-
公开(公告)号:US4210961A
公开(公告)日:1980-07-01
申请号:US943695
申请日:1978-09-19
Applicant: Duane L. Whitlow , Azra Sasson
Inventor: Duane L. Whitlow , Azra Sasson
CPC classification number: G06F7/22 , Y10S707/99937
Abstract: A method is disclosed of sorting data in an electronic data processing system utilizing a digital computer and at least one random access device such as, for example, a magnetic disc or drum unit. A first simulation is performed to optimize the utilization of the resources of the data processing system. This simulation is based on information relating to characteristics of the data processing system available to the sorting function as well as known or predicted information concerning the data to be sorted. The input data is ordered into a plurality of strings of data which are generated by a selection technique known as quadratic selection replacement. The number of strings generated is minimized by determining if data bias exists, and if unfavorable or negative bias is discovered, the order of string generation is reversed. The ordered strings of data are written into random access storage in sets of ordered data blocks with each block having at least one link each to the data blocks logically positioned in front and in back thereof so that the logical position of each block in the string is well defined. After the initial string generation phase has been completed, the sequence in which the now existing strings will be selected for merging into fewer strings is computed so that the data in the strings will be manipulated a minimum number of times. The data blocks of the merged strings are then read into the working storage area of the digital computer where the data blocks are merged into still fewer strings. The working storage area is divided into a record hold area and generally, at least three buffers, an input buffer, an output buffer, and an active buffer from which the data is transferred into and from the record hold area. The order in which the data blocks already in the working storage area will become exhausted is maintained and hence the order in which new data blocks will be required is anticipated before they are needed. Thus, the read-write heads (hereinafter referred to as readheads) in the random access storage devices can be positioned at the proper track location for reading a required block before the block is needed. This is one result of the synchronization of read and write operations which thereby reduces the input/output time. This merge process is repeated until the number of strings has been reduced to a number equal to the order of the final merge. That number of strings is then read from the random access storage area into the working storage area of the computer and again merged until only one string or set of strings remains which string or set of strings contains the input data ordered in the desired sequence.
-
公开(公告)号:US4037205A
公开(公告)日:1977-07-19
申请号:US578685
申请日:1975-05-19
Applicant: Murray Edelberg , Lloyd R. Schissler
Inventor: Murray Edelberg , Lloyd R. Schissler
CPC classification number: G06F17/30982 , G06F7/22 , G11C19/00 , Y10S707/99937
Abstract: A digital memory comprising a plurality of contiguous circulating serial data storage loops and a plurality of processing elements. Each pair of adjacent loops are coupled to a processing element such that each loop is coupled through two processing elements displaced by one-half loop length. Each processing element includes a serial comparator and a loop control circuit for either isolating or exchanging the data on the two incident loops. Additional circuits are included in each processing element which in combination with the comparator and loop control provide the memory, as the loop data circulates through the processing elements, with the capability of performing an ascending or descending sort of its contents, an associative search of its contents with retrieval of located data, an updating of stored data records, a loading or unloading of a record file or data base and a dynamic reconfiguration of the memory structure.
Abstract translation: 一种数字存储器,包括多个连续的循环串行数据存储环路和多个处理单元。 每对相邻的环路耦合到处理元件,使得每个环路通过两个被半个环路长度移位的处理元件耦合。 每个处理元件包括串行比较器和用于隔离或交换两个入射回路上的数据的环路控制电路。 附加电路包括在每个处理元件中,其结合比较器和循环控制提供存储器,因为循环数据通过处理元件循环,具有执行其内容的升序或降序排序的能力,其关联搜索 检索定位数据的内容,存储的数据记录的更新,记录文件或数据库的加载或卸载以及存储器结构的动态重新配置。
-
公开(公告)号:US3102997A
公开(公告)日:1963-09-03
申请号:US52560
申请日:1960-01-05
Applicant: GERHARD DIRKS
Inventor: GERHARD DIRKS
IPC: G06F7/22
CPC classification number: G06F7/22
-
公开(公告)号:US2974306A
公开(公告)日:1961-03-07
申请号:US47214454
申请日:1954-11-15
CPC classification number: G06F7/22 , G06F3/0601 , G06F7/026 , G06F15/08 , G06F2003/0698 , Y10S707/99937
-
公开(公告)号:US2911277A
公开(公告)日:1959-11-03
申请号:US36732653
申请日:1953-07-10
Applicant: COOK ELECTRIC CO
Inventor: BELLAMY JOHN C
IPC: G06F7/22
CPC classification number: G06F7/22
-
公开(公告)号:US2847657A
公开(公告)日:1958-08-12
申请号:US43094354
申请日:1954-05-19
Applicant: INT STANDARD ELECTRIC CORP
Inventor: CLIFFORD HARTLEY GEORGE , HARRY BRAY FREDERICK
-
公开(公告)号:US2539998A
公开(公告)日:1951-01-30
申请号:US9386449
申请日:1949-05-18
Applicant: IBM
Inventor: HOLLAND-MARTIN CYRIL G , BOWYER ALDRED W
