Unit architecture with rear power rail
By introducing a rear-side power distribution network architecture, heterogeneous integration of power rails and power supply across multiple voltage domains are achieved in the standard unit, solving the problem of power rail continuity requirements in existing technologies and improving design efficiency and area utilization.
Patent Information
- Application Number
- CN202111302958.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Priority Date
- 2020-11-06
- Filing Date
- 2021-11-04
- Publication Date
- 2026-11-10
- Estimated Expiration
- 2041-11-04
AI Technical Summary
Existing EDA tools lack effective support for connecting different power domains and grounding grids, resulting in the need for continuity of power rails and grounding rails in standard unit layouts, which affects design efficiency and area utilization.
It adopts a rear-side power distribution network (BPDN) architecture, which integrates the rear power rail with the heterogeneous rail of the front power network in the standard unit. It utilizes embedded through-hole coupling and special filling units to realize the interruption and isolation of the power rail, and supports power supply in multiple voltage domains.
It improves the area utilization efficiency of standard unit design, supports power supply in multiple voltage domains, enhances the flexibility of power rails and design, and reduces area overhead.
Smart Images

Figure CN114444428B_ABST
Abstract
Description
Background Technology
[0001] This section aims to provide information relevant to understanding the various techniques described herein. As the title of this section suggests, this is a discussion of related techniques and should in no way imply that they are prior art. Generally speaking, related techniques may or may not be considered prior art. Therefore, it should be understood that any statement in this section should be interpreted in this sense and does not constitute any endorsement of prior art.
[0002] In semiconductor design, standard cells typically include transistors and interconnect structures to implement various digital circuits for the logic functions used in the design. Standard cell layout refers to the physical representation of digital circuits used in different forms in circuit design using electronic design automation (EDA) tools. Standard cells with different circuits and functions are available in a standard cell library, and standard cells may have characteristics for associating themselves with the standard cell library. These characteristics, along with design rules, facilitate placement in a design block, and each standard cell has power supply rails and ground supply rails at fixed locations relative to the row boundary of each row of standard cells. These power rails and ground rails provide primary power, and when these cells are placed in a block design, the power rails and ground rails of the standard cells in the block are connected to a single power grid and ground grid. Conventional EDA tools often lack support for power rails and / or ground rails connected to different power grids or ground grids in the same power / voltage domain. Additionally, some power supplies connected to standard cells pass through metal pins without passing through power rails / ground rails, and design blocks with standard cells require continuous power rails / ground rails in designated areas of the standard cells. Therefore, there is a need to improve the physical design layout of standard cells. Attached Figure Description
[0003] This document describes specific implementations of various techniques with reference to the accompanying drawings. However, it should be understood that the drawings merely illustrate the various specific implementations described herein and are not intended to limit the implementation methods of the various techniques described herein.
[0004] Figure 1 The diagram illustrates the embedded power rail (BPR) process technology used in physical design according to various specific implementations described herein.
[0005] Figure 2 The diagram illustrates various specific implementations of Back-of-House Power Distribution Network (BPDN) architectures as described herein.
[0006] Figures 3A-3B Additional diagrams illustrating the BPDN architecture according to various specific implementations described herein are shown.
[0007] Figures 4A-4BAdditional diagrams illustrating the BPDN architecture according to various specific implementations described herein are shown.
[0008] Figures 5A-5D Additional diagrams illustrating the BPDN architecture according to various specific implementations described herein are shown.
[0009] Figure 6 Another diagram shows the BPDN architecture according to the various specific implementations described in this article.
[0010] Figures 7A-7E Additional diagrams illustrating the BPDN architecture according to various specific implementations described herein are shown.
[0011] Figure 8 A diagram is shown of a system for providing back-side power distribution technology by physical design, according to a specific implementation described herein.
[0012] Figure 9 A diagram illustrating the composition and placement and routing (PNR) process according to various specific implementations described herein.
[0013] Figure 10 A diagram showing the placement of standard unit designs according to various specific implementations described herein. Detailed Implementation
[0014] The various embodiments described herein relate to chip-scale integrated circuit (IC) design schemes and techniques for implementing various efficient backside power supply and distribution methods in physical structures. These schemes and techniques provide heterogeneous rail integration with a backside power distribution network (BPDN) in a standard cell architecture. In various embodiments, backside power / ground rail disconnection can be efficiently implemented to connect the frontside rail of a standard cell to a secondary power / ground supply using various cell edge types and / or power gap cells with special fill cells. When standard cells are constructed using various process technologies that allow power / ground rails on the backside of the wafer, the backside rails may require continuous disconnection to provide secondary power / ground to the standard cell or at least some portions thereof. Continuous disconnection of the backside rails can reduce the area overhead for disconnecting the backside rails by using automated placement rules and special cells, which may involve the use of gap cells and special fill cells based on various defined edge types.
