Configurable scan chain architecture for multi-port memory
By designing an embedded full-scan chain architecture for efficient multi-port memory applications, the problem of large-scale scanning chain occupancy and difficult timing matching in semiconductor manufacturing is solved, area efficiency and timing matching is achieved, multiple port configurations are supported, and design flexibility is improved Sex and efficiency.
Patent Information
- Application Number
- CN202310694613.9
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Priority Date
- 2022-09-26
- Filing Date
- 2023-06-13
- Publication Date
- 2024-11-12
AI Technical Summary
In conventional semiconductor manufacturing designs, the embedded scanning chain occupies a large area, resulting in complex integrated circuit design and asynchronous clock design increases the complexity of the circuit architecture, making it difficult to achieve area efficiency and timing matching.
An embedded full scan chain architecture for efficient and scalable multi-port memory applications is designed. By reducing the number of scan shift chains and unified scan chains, the observable latch at the control input is reduced, and area effectiveness and timing are achieved. Match, supports different number of read/write port configurations.
It realizes the reuse of the functional input/output path of the DFT scan chain without adding non-functional latch, which reduces area occupation and matches the timing of different operating modes, supports multiple port configurations, and improves the Design flexibility and efficiency.
Smart Images

Figure CN118942523A_ABST
Abstract
Description
Background Art
[0001] This section is intended to provide information relevant to understanding the various technologies described herein. As the title of this section implies, this is a discussion of related art and should in no way be considered prior art. Generally speaking, related art may or may not be considered prior art. Therefore, it should be understood that any statement in this section should be understood in this sense and is not intended to be any admission of prior art.
[0002] In conventional semiconductor manufacturing designs, embedded scan chains may include scan chains in which write data inputs, write enable, and read data outputs include full scan flip-flops. Additionally, these embedded scan chains may be used in asynchronous clock designs, in which data (D) and write enable (WEN) capture uses a write clock (WCLK) and the Q output uses a read clock (RCLK). These many signals and associated circuits may result in a large area footprint on the circuit architecture. Therefore, in some modern circuit designs, it may be necessary to redesign the scan chain to reduce the area of the integrated circuit. BRIEF DESCRIPTION OF THE DRAWINGS
[0003] The specific implementations of various technologies are described herein with reference to the accompanying drawings. However, it should be understood that the accompanying drawings only illustrate various specific implementations described herein and are not intended to limit the implementation of the various technologies described herein.
[0004] Figure 1A to Figure 1B A diagram illustrating a scan chain architecture for a multi-port memory application according to various implementations described herein.
[0005] FIG. 2A to FIG. 2C A diagram illustrating a scan capture output table for a multi-port memory application according to various implementations described herein.
[0006] Figure 3 A diagram showing a scan capture input-output (IO) signal table for a multi-port memory application according to an implementation described herein.
[0007] Figure 4 A diagram illustrating a scan chain architecture for a multi-port memory application according to various implementations described herein.
[0008] Figure 5 A diagram showing a scan capture input-output (IO) signal table for a multi-port memory application according to an implementation described herein.
[0009] Figure 6 A diagram illustrating a scan chain architecture for a multi-port memory application according to various implementations described herein. DETAILED DESCRIPTION
[0010] Various embodiments described herein relate to efficient and scalable DFT (design for testability) scan chain architecture designs for multi-port memory applications in various scan chain applications. Various embodiments described herein provide embedded full scan chain designs, schemes, and techniques for various integrated circuit (IC) memory applications. Various schemes and techniques described herein enable area-efficient memory designs with embedded full scan chains by reducing the number of scan shift chains in, for example, input / output (I / O) column multiplexers (colmux). In addition, various schemes and techniques described herein can further address flip-flop count reduction by unifying scan chains in the presence of asynchronous read and write clock domains and by reducing observability latches at control inputs.
[0011] Various implementations described herein implement area-efficient and / or scalable solutions for embedded full scan chains in multi-port memory applications. Thus, in various applications, the scan chain architecture design described herein enables the reuse of functional input / output (IO) data path latches of the DFT scan chain without adding any non-functional latches in order to improve area and match timing between operating modes. Thus, the various solutions and implementations described herein are configurable and / or scalable to any number of read / write ports, such as 12R6W, 9R9W, 6R12W, etc.
[0012] Various embodiments described herein implement area-efficient and / or scalable solutions for multi-port memory designs having many read and write ports and / or having unequal numbers of read and write ports. In addition, various embodiments described herein reduce the number of scan chains for multi-port memories from many to one, wherein the scan chain schemes and techniques described herein reduce the number of scan flip-flops by unifying the scan chains, even in the case of many and unequal numbers of read and write ports in the multi-port memory. Some novel features and characteristics described herein can reuse the functional data path input-output (IO) latches of the DFT scan chain, so as to thereby reduce the area impact and match the timing between different operating modes without additional latches. In addition, even with unequal numbers of read ports (RP) and write ports (WP), these read-write ports are controllable, scalable, and observable by scan capture / shift, and these read-write ports can be easily extended to different number of port configurations, so that different RP-WP configurations can be implemented and used in various multi-port memory architecture designs. In some cases, the number of read ports (RP) may be equal to the number of write ports (WP), such as 3R3W, 7R7W, 9R9W, etc. In addition, in other cases, the number of read ports (RP) may be greater than the number of write ports (WP), such as 12R6W, 9R8W, 4R3W, etc. In addition, in various other cases, the number of read ports (RP) may be less than the number of write ports (WP), such as 6R12W, 5R7W, 4R12W, etc.
[0013] This article will now refer to Figures 1 to Figure 6 Various implementations of scan chain architectures for memory applications are described in detail.
[0014] Figure 1A Diagram 100A shows a scan chain architecture 104A for a multi-port memory application according to various implementations described herein.
[0015] In various specific implementations, Figure 1AThe scan chain architecture 104A in the present invention may be implemented as a system or device having integrated circuits (ICs) and components that are arranged and coupled together as an assembly or some combination of parts that can provide a physical circuit layout design and related structures. In various cases, methods of designing, manufacturing, constructing, and / or providing the scan chain architecture 104A as an integrated system or device may involve using the IC circuit components described herein to thereby implement various configurable scan chain architecture schemes and techniques associated therewith. In addition, the scan chain architecture 104A may be integrated with various computing circuits and related components on a single chip, and the scan chain architecture 104A may be implemented in embedded systems for automotive, electronic, mobile, server, and Internet of Things (IoT) applications.