Abstract: 655,674. Statistical apparatus. MARTIN, C. G. HOLLAND-, and BOWYER, A. W. Nov. 3, 1947, No. 29288. [Class 106 (i)] A record-card controlled machine for searching for data contained in unknown columns of the cards, comprises card feeding means, data sensing means from which data is presented to a search unit column by column, and card separating or data reproducing means. The search unit includes a chain of contacts manually set to represent desired data, interspersed with contacts automatically set by the sensing means ; other similar chains are brought into operation when the first chain registers an agreement. In the described embodiments the data searched for consists of patent subject-matter, particularly chemical data which can be expressed in alphabetical and numerical characters and represented by a card code set out in the Specification. The required data is set up manually on a switchboard, and appropriate plug connections are made (Fig. 1, not shown). Upon depression of a start key, the cards are fed endwise from a magazine 102, Fig. 2, by mechanism including a one-revolution clutch 133 controlled by a magnet CFCM and driving an eccentric 131 which reciprocates a sector driving the picker blade 121. The usual card lever control contacts CLC are closed as the card passes, and sensing brushes 126 read the cards column by column. After sensing, the cards are passed on by the feeding rollers to a pocket RP, but when a card is found to contain the required data, a magnet SM is energized and its armature lever 127 depresses a deflector 128 which guides the selected card into a pocket SP. The search unit according to a first embodiment is shown diagrammatically in Fig. 3. In the first chain of contacts, switches 8a ... 12a represent the manually settable means, while switches a in this chain are controlled by magnets R8 ... R12 each in the circuit of a card sensing brush 126. Where a card column contains the same data as that represented by the manually set switches, a circuit is completed through the first chain of contacts energizing the pick-up coil of a relav R 16 which closes contacts R16b in the second chain, thus preparing it for the next stage of comparison. Relay R16 also sets up a holding circuit, the duration of which depends on the plug connection from the socket 31. If taken to one of the sockets 26, the relay R16 is held until the following card column has been sensed when, if there is again agreement, relay R17 will be energized, after which contacts C2 open and relay R16 is deenergized. When the required data has been sensed and the appropriate relays R16 ... R20 have been successively energized, a plug connection to a socket 33 completes the circuit of a relay R14 which establishes a holding circuit through its contacts a and prepares a circuit for the magnet SM through its contacts b. Energization of magnet SM, upon closure of contacts C6 at an appropriate time, causes the required card to be guided into the pocket SP as described above. Various possible searches are described, and means are provided for skipping certain data (such as numerical data) in particular cases. Moreover, a combination of two groups of data may be searched for. In a second embodiment an electronic search unit is employed for greater speed. Each chain of the search unit is built up from valve units, Fig. 5a, which, in the unset state, have plug connections between 200 and 201, 202 and 203, and across sockets 210, whereby the valve is, normally, non-conducting. - In the set state, sockets 201 and 202, 203 and 200 are connected and the valves are, normally, conducting. When a card hole is sensed, a positive impulse arrives via a wire 207 and has the effect of switching a non-conducting valve (unset state), " ON or a conducting valve (set state) " OFF." Thus, agreement between the setting of the valve units and the sensed data is indicated when all the valves in the chain are in the " OFF " condition. The anodes of all the valves in a chain are connected to the H.T. supply via a plug point 273, Fig. 5b, and resistor 211 across which there is a voltage drop, as long as any valve in the chain is conducting, sufficient to prevent a valve V13 from conducting. When, however, all the valves become non-conducting the voltages on the grid and cathode of V13 become equal and the valve conducts causing a negative pulse to pass through a diode V14 to the grid of a triode V15 comprising the normally conducting valve of a flip-flop V15, V16. The impulse momentarily reverses the flip-flop, and as the valve V16 returns to its normal state its rising anode potential is applied to the grid of a cathode follower V17 causing a positive pulse to be developed across the load 223 which, being supplied to the grid of a thyratron V18 causes the latter to strike. The anode of V18 may be connected via plug point 229 to the common cathode line 208 of the search valves in the next chain, so that when V18 becomes conductive the necessary potential difference is supplied to render the next chain effective. V18 is extinguished when a negative impulse from a control unit, arriving at a plug point 233, causes a large amplitude positive pulse across the anode load 230 of a valve V19. This pulse, through a condenser 228, drives the cathode of V18 sufficiently positive to extinguish the thyratron. The control unit comprises a unit supplying a short-time sensing impulse employing a flip-flop circuit triggered by the closure of cam contacts, a similar unit supplying the resetting impulse referred to above, and a further unit for cancelling the resetting impulse when column skipping is permissible. The general operation of the electronic search unit (Fig. 6, not shown) is similar to that of the first embodiment. Instead of separating the cards bearing the required data means may be provided for reproducing the required data from the cards which may then remain in their original order. The data which could be represented in the card by punching, marking, embossing or magnetizing may be sensed while the card is at rest and then read out column by column by an 80-position commutator in the manner described in Specification 623,283.