[0015] The various schemes and techniques described herein can be applied to various cell design processes that force power rail interruptions in primary rail structures by implementing heterogeneous rail integration in standard cell placement and cabling (PNR) through some cell design processes that utilize interruptions in the primary power rail. Some cell design methods and techniques can also be combined with single-rail or multi-rail configurations, and some variations of multi-row cell designs can have additional pathways to the primary supply rail and / or secondary supply rail. Some novel cell types can complement the cell design techniques provided herein, and some power rail segments of the same network can be compatible with each other and can be placed together in the PNR to obtain the benefits of the various ideas presented herein.
[0016] The various solutions and techniques described herein can achieve area-efficient composition and placement and routing (PNR) processes by utilizing standard cell layouts, structures, and methods specifically designed for cell design processes when standard cells require electrically isolated structures (such as NWELLs with different potentials, power supply rails coupled to secondary power supplies that differ from the primary power supply in the cell design). The various solutions and techniques described herein are primarily applicable to semiconductor processes supporting rear-side power distribution networks (BPDNs) and having rear-side power rails (or embedded power rails) coupled to transistors on the front side via embedded vias to interrupt the primary power rail and / or ground rail and to insert secondary power segments and / or ground rail segments. Additionally, the various solutions and techniques described herein can be applied to semiconductor processes with front-side power rails to insert secondary rail segments by interrupting the primary power rail and to isolate NWELLs with different potentials in an area-efficient manner. Different potentials can refer to differences in voltage values and / or different power supplies.
[0017] This article will refer to Figures 1-8 This paper describes in detail the various specific implementations of physical layout design techniques for back-side power distribution.
[0018] Figure 1 Figure 100 illustrates a specific implementation of the embedded power rail (BPR) process technology 104 for physical design, as described herein.
[0019] In various specific implementations, BPR process technology 104 can be implemented with systems or devices having various integrated circuit (IC) components arranged and coupled together as an assembly or combination of parts providing physical circuit design and / or related structures. In some cases, methods of designing, providing, fabricating, and / or manufacturing systems and / or devices using BPR process technology 104 may involve using the various IC circuit components described herein to implement various associated back-side power distribution schemes and techniques. BPR process technology 104 can be integrated with various computing circuits and / or related fabrication components on a single chip structure, and BPR process technology 104 can be implemented in various embedded systems for automotive, electronics, mobile, server, and Internet of Things (IoT) applications (including remote sensor nodes).
[0020] like Figure 1 As shown, the BPR process technology 104 can utilize a Front Power Network (FSPN) architecture 108 configured for memory applications, such as, for example, power distribution for a core memory array. In some cases, the Front Power Network (FSPN) 108 may include a front metal layer (FML) supplying the core voltage in different power domains, wherein the power domains of the core voltage may refer to the primary power domain (PPWR) and / or the secondary power domain (SPWR).
[0021] In some implementations, the front power network (FSPN) 108 may include various types of components, circuits, and metal layers. For example, the FSPN 108 may include front power rails coupled to memory circuitry having, for example, an array of bit cells arranged in columns and rows, as well as header logic and power gating control logic. Additionally, in some cases, the FSPN 108 may include any number (N) of front metal layers (FMLs) and front interlayer vias (FSVs) coupling different FMLs together.
[0022] BPR process technology 104 can also utilize a rear-side power network (BSPN) architecture 118, which provides rear-side power distribution for memory applications, such as for use with various memory control circuitry, including column multiplexer circuitry, sense amplifier circuitry, power gating input / output circuitry, and / or power gating control circuitry. Additionally, the rear-side power network (BSPN) 118 can be configured to supply core voltage in different power domains, including primary power domains (PPWR) and / or secondary power domains (SPWR) for front-side memory control logic. Furthermore, the rear-side power distribution network (BSPN) 118 can be configured to supply different types of source voltages, including, for example, core voltage, peripheral voltage, and / or ground.
[0023] In some implementations, the Backside Power Network (BSPN) 118 may include one or more backside metal layers (BMLs) supplying power in multiple domains (PPWR, SPWR), including, for example, embedded metal layers. Additionally, the BSPN 118 may have backside power rails and backside interlayer vias (BSVs) formed together with the backside metal layers (BMLs). The BSPN 118 may also include backside power connection bumps (BSBs).