[0016] like Figure 1A As shown, the scan chain architecture 104A may include a first data path circuit having an input latch (p2 latch) that receives data (D) from a plurality of write ports (WP) and provides a first latched data (Q) signal based on a clock signal (CLKB). In various applications, the write port (WP) may be associated with a plurality of D input ports (DB0, DB1, ..., DB8), and other inputs may include a scan input (SI) and a design for testability (DFT) random access memory (RAM) bypass (BYP) signal, which may be referred to as a DFTRAMBYP signal, and (&) a sense amplifier enable signal (SE). In addition, the inverted DFTRAMBYP signal may be referred to as a !DFTRAMBYP signal.
[0017] In some implementations, the scan chain architecture 104A may have a plurality of input logic gates (L1) and a plurality of multiplexers (M1) that receive D input signals (DB0, DB1, ..., DB8) and provide data (D) input signals to input latches (p2 latches). Figure 1A As shown, the scan chain architecture 104A may include multiple input latches (eg, two p2 latches) per data port that latch multiple input data signals (DB0, DB1, ..., DB8) provided by an input logic gate (L1) and a multiplexer (M1).
[0018] Additionally, in some applications, multiple D input signals (DB0, DB1, ..., DB11) may be provided to the L1 logic gate and / or the M1 multiplexer, and the output from the L2 logic gate and / or the M1 multiplexer may be provided to the D input of the input latch (p2 latch). Additionally, the M1 multiplexer may provide an output signal to the D input of the input latch (p2 latch) based on the !DFTRAMBYP signal. Additionally, in various applications, the L1 logic gate may be referred to as an AND gate, and the M1 multiplexer may be referred to as a 2 to 1 multiplexer; however, various other configurations may be possible.
[0019] In some implementations, the scan chain architecture 104A may include a second data path circuit having a logic gate (L2, M2) that receives the first latch data (Q) from the input latch (p2 latch) and provides the first latch data (Q) based on the read bit line signal (RPBL). In addition, in some applications, multiple latched Q output signals from multiple input latches (p2 latches) of each port may be provided to the L2 logic gate, and the output from the L2 logic gate may be provided to the M2 multiplexer together with the RPBL signal. In addition, the M2 multiplexer may provide an output signal to the D input of the output latch (p1 latch) based on the DFTRAMBYP signal and the RPBL signal.
[0020] In some implementations, the scan chain architecture 104A may also include a third data path circuit having an output latch (p1 latch) that receives a first latched data (e.g., Q output) signal from a logic gate (L2, M2) and then provides a second latched data (QA0, QA1, QA2) to a plurality of read ports (RP) based on a clock signal (CLKA). Additionally, in some applications, one or more of the second latched data (QA0, QA1, QA2) may be provided as a feedback signal to the M1 multiplexer, and additionally, at least one of the second latched data signals (e.g., QA2) may be used as a scan output (SO) signal. Additionally, in some applications, the CLKA signal is equal to the CLKB signal. In various applications, the L2 logic gate may be referred to as an OR gate, and the M2 multiplexer may be referred to as a 2 to 1 multiplexer; however, various other configurations may be possible.
[0021] In various applications, as described in more detail herein, the number of read ports (RP) may be less than (<) the number of write ports (WP), the number of read ports (RP) may be greater than (>) the number of write ports (WP), or the number of read ports (RP) may be equal to (=) the number of write ports (WP). In various applications, the scan chain architecture 104A may refer to a configurable scan chain architecture, and thus, in various applications, the scan chain architecture 104A may be implemented in a device having a memory design for testability (DFT) design for multi-port memory applications.
[0022] For example, in some applications, such as Figure 1AAs shown, the number of D input ports (e.g., DB0, DB1, ..., DB6) of the input latch (p2 latch) may refer to 6 write ports (e.g., 6WP or 6W), and the number of read ports (e.g., QA0, QA1, QA2) of the output latch (p1 latch) may refer to 3 read ports (e.g., 3RP or 3R). In this scenario, the scan chain architecture 104A may be configured as a 3 read and 6 write (3R6W) scan chain architecture that may be used for a multi-port memory architecture. In various specific implementations, such as, Figure 1A The configuration shown may be selectively configurable (or extended) as further described herein for any other number (X) of read ports (RP) and any other number (Y) of write ports (WP) (e.g., a 6 read and 12 write (6R12W) scan chain architecture).
[0023] In various specific implementations, the clock circuit can be used to provide a clock signal (CLKB) to the input latch (p2 latch) and also provide a clock signal (CLKA) to the output latch (p1 latch). Therefore, the plurality of clock signals can include a first clock signal (CLKA) that can be used as a scan shift signal during the first operating mode, and the plurality of clock signals can also include a second clock signal (CLKB) that can be used as a scan capture signal during the second operating mode.
[0024] In various specific implementations, such as reference Figure 1A As shown, the input latch (p2 latch) may include multiple latches (e.g., two latches) for each data bit, the output latch (p1 latch) may include a single latch (e.g., one latch) for each data bit, and the logic gate (L2, M2) may be coupled between the input latch (p2 latch) and the output latch (p1 latch), wherein the logic gate (L2, M2) may combine the first latch data from each data port of the multiple latches (e.g., two latches for each data bit) into a single latch data (e.g., one latch for each data bit) provided to the output latch (p1 latch). In various applications, the number of input latches (e.g., two p2 latches for each data bit) is greater than the number of output latches (e.g., one p1 latch for each data bit). Thus, the input latch (p2 latch) includes two (2) latches for each data bit, and the output latch (p1 latch) includes one (1) latch for each data bit.
[0025] In various specific implementations, such as reference Figure 1AAs shown, the logic gates (L2, M2) include multiple logic gates for each data port, including, for example, a first logic gate (L2) and a second logic gate (M2). In addition, the first logic gate (L2) receives multiple first latched data for each data port from at least two (2) input latches of the input latches (p2 latches), and provides a single first latched data to the second logic gate (M2). In addition, the second logic gate (M2) receives the single first latched data from the first logic gate (L2), receives a read bit line signal (RPBL), and provides a single first latched data to the output latch (p1 latch) of each data port based on a design for testability (DFT) signal (DFTRAMBYP).