-
88.
公开(公告)号:US20240111996A1
公开(公告)日:2024-04-04
申请号:US18449532
申请日:2023-08-14
Applicant: Incucomm, Inc.
Inventor: Randal Allen , Steven D. Roemerman , John P. Volpi
IPC: G06N3/045 , G06F7/22 , G06F11/34 , G06F17/11 , G06F17/16 , G06F17/18 , G06F30/27 , G06N3/08 , G06N7/01 , G06N20/00
CPC classification number: G06N3/045 , G06F7/22 , G06F11/3452 , G06F17/11 , G06F17/16 , G06F17/18 , G06F30/27 , G06N3/08 , G06N7/01 , G06N20/00 , G06F17/142
Abstract: Disclosed are systems and methods for operations and maintenance (O&M) systems for controlling the operation of a second system, wherein the second system includes a plurality of objects (i.e., components and/or subsystems). The O&M system comprises means for evaluating at least one of the plurality of objects using a first model to produce a first prediction of a characteristic of the object; evaluating at least one of the plurality of objects using a second model to produce a second prediction of the characteristic of the object, the second model being dissimilar to the first model; generating a final prediction of the characteristic of the object as a function of dynamic weightings of the first prediction and the second prediction; and, executing an operational or maintenance decision with respect to the second system as a function of the final prediction of the characteristic of the at least one of the plurality of objects.
-
公开(公告)号:US20240061647A1
公开(公告)日:2024-02-22
申请号:US18260134
申请日:2021-11-22
Applicant: SONY GROUP CORPORATION
Inventor: RYOTARO SHIRAI
Abstract: [Object]
A calculation in which a natural number is expressed by four square numbers is performed at higher speed.
[Solving Means]
Provided is an information processing apparatus including a calculation unit that performs a calculation to express a certain natural number by the sum of a first square number, a second square number, a third square number, and a fourth square number. The calculation unit calculates, as the square root of the first square number, the largest value or a value smaller by two than the largest value among even numbers equal to or less than the square root of the natural number and then searches for the respective square roots of the second square number, the third square number, and the fourth square number.-
公开(公告)号:US09928063B2
公开(公告)日:2018-03-27
申请号:US15267668
申请日:2016-09-16
Applicant: Intel Corporation
Inventor: Elmoustapha Ould-Ahmed-Vall , Kshitij A. Doshi , Suleyman Sair , Charles R. Yount
CPC classification number: G06F9/30036 , G06F7/22 , G06F7/544 , G06F9/30018 , G06F9/30021 , G06F9/30101 , G06F9/30145 , G06F9/3016 , G06F11/1048 , G06F11/1479
Abstract: Instructions and logic provide vector horizontal majority voting functionality. Some embodiments, responsive to an instruction specifying: a destination operand, a size of the vector elements, a source operand, and a mask corresponding to a portion of the vector element data fields in the source operand; read a number of values from data fields of the specified size in the source operand, corresponding to the mask specified by the instruction and store a result value to that number of corresponding data fields in the destination operand, the result value computed from the majority of values read from the number of data fields of the source operand.
-
-
-
-
-
-
-
-
-