ごとの前記演算処理の起点から、所与の数式を解くことによって、特定の数列を算出する処理と、
前記特定の数列に基づいて、準乱数列を生成する処理と
をコンピュータに実行させるためのプログラムを格納した、一時的でない記録媒体。
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2022568132A JPWO2022124010A5 (ja) | 2021-11-15 | 演算制御装置、演算方法、およびプログラム | |
US18/039,471 US20230418601A1 (en) | 2020-12-07 | 2021-11-15 | Arithmetic and control device, arithmetic and control method, and recording medium |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2020-202595 | 2020-12-07 | ||
JP2020202595 | 2020-12-07 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022124010A1 true WO2022124010A1 (ja) | 2022-06-16 |
Family
ID=81974343
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2021/041830 WO2022124010A1 (ja) | 2020-12-07 | 2021-11-15 | 演算制御装置、演算制御方法、および記録媒体 |
Country Status (2)
Country | Link |
---|---|
US (1) | US20230418601A1 (ja) |
WO (1) | WO2022124010A1 (ja) |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO1996018144A1 (fr) * | 1994-12-05 | 1996-06-13 | International Business Machines Corporation | DISPOSITIF ET PROCEDE DE PRODUCTION QUASI-ALEATOIRE DE NOMBRES, ET DISPOSITIF ET PROCEDE D'INTEGRATION MULTIPLE DE LA FONCTION f |
-
2021
- 2021-11-15 US US18/039,471 patent/US20230418601A1/en active Pending
- 2021-11-15 WO PCT/JP2021/041830 patent/WO2022124010A1/ja active Application Filing
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO1996018144A1 (fr) * | 1994-12-05 | 1996-06-13 | International Business Machines Corporation | DISPOSITIF ET PROCEDE DE PRODUCTION QUASI-ALEATOIRE DE NOMBRES, ET DISPOSITIF ET PROCEDE D'INTEGRATION MULTIPLE DE LA FONCTION f |
Non-Patent Citations (1)
Title |
---|
TEZUKA, SHU; FUSHIMI, MASANORI: "Multiple Integration using Low-Discrepancy Sequences Generated by Fibonacci Polynomials", IPSJ SIG NOTES, vol. 92, no. 26, 21 March 1992 (1992-03-21), JP , pages 27 - 32, XP009537614, ISSN: 0919-6072 * |
Also Published As
Publication number | Publication date |
---|---|
US20230418601A1 (en) | 2023-12-28 |
JPWO2022124010A1 (ja) | 2022-06-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP4152215A1 (en) | Quantum state preparation circuit generation method and apparatus, chip, device, and program product | |
CN112884153B (zh) | 处理数据的方法和相关装置 | |
AU2022283685A1 (en) | Quantum neural network training method and system, electronic device, and medium | |
CN111914378B (zh) | 一种单振幅量子计算模拟方法及装置 | |
US20190138922A1 (en) | Apparatus and methods for forward propagation in neural networks supporting discrete data | |
CN115577791B (zh) | 基于量子系统的信息处理方法及装置 | |
Lloyd et al. | A cubic differential system with nine limit cycles | |
CN115759269B (zh) | 特征信息的确定方法、装置、设备以及存储介质 | |
WO2020223850A1 (en) | System and method for quantum circuit simulation | |
Chang et al. | Least-squares solutions to polynomial systems of equations with quantum annealing | |
WO2023103754A1 (zh) | 量子体系下的热化态制备方法、设备及存储介质 | |
US11526761B2 (en) | Neural network training with decreased memory consumption and processor utilization | |
JP2018163396A (ja) | 区分線形近似関数生成装置および方法 | |
WO2022124010A1 (ja) | 演算制御装置、演算制御方法、および記録媒体 | |
US11080365B2 (en) | Solving lattice problems using annealing | |
JP7091930B2 (ja) | テンソルデータ計算装置、テンソルデータ計算方法及びプログラム | |
CN114021729B (zh) | 量子电路操作方法及系统、电子设备和介质 | |
CN116932991A (zh) | 一种形式化的多标量乘法分析与计算加速方法 | |
Wicaksono et al. | Implementation of Shor’s quantum factoring algorithm using projectQ framework | |
JP6067596B2 (ja) | ペアリング演算装置、マルチペアリング演算装置、プログラム | |
Larasati et al. | Simulation of modular exponentiation circuit for shor's algorithm in qiskit | |
US20190073584A1 (en) | Apparatus and methods for forward propagation in neural networks supporting discrete data | |
US9355363B2 (en) | Systems and methods for virtual parallel computing using matrix product states | |
CN116340714B (zh) | 一种张量数据处理方法及相关装置 | |
US20230281267A1 (en) | Calculation device, calculation program, recording medium, and calculation method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 21903114 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 18039471 Country of ref document: US |
|
ENP | Entry into the national phase |
Ref document number: 2022568132 Country of ref document: JP Kind code of ref document: A |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 21903114 Country of ref document: EP Kind code of ref document: A1 |