[0026] Figure 1B FIG. 100B shows a scan chain architecture 104B for a multi-port memory application according to a specific implementation described herein. In some applications, two back-to-back latches may be used to implement a flip-flop. Figure 1B As shown, the scan chain architecture 104B may be configured as a repeating circuit configured as a repeating (or daisy-chained) scan chain architecture (SCA) 104A, wherein at least one output read port (e.g., RP: QA2) is coupled to the scan input (S1) of the next SCA 104A in the daisy-chain configuration. In this case, one or more of the input write ports (e.g., DB1, DB2, DB3, DB4) may be provided to a plurality of individual latches.
[0027] In addition, in some applications, such as Figure 1B As shown, the number of D input ports (e.g., DB0, DB1, ..., DB11) may refer to 12 write ports (e.g., 12WP or 12W), and the number of read ports (e.g., QA0, QA1, ..., QA5) of the output latch (p1 latch) may refer to 6 read ports (e.g., 6RP or 6R). In this scenario, the scan chain architecture 104B may be configured as a 6 read and 12 write (6R12W) scan chain architecture that can be used for a multi-port memory architecture. In addition, Figure 1B The configuration shown may be selectively configurable for any other number (X) of read ports (RP) and any other number (Y) of write ports (WP), for example, as further described herein.
[0028] For example, the input write ports (DB8[0], DB2[0], DB7[0], DB1[0], DB6[0], DB0[0], SI[0]) may be provided to SCA 104A1, which then provides the output read ports (QA2[0], QA1[0], QA0[0]). For another example, the input write ports (DB8[1], DB2[1], DB7[1], DB1[1], DB6[1], DB0[1], QA2[0]) may be provided to SCA 104A2, which then provides the output read ports (QA2[1], QA1[1], QA0[1]). As another example, input write ports (DB8
[33] , DB2
[33] , DB7
[33] , DB1
[33] , DB6
[33] , DB0
[33] , QA2[0]) may be provided to SCA 104A3, which then provides output read ports (SO[0], QA2
[33] , QA1
[33] , QA0
[33] ).
[0029] For another example, the input write ports (DB11[0], DB5[0], DB10[0], DB4[0], DB9[0], DB3[0], SI[1]) may be provided to SCA 104A4, which then provides the output read ports (QA5[0], QA4[0], QA3[0]). For another example, the input write ports (DB11[1], DB5[1], DB10[1], DB4[1], DB9[1], DB3[1], QA5[0]) may be provided to SCA 104A5, which then provides the output read ports (QA5[1], QA4[1], QA3[1]). As another example, input write ports (DB11
[33] , DB5
[33] , DB10
[33] , DB4
[33] , DB9
[33] , DB3
[33] , QA5[1]) may be provided to SCA 104A6, which then provides output read ports (SO[1], QA5
[33] , QA4
[33] , QA3
[33] ).
[0030] FIG. 2A to FIG. 2C Various diagrams of a scan capture output table 204 for a multi-port memory application according to a specific implementation described herein are shown. Specifically, Figure 2A A diagram 200A shows a scan capture output table 204A for a multi-port memory application, where the number of read ports (RP: X) is equal to (=) the number of write ports (WP: Y), Figure 2B FIG. 200B shows a scan capture output table 204B for a multi-port memory application, where the number of read ports (RP: X) is greater than (>) the number of write ports (WP: Y), and Figure 2CDiagram 200C shows a scan capture output table 204C for a multi-port memory application where the number of read ports (RP: X) is less than (<) the number of write ports (WP: Y).
[0031] Figure 2A A diagram 200A shows a scan capture output table 204A for a multi-port memory application, where the number of read ports (RP: X) is equal to (=) the number of write ports (WP: Y). In some implementations, when the scan capture output table 204A is configured for a multi-port memory having an equal number of read ports (RP) and write ports (WP), it may refer to a table describing Figure 1A 1. Data input-output (IO) table of IO characteristics of the scan chain architecture 104A in FIG. 1, where RP=WP.
[0032] like Figure 2A As shown, the scan capture output table 204A provides a 6R6W multi-port memory with six read ports (6RP) and six write ports (6WP). The read ports (RP:R:X) may refer to the Figure 1A The six (6) Q outputs (6R: QA0, QA1, ..., QA5) of the output latch (ph1 latch) of the scan chain architecture 104A in FIG. 1 and the write port (WP: W: Y) can be pointed to Figure 1A The six (6) D inputs (6W: DB0, DB1, ..., DB5) of the input latch (ph2 latch) of the scan chain architecture 104A in FIG.
[0033] In various specific implementations, Figure 1A The scan chain architecture 104A shown is scalable and can therefore be modified and / or configured to have any number of DB inputs and any number of QA outputs to provide a 6R6W multi-port memory with six read ports (6RP) and six write ports (6WP). However, many other RP=WP configurations are possible, such as 3R3W, 4R4W, 5R5W, etc. For example, in some applications, such as Figure 6 As shown, Figure 1A The illustrated scan chain architecture 104A may be modified and / or configured to provide a 3R3W scan chain architecture 604 .
[0034] Figure 2B FIG. 200B shows a scan capture output table 204B for a multi-port memory application, where the number of read ports (RP: X) is at least greater than (>) the number of write ports (WP: Y). In various implementations, when the scan capture output table 204B is configured for a multi-port memory having a greater number of read ports (RP) than write ports (WP), it may refer to a table describing Figure 1A1. Data input-output (IO) table of IO characteristics of the scan chain architecture 104A in FIG. 1, where RP>WP.
[0035] In various applications, Figure 1A The scan chain architecture 104A shown may refer to a configurable scan chain architecture, and therefore, in some implementations, Figure 1A The scan chain architecture 104A shown may be implemented in a device having 6 read ports (RP) and 4 write ports (WP), wherein the number of read ports (e.g., 6RP:6R:6X) may be greater than the number of write ports (e.g., 4WP:4W:4Y). Thus, in various applications, the number of write ports (WP:W:Y) may be equal to or greater than two (2) write ports (2WP:2W:2Y), and the number of read ports (RP:R:X) may be equal to or greater than three (3) write ports (3RP:3R:3X), such that the number of read ports (RP) is at least greater than the number of write ports (WP), wherein RP>WP.
[0036] like Figure 2B As shown, the scan capture output table 204B provides a 6R4W multi-port memory with six read ports (6RP) and six write ports (4WP). The read ports (RP:R:X) may refer to the Figure 1A The six (6) Q outputs (6R: QA0, QA1, ..., QA5) of the output latch (ph1 latch) of the scan chain architecture 104A in FIG. 1 and the write port (WP: W: Y) can be pointed to Figure 1A The four (4) D inputs (4W: DB0, DB1, ..., DB3) of the input latch (ph2 latch) of the scan chain architecture 104A in FIG. In some applications, additional QA outputs may be mirrored during scan capture / shift, for example, referring to the 6R4W configuration, the DB3 output may be captured to QA3, QA4, and QA5.
[0037] In various specific implementations, Figure 1A The scan chain architecture 104A shown is scalable and can therefore be modified and / or configured to have any number of DB inputs and any number of QA outputs to provide a 6R4W multi-port memory with six read ports (6RP) and four write ports (4WP). However, other RP=WP configurations may be possible, such as 4R3W, 5R4W, 5R3R, etc. For example, in some applications, such as Figure 4 As shown, Figure 1A The illustrated scan chain architecture 104A may be modified and / or configured to provide a 4R3W scan chain architecture 404 .
[0038] Figure 2CFIG200C shows a scan capture output table 204C for a multi-port memory application, where the number of read ports (RP: X) is less than (<) the number of write ports (WP: Y). In some implementations, when the scan capture output table 204C is configured for a multi-port memory having a smaller number of read ports (RP) than write ports (WP), it may refer to a table describing Figure 1A The data input-output (IO) table of the IO characteristics of the scan chain architecture 104A in FIG. 1 is shown in FIG. <WP。
[0039] In various applications, Figure 1A The scan chain architecture 104A shown may refer to a configurable scan chain architecture, and therefore, in some implementations, Figure 1A The scan chain architecture 104A shown may be implemented in a device having 6 read ports (RP) and 12 write ports (WP), wherein the number of read ports (e.g., 6RP:6R:6X) may be less than the number of write ports (e.g., 12WP:12W:12Y). Thus, in various applications, the number of write ports (WP:W:Y) may be equal to or greater than three (3) write ports (3WP:3W:3Y), and the number of read ports (RP:R:X) may be equal to or greater than two (2) read ports (2RP:2R:2X), such that the number of read ports (RP) is at least less than the number of write ports (WP), wherein the number of RP is less than the number of write ports (WP). <WP。
[0040] In some applications, the number of write ports (WP:W:Y) may be greater than (>) the number of read ports (RP:R:X), and therefore, in some scenarios, the number of read ports (RP:R:X) may be equal to or greater than (>) two (2) read ports (2RP:2R:2X), and the number of write ports (WP:W:Y) may be equal to or greater than (>) three (3) write ports (3WP:3W:3Y), so that the number of write ports (WP) is at least greater than (>) the number of read ports (RP), where WP>RP.
[0041] like Figure 2C As shown, the scan capture output table 204C provides a 6R12W multi-port memory with six read ports (6RP) and twelve write ports (12WP). The read ports (RP:R:X) may refer to the Figure 1A The six (6) Q outputs (6R: QA0, QA1, ..., QA5) of the output latch (ph1 latch) of the scan chain architecture 104A in FIG. 1 and the write port (WP: W: Y) can be pointed to Figure 1AThe twelve (12) D inputs (12W: DB0, DB1, ..., DB11) of the input latches (ph2 latches) of the scan chain architecture 104A in FIG. In various applications, the two DB inputs may be OR-gated into the QA latches, for example, referring to the 6R12W configuration, DB0-DB5 are captured into QA0-QA5, and QA6-QA11 are captured into QA0-QA5.
[0042] In various specific implementations, Figure 1A The scan chain architecture 104A shown is scalable and can therefore be modified and / or configured to have any number of DB inputs and any number of QA outputs to provide a 6R12W multi-port memory with six read ports (6RP) and twelve write ports (12WP). However, other RP=WP configurations may be feasible, such as 3R6W, 4R6W, 4R8W, 5R8W, etc. For example, in various applications, such as Figure 1A As shown, the scan chain architecture 104A may be modified and / or configured to provide a 6R12W scan chain architecture 104A.
[0043] Figure 3 A diagram 300 is shown of a scan capture input-output (IO) signal table 304 for a multi-port memory according to implementations described herein.
[0044] In various applications, Figure 1A The scan chain architecture 104A shown is configured to provide a multi-port memory Figure 3 The configurable scan chain architecture of the scan capture IO table 304 is shown. Therefore, in some specific implementations, such as Figure 1A The scan chain architecture 104A in may be implemented as a 6R12W device having 6 read ports (RP) and 12 write ports (WP), where the number of read ports (e.g., 6RP:6R:6X) may be less than the number of write ports (e.g., 12WP:12W:12Y).
[0045] In various specific implementations, Figure 1A The scan chain architecture 104A shown can be configured to support unequal numbers of DB input ports and QA output ports, where functional latches can be reused to reduce area impact and to match timing between functional memory mode and DFT mode (i.e., design for testability mode). In addition, in some applications, multiple synchronous clocks (e.g., CLKA, CLKB) can be used, such that CLKA=CLKB. In addition, in some applications, the scan chain shift order on QB / QA can refer to SI->QA0->QA1->…->QA4->QA5 / SO. In other applications, the present invention can be modified, configured and / or extended. Figure 1AThe scan chain architecture 104A shown to support an additional write port using an OR3 (i.e., 3-input "OR" gate), where for example with reference to a 4R12W configuration, the 3-DB inputs can be "OR" gated into a single (1) QA.
[0046] As Figure 3 shown, the scan capture IO table 304 refers to a 6R12W scan capture IO table, where the number of read ports is less than the number of write ports, such that RP(X) < WP(Y). In the functional read mode from read port 0 (QA0), as shown in the first row of Table 304, for a read (R) operation, CLKA = CLKB and DFTRAMBYP = 0, and thus, QA0 is stored / shifted into the memory address location AA0, i.e., MEM(@AA0), and QA5 is stored / shifted into the memory address location AA5, i.e., MEM(@AA0).
[0047] Additionally, in some applications, in the capture write mode of the write port DB-input (DB*), as shown in the second row of Table 304, for a write (W) operation, CLKA = CLKB and DFTRAMBYP = 1, and scan enable = 0, and thus, QA0 is used to capture DB0 - DB6, and QA5 is used to capture DB5 - DB11.
[0048] Additionally, in some applications, in the scan-shift mode of the scan input (SI), as shown in the third row of Table 304, for a scan-input operation, CLKA = CLKB and DFTRAMBYP = 1, and scan enable = 1, and thus QA0 is scan-shifted, and QA5 is scan-shifted.
[0049] Figure 4 FIG. 400 shows a scan chain architecture 404 for multi-port memory applications according to a specific implementation described herein. Referring Figure 4 , various circuit components as described herein are similar in scope, features, and functionality to various related components shown and described in the scan chain architecture 104A as shown in reference Figure 1A . In various applications, Figure 4 the scan chain architecture 404 shown can be implemented in a 4R3W device having 4 read ports (RP) and 3 write ports (WP), where the number of read ports (e.g., 4RP: 4R: 4X) can be greater than the number of write ports (e.g., 3WP: 3W: 3Y).
[0050] In various specific implementations, Figure 4The scan chain architecture 404 in the present invention may be implemented as a system or device having integrated circuits (ICs) and components that are arranged and coupled together as an assembly or some combination of parts that can provide a physical circuit layout design and related structures. In various cases, methods of designing, manufacturing, constructing, and / or providing the scan chain architecture 404 as an integrated system or device may involve using the IC circuit components described herein to thereby implement various configurable scan chain architecture schemes and techniques associated therewith. In addition, the scan chain architecture 404 may be integrated with various computing circuits and related components on a single chip, and the scan chain architecture 404 may be implemented in embedded systems for automotive, electronic, mobile, server, and Internet of Things (IoT) applications.
[0051] like Figure 4 As shown, the scan chain architecture 404 may include a first data path circuit having an input latch (p2 latch) that receives data (D) from a plurality of write ports (WP) and provides a first latched data (Q) signal based on a clock signal (CLKB). In various applications, the write port (WP) may be associated with a plurality of D input ports (DB0, DB1, DB2), and other inputs may include a scan input (SI) and a DFTRAMBYP signal, and (&) a sense amplifier enable signal (SE). In addition, the inverted DFTRAMBYP signal may be referred to as a !DFTRAMBYP signal.
[0052] In some implementations, the scan chain architecture 404 may have multiple input multiplexers (M1) that receive D input signals (DB0, DB1, DB2) and provide data (D) input signals to input latches (p2 latches). Figure 4 As shown, the scan chain architecture 404 may have a single input latch (eg, one p2 latch) per data port that latches the input data signals ( DB0 , DB1 , DB2 ) provided by the input multiplexer ( M1 ).
[0053] Additionally, in some applications, multiple D input signals (DB0, DB1, DB2) may be provided to the M1 multiplexer, and the output from the M1 multiplexer may be provided to the D input of the input latch (p2 latch). Additionally, the M1 multiplexer may provide an output signal to the D input of the input latch (p2 latch) based on the !DFTRAMBYP signal. In various applications, the M1 multiplexer may be referred to as a 2 to 1 multiplexer; however, various other configurations may be possible.
[0054] In some implementations, the scan chain architecture 404 may include a second data path circuit having a multiplexer (M2) that receives the first latched data (Q) from the input latch (p2 latch) and provides the first latched data (Q) based on the read bit line signal (RPBL). Additionally, in some applications, a single latched Q output signal from multiple input latches (p2 latches) of each port may be provided to the M2 multiplexer, and the M2 multiplexer may provide an output signal to the D input of the output latch (p1 latch) based on the DFTRAMBYP signal and the RPBL signal.
[0055] In some implementations, the scan chain architecture 404 may also include a third data path circuit having an output latch (p1 latch) that receives a first latched data (e.g., Q output) signal from the M2 multiplexer and then provides second latched data (QA0, QA1, QA2, QA3) to a plurality of read ports (RP) based on a clock signal (CLKA). Additionally, in some applications, one or more of the second latched data (QA0, QA1) may be provided as a feedback signal to the M1 multiplexer, and further, at least one of the second latched data signals (e.g., QA2) may be used as a scan output (SO) signal. In some applications, the CLKA signal is equal to the CLKB signal. In some applications, the M2 multiplexer may be referred to as a 2 to 1 multiplexer; however, various other configurations may be possible.
[0056] In various applications, as described in more detail herein, the number of read ports (RP) may be greater than (>) the number of write ports (WP), wherein the number of output latches (p1 latches) is greater than the number of input latches (p2 latches). Additionally, in some applications, the number of read ports (RP:R:X) is greater than the number of write ports (WP:W:Y), such that RP>WP. Thus, in various applications, the scan chain architecture 404 may refer to a configurable scan chain architecture, wherein the scan chain architecture 404 may be implemented in a 4R3W device having a memory design for testability (DFT) design for multi-port memory applications.
[0057] For example, in some applications, such as Figure 4 As shown, the number of D input ports of the input latch (p2 latch) (e.g., DB0, DB1, DB2) can refer to 3 write ports (e.g., 3WP or 3W), and the number of read ports of the output latch (p1 latch) (e.g., QA0, QA1, QA2, QA3) can refer to 4 read ports (e.g., 4RP or 4R). In this scenario, the scan chain architecture 404 can be configured as a 4-read and 3-write (4R3W) scan chain architecture that can be used for a multi-port memory architecture. In addition, Figure 4The configuration shown may be selectively configurable, for example, as described herein, for any other number (X) of read ports (RP) and any other number (Y) of write ports (WP).
[0058] In various specific implementations, the clock circuit can be used to provide a clock signal (CLKB) to the input latch (p2 latch) and also provide a clock signal (CLKA) to the output latch (p1 latch). Therefore, the plurality of clock signals can include a first clock signal (CLKA) that can be used as a scan shift signal during the first operating mode, and the plurality of clock signals can also include a second clock signal (CLKB) that can be used as a scan capture signal during the second operating mode.
[0059] In various specific implementations, such as reference Figure 4 As shown, the input latch (p2 latch) may include a single latch (one p2 latch) for each data bit, the output latch (p1 latch) may include a single latch (one latch) for each data bit, and the M2 multiplexer may be coupled between the input latch (p2 latch) and the output latch (p1 latch). In addition, in various applications, the number of input latches (e.g., one p1 latch for each data bit) is less than the number of output latches (e.g., one or two p1 latches for each data bit). Therefore, the input latch (p2 latch) provides 3 write ports (3WP:3W:3Y), and the output latch (p1 latch) provides 4 read ports (4RP:4R:4X), where 4RP>3WP. Therefore, the number of output latches (4 p1 latches) is greater than the number of input latches (3 p2 latches), and therefore, the number of read ports (4RP) is greater than the number of write ports (3WP).
[0060] Figure 5 A diagram 500 is shown of a scan capture input-output (IO) signal table 504 for a multi-port memory according to implementations described herein.
[0061] In various applications, Figure 1A The scan chain architecture 104A shown is configured to provide a multi-port memory Figure 5 The configurable scan chain architecture of the scan capture IO table 504 is shown. Therefore, in some specific implementations, Figure 1A The illustrated scan chain architecture 104A may be implemented in a 4R3W device having 4 read ports (RP) and 3 write ports (WP), such that the number of read ports (eg, 4RP:4R:4X) may be greater than the number of write ports (eg, 3WP:3W:3Y).
[0062] In various specific implementations, Figure 1AThe scan chain architecture 104A shown can be configured to support unequal numbers of DB input ports and QA output ports, where functional latches can be reused to reduce area impact and to match timing between functional memory mode and DFT mode (i.e., design for testability mode). In addition, in some applications, multiple synchronous clocks (e.g., CLKA, CLKB) can be used, such that CLKA=CLKB. In addition, in some applications, the scan chain shift order on QB / QA can refer to SI[0]->QA0[0]->QA1[0]->QA2[0]->QA0[1]->..., and further, during the scan capture / shift operation mode, QA3[*]=QA2[*].
[0063] like Figure 5 As shown, the scan capture IO table 504 may refer to a 4R3W scan capture IO table, wherein the number of read ports (RP) may be greater than the number of write ports (WP), such that RP(X)>WP(Y). In a functional read mode from read port 0 (QA0), as shown in the first row of table 504, for a read (R) operation, CLKA=CLKB and DFTRAMBYP=0, and in some specific implementations, QA0 is stored / shifted to memory address location AA0, i.e., MEM (@AA0), and, further, QA1 is stored / shifted to memory address location AA1, i.e., MEM (@AA1), and further, QA2 is stored / shifted to memory address location AA2, i.e., MEM (@AA2), and further, QA3 is stored / shifted to memory address location AA3, i.e., MEM (@AA3).
[0064] In addition, in some applications, in the capture write mode of the write port DB-input (DB*), as shown in the second row of Table 504, for the write (W) operation, CLKA=CLKB and DFTRAMBYP=1, and scan enable=0, and in some specific implementations, QA0 is used to capture DB0, and QA1 is used to capture DB1, and QA2 is used to capture DB2, and QA3 is used to capture DB3.
[0065] In addition, in some applications, in the scan-shift mode of scan input (SI), as shown in the third row of Table 504, for the scan-in operation, CLKA=CLKB and DFTRAMBYP=1, and scan enable=1, and therefore QA0 is scan-shifted, QA1 is scan-shifted, QA2 is scan-shifted, and, QA3 is scan-shifted.
[0066] Figure 6 A diagram 600 is shown of a scan chain architecture 604 for a multi-port memory application according to an implementation described herein. Figure 6, the various circuit components described herein are similar in scope, features, and functionality to those described in reference Figure 1A The various related components shown and described in the scan chain architecture 104A shown. In various applications, Figure 6 The scan chain architecture 604 shown may be implemented in a 3R3W device having 3 read ports (RP) and 3 write ports (WP), where the number of read ports (eg, 3RP:3R:3X) may be less than the number of write ports (eg, 3WP:3W:3Y).
[0067] In various specific implementations, Figure 6 The scan chain architecture 604 in the embodiment of the present invention may be implemented as a system or device having integrated circuits (ICs) and components arranged and coupled together as an assembly or some combination of parts that may provide a physical circuit layout design and related structures. In various cases, methods of designing, manufacturing, constructing, and / or providing the scan chain architecture 604 as an integrated system or device may involve using the IC circuit components described herein to thereby implement various configurable scan chain architecture schemes and techniques associated therewith. Furthermore, the scan chain architecture 604 may be integrated with various computing circuits and related components on a single chip, and the scan chain architecture 604 may be implemented in embedded systems for automotive, electronic, mobile, server, and Internet of Things (IoT) applications.
[0068] like Figure 6 As shown, the scan chain architecture 604 may include a first data path circuit having an input latch (p2 latch) that receives data (D) from a plurality of write ports (WP) and provides a first latched data (Q) signal based on a clock signal (CLKB). In various applications, the write port (WP) may be associated with a plurality of D input ports (DB0, DB1, DB2), and other inputs may include a scan input (SI) and a DFTRAMBYP signal, and (&) a sense amplifier enable signal (SE). In addition, the inverted DFTRAMBYP signal may be referred to as a !DFTRAMBYP signal.
[0069] In some implementations, the scan chain architecture 604 may have multiple input multiplexers (M1) that receive D input signals (DB0, DB1, DB2) and provide data (D) input signals to input latches (p2 latches). Figure 6 As shown, the scan chain architecture 604 may have a single input latch (eg, one p2 latch) per data port that latches the input data signals ( DB0 , DB1 , DB2 ) provided by the input multiplexer ( M1 ).
[0070] Additionally, in some applications, multiple D input signals (DB0, DB1, DB2) may be provided to the M1 multiplexer, and the output from the M1 multiplexer may be provided to the D input of the input latch (p2 latch). Additionally, the M1 multiplexer may provide an output signal to the D input of the input latch (p2 latch) based on the !DFTRAMBYP signal. In various applications, the M1 multiplexer may be referred to as a 2 to 1 multiplexer; however, various other configurations may be possible.
[0071] In some implementations, the scan chain architecture 604 may include a second data path circuit having a multiplexer (M2) that receives the first latched data (Q) from the input latch (p2 latch) and provides the first latched data (Q) based on the read bit line signal (RPBL). Additionally, in some applications, a single latched Q output signal from multiple input latches (p2 latches) of each port may be provided to the M2 multiplexer, and the M2 multiplexer may provide an output signal to the D input of the output latch (p1 latch) based on the DFTRAMBYP signal and the RPBL signal.
[0072] In some implementations, the scan chain architecture 604 may also include a third data path circuit having an output latch (p1 latch) that receives a first latched data (e.g., Q output) signal from the M2 multiplexer and then provides second latched data (QA0, QA1, QA2) to a plurality of read ports (RP) based on a clock signal (CLKA). Additionally, in some applications, one or more of the second latched data (QA0, QA1) may be provided as a feedback signal to the M1 multiplexer, and further, at least one of the second latched data signals (e.g., QA2) may be used as a scan output (SO) signal. In some applications, the CLKA signal is equal to the CLKB signal. In some applications, the M2 multiplexer may be referred to as a 2 to 1 multiplexer; however, various other configurations may be possible.
[0073] In various applications, as described in more detail herein, the number of read ports (RP) may be equal to (=) the number of write ports (WP), and thus, in various applications, the number of read ports (RP:R:X) is therefore equal to the number of write ports (WP:W:Y), such that RP=WP. Thus, in some applications, the scan chain architecture 604 may refer to a configurable scan chain architecture, wherein the scan chain architecture 604 may be implemented in a 3R3W device having a memory design for testability (DFT) design for multi-port memory applications.
[0074] For example, in some applications, such as Figure 6As shown, the number of D input ports of the input latch (p2 latch) (e.g., DB0, DB1, DB2) can refer to 3 write ports (e.g., 3WP or 3W), and the number of read ports of the output latch (p1 latch) (e.g., QA0, QA1, QA2) can refer to 3 read ports (e.g., 3RP or 3R). Therefore, in some scenarios, the scan chain architecture 604 can be configured as a 3 read and 3 write (3R3W) scan chain architecture that can be used for a multi-port memory architecture. In addition, Figure 6 The configuration shown may be selectively configurable, for example, as described herein, for any other number (X) of read ports (RP) and any other number (Y) of write ports (WP).
[0075] In various specific implementations, the clock circuit can be used to provide a clock signal (CLKB) to the input latch (p2 latch) and also provide a clock signal (CLKA) to the output latch (p1 latch). Therefore, the plurality of clock signals can include a first clock signal (CLKA) that can be used as a scan shift signal during the first operating mode, and the plurality of clock signals can also include a second clock signal (CLKB) that can be used as a scan capture signal during the second operating mode.
[0076] In various specific implementations, such as reference Figure 6 As shown, the input latch (p2 latch) may include a single latch (one p2 latch) for each data bit, the output latch (p1 latch) may include a single latch (one latch) for each data bit, and the M2 multiplexer may be coupled between the input latch (p2 latch) and the output latch (p1 latch). In addition, in various applications, the number of input latches (e.g., one p1 latch for each data bit) may be equal to the number of output latches (e.g., one p1 latch for each data bit). Therefore, the input latch (p2 latch) provides 3 write ports (3WP:3W:3Y), and the output latch (p1 latch) provides 3 read ports (3RP:3R:3X), where 3RP=3WP. Therefore, the number of output latches (3 p1 latches) is equal to the number of input latches (3 p2 latches), and therefore, the number of read ports (3RP) is equal to the number of write ports (3WP).
[0077] It is contemplated that the subject matter of the claims is not limited to the specific implementations and illustrations provided herein, but rather includes modifications of those specific implementations according to the claims, including portions of specific implementations and combinations of elements of different specific implementations. It should be understood that in the development of any such specific implementation, as in any engineering or design project, many specific implementation-specific decisions should be made to achieve the developer's specific goals, such as compliance with system-related and business-related constraints, which may vary between different specific implementations. In addition, it should be understood that such development work may be complex and time consuming, but is nevertheless a routine task of design, manufacture, and production for a person of ordinary skill having the benefit of this disclosure.
[0078] Various implementations of a device having a first data path circuit having an input device that receives data from a plurality of write ports and provides first data are described herein. The device may have a second data path circuit having a logic gate that receives the first data from the input device and provides the first data based on a read bit line signal. The device may have a third data path circuit having an output device that receives the first data from the logic gate and provides second data to a plurality of read ports. Additionally, the number of read ports may be greater than the number of write ports.
[0079] Various implementations of a device having a first data path circuit are described herein, the first data path circuit having an input latch that receives data from a plurality of write ports and provides a first latched data. The device may have a second data path circuit having a logic gate that receives the first latched data from the input latch and provides the first latched data based on a read bit line signal. The device may have a third data path circuit having an output latch that receives the first latched data from the logic gate and provides the second latched data to a plurality of read ports. In addition, the number of write ports may be greater than the number of read ports.
[0080] Various implementations of a device having a first data path circuit are described herein, the first data path circuit having an input latch that receives data from a plurality of write ports and provides a first latched data. The device may have a second data path circuit having a logic gate that receives the first latched data from the input latch and provides the first latched data based on a read bit line signal. The device may have a third data path circuit having an output latch that receives the first latched data from the logic gate and provides the second latched data to a plurality of read ports. In addition, the number of output latches is greater than the number of input latches, and the number of read ports is greater than the number of write ports.
[0081] Reference has been made in detail to various specific implementations, examples of which are shown in the accompanying drawings and diagrams. In the following specific description, many specific details are set forth to provide a thorough understanding of the disclosure provided herein. However, the disclosure provided herein may be practiced without these specific details. In some other cases, well-known methods, procedures, components, circuits, and networks are not described in detail so as not to unnecessarily obscure the details of the embodiments.
[0082] It should also be understood that although the terms "first", "second", etc. may be used to describe various elements in this article, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, a first element may be referred to as a second element, and similarly, a second element may be referred to as a first element. The first element and the second element are both elements, but they are not considered to be the same element.
[0083] The terms used in the description of the disclosure provided herein are for the purpose of describing a specific implementation, and are not intended to limit the disclosure provided herein. As used in the disclosure provided herein and the description of the appended claims, the singular forms "one", "a kind of" and "the" are intended to also include plural forms, unless the context clearly indicates otherwise. As used herein, the term "and / or" refers to and encompasses any and all possible combinations of one or more of the associated listed items. When used in this specification, the terms "include", "comprise" and / or "contain" specify the presence of the features, integers, steps, operations, elements and / or parts, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, parts and / or their groups.
[0084] As used herein, the term "if" may be interpreted to mean "when" or "at" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrase "if it is determined that" or "if [the condition or event] is detected" may be interpreted to mean "upon determining that" or "in response to determining that" or "upon detecting [the condition or event]" or "in response to detecting [the condition or event]," depending on the context. The terms "up" and "down"; "upper" and "lower"; "upwardly" and "downwardly"; "below" and "above"; and other similar terms indicating relative positions above or below a given point or element may be used in conjunction with some specific implementations of the various techniques described herein.
[0085] While the foregoing relates to various implementations of the various schemes and techniques described herein with reference to a configurable scan chain architecture for a multi-port memory, various other implementations are contemplated based on the disclosure herein, which may be determined by the claims appended hereto below.
[0086] Although the subject matter described herein is described in language specific to structural features and / or methodological acts, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described herein are disclosed as exemplary forms of implementing the claims.
Claims
1. A device comprising: a first data path circuit having an input device that receives data from a plurality of write ports and provides first data; a second data path circuit having a logic gate that receives the first data from the input device and provides the first data based on a read bit line signal; and a third data path circuit having an output device that receives the first data from the logic gate and provides second data to a plurality of read ports, The number of read ports is greater than the number of write ports.
2. The apparatus of claim 1, wherein the apparatus comprises a scan chain architecture having a memory design for testability (DFT) design for multi-port memory applications.
3. The apparatus according to claim 1, wherein: The input device has a plurality of latches for each data bit; The output device has a single latch for each data bit; and The logic gate is coupled between the input device and the output latch, wherein the logic gate combines the first data from the plurality of latches into a single latched data that is provided to the output latch. The device of claim 1 , wherein the number of input devices is greater than the number of output devices.
5. The device of claim 1, wherein the input device comprises two (2) latches per data bit, and wherein the output device comprises one (1) latch per data bit.
6. The apparatus of claim 1, wherein the number of write ports is equal to or greater than two (2) write ports, and wherein the number of read ports is equal to or greater than three (3) write ports, and wherein the number of read ports is at least greater than the number of write ports.
7. The apparatus of claim 1, wherein: The logic gate includes a plurality of logic gates, wherein the plurality of logic gates includes a first logic gate and a second logic gate, The first logic gate receives a plurality of first data from at least two (2) input latches of the input device and then provides a single first latched data to the second logic gate, and The second logic gate receives the single first data from the first logic gate, receives the read bit line signal, and provides the single first latch data to the output latch based on a design for testability (DFT) signal.
8. A device comprising: a first data path circuit having an input latch that receives data from a plurality of write ports and provides first latched data; a second data path circuit having a logic gate that receives the first latched data from the input latch and provides the first latched data based on a read bit line signal; and a third data path circuit having an output latch that receives the first latched data from the logic gate and provides second latched data to a plurality of read ports, The number of write ports is greater than the number of read ports.
9. The apparatus of claim 8, wherein the apparatus comprises a scan chain architecture having a memory design for testability (DFT) design for multi-port memory applications.
10. The apparatus of claim 8, wherein: The input latch has a plurality of latches for each data bit; The output latch has a single latch for each data bit; and The logic gate is coupled between the input latch and the output latch, wherein the logic gate combines the first latch data from the plurality of latches into a single latch data provided to the output latch.
11. The apparatus of claim 8, wherein the number of input latches is greater than the number of output latches.
12. The apparatus of claim 8, wherein the input latch comprises two (2) latches per data bit, and wherein the output latch comprises one (1) latch per data bit.
13. The apparatus of claim 8, wherein the number of read ports is equal to or greater than two (2) read ports, and wherein the number of write ports is equal to or greater than three (3) write ports, and wherein the number of write ports is at least greater than the number of read ports.
14. The apparatus of claim 8, wherein: The logic gate includes a plurality of logic gates, wherein the plurality of logic gates includes a first logic gate and a second logic gate, The first logic gate receives a plurality of first latch data from at least two (2) of the input latches and then provides a single first latch data to the second logic gate, and The second logic gate receives the single first latch data from the first logic gate, receives the read bit line signal, and provides the single first latch data to the output latch based on a design for testability (DFT) signal.
15. A device comprising: a first data path circuit having an input latch that receives data from a plurality of write ports and provides first latched data; a second data path circuit having a logic gate that receives the first latched data from the input latch and provides the first latched data based on a read bit line signal; and a third data path circuit having an output latch that receives the first latched data from the logic gate and provides second latched data to a plurality of read ports, Wherein the number of output latches is greater than the number of input latches, and wherein the number of read ports is greater than the number of write ports.
16. The apparatus of claim 15, wherein the apparatus comprises a scan chain architecture having a memory design for testability (DFT) design for multi-port memory applications.
17. The apparatus of claim 15, wherein: The input latches have a single input latch for each data bit; The output latches have a single output latch for each data bit; and The logic gate is coupled between the input latch and the output latch, The logic gate receives the first latch data from the single input latch and provides the first latch data to the single output latch based on the read bit line signal.
18. The apparatus of claim 15, wherein the input latch comprises two (2) latches per data bit, and wherein the output latch comprises one (1) latch per data bit.
19. The apparatus of claim 15, wherein the number of write ports is equal to or greater than two (2) write ports, and wherein the number of read ports is equal to or greater than three (3) write ports, and wherein the number of read ports is at least greater than the number of write ports.
20. The apparatus of claim 15, wherein the number of write ports is equal to or greater than three (3) write ports, and wherein the number of read ports is equal to or greater than three (4) write ports, and wherein the number of read ports is at least greater than the number of write ports.