[0024] BPR process technology 104 may also utilize a front-to-back transition architecture 114 with an embedded transition via (BTV) for coupling an embedded backside power network (BSPN) to a frontside power network (FSPN). In some cases, the embedded transition via (BTV) 124 may be configured to provide a power transition between the backside power network (BSPN) 118 and the frontside power network (FSPN) 108. Therefore, BPR process technology 104 may be configured to transition the backside power rails of the backside power network to the frontside power rails of the frontside power network to provide a power tap from the backside power network (BSPN) to the memory circuitry. Additionally, the front-to-back transition 114 may be referred to as an embedded transition architecture, which may have a backside-to-frontside transition unit and an embedded transition via 124 providing a coupling transition between the backside power network (BSPN) 118 and the frontside power network (FSPN) 108.
[0025] In some specific implementations, the rear power rail (BPR) technology 104 can be configured to operate as a power distribution network architecture, utilizing the rear embedded metal in the rear power network (BSPN) of the rear power rail, and also utilizing the front metal of the front power network (FSPN) of the front power rail. Alternatively, the front power network (FSPN) can use the front metal, and the rear power network (BSPN) can use the rear embedded metal disposed below the front metal of the front power network (FSPN). Furthermore, the transition architecture can refer to an embedded transition architecture disposed between the rear embedded metal of the rear power network (BSPN) and the front metal of the front power network (FSPN).
[0026] In various specific implementations, the various schemes and techniques described herein can provide a power distribution network architecture for transitioning power from the rear to the front using embedded power rails. Additionally, the power distribution network architecture can be configured to use embedded metal on the rear for the rear power rails and metal on the front for the front power rails. Furthermore, in some cases, the power distribution network architecture can utilize header-based power tapping circuits in multiple different voltage domains (such as, for example, PPWR and SPWR) and grounding (GND). The following references... Figures 2-7ETo describe these aspects and various other features in more detail.
[0027] References for various physical designs Figure 1 Embedded Power Rail (BPR) Technology 104 Figures 2-7E Various physical layout patterns and diagrams associated with the Back-of-House Power Distribution Network (BPDN) architecture according to various specific implementations described herein are illustrated. In some specific implementations, Figures 2-7E The diagram illustrates various BPDN cell architectures with power rail structures providing primary power (PPWR), secondary power (SPWR), and ground (GND), as well as cell boundaries (CB) and available equipment spans (USOD). Additionally, Figures 2-7E The various diagrams illustrate power rails with a power rail interruption (or rail break or rail gap) formed by a reference primary power (PPWR), which allows for the insertion of other power rails for secondary power (SPWR). Additionally, Figures 2-7E The various diagrams in the document illustrate power rails with modifiable features, including, for example, various edge types, various location orientations, and / or various adjacency configurations.
[0028] Figure 2 Figure 200 illustrates a Back-of-House Power Distribution Network (BPDN) architecture 204 according to a specific implementation described herein.
[0029] like Figure 2 As shown, the BPDN architecture 204 may include a primary power rail (PPWR) coupled to the primary power rail (PPWR) and a ground power rail (GND) coupled to ground (GND). The BPDN architecture 204 may refer to a cell architecture having a cell boundary (CB) and a device available span (USOD). In various cases, the cell architecture may refer to a standard cell architecture, and the cell boundary (CB) may refer to one or more areas defining the device available span (USOD), which may be enclosed within the cell boundary (CB). The primary power rail (PPWR) and / or the ground power rail (GND) may be formed from the same or different back metal layers (BML).
[0030] Figures 3A-3B Additional diagrams illustrating the BPDN architecture according to various specific implementations described herein are shown. In particular, Figure 3A Figure 300A shows a BPDN architecture 304A with primary power (PPWR) and secondary power (SPWR) in a single row, and Figure 3B Figure 300B shows a BPDN architecture 304B with primary power (PPWR) and secondary power (SPWR) in multiple rows.
[0031] like Figure 3AAs shown, the BPDN architecture 304A may have a primary power rail (PPWR) coupled to the primary power (PPWR) and a ground power rail (GND) coupled to ground (GND). The BPDN architecture 304A may refer to a cell architecture having a cell boundary (CB) and a device available span (USOD). In some cases, the primary power rail (PPWR) may be interrupted, and the secondary power rail (SPWR) may be inserted into the gap space defined by the primary power interruption 308. Additionally, the primary power rail (PPWR) provides the primary power supply (PPWR), and the secondary power rail (SPWR) provides the secondary power supply (SPWR). Furthermore, in various cases, the primary power rail (PPWR), secondary power rail (SPWR), and / or ground power rail (GND) may be formed from the same or different back metal layers (BML).
[0032] In some implementations, the primary power interruption 308 provides a gap space interval for the insertion of the secondary power supply (SPWR). Additionally, in some cases, the primary power supply (PPWR) may be provided in a first power domain (such as, for example, the core power domain (PPWR)) and the secondary power supply (SPWR) may be provided in a second power domain different from the first power domain (such as, for example, the global core power domain (SPWR)).
[0033] In some specific implementations, Figure 3A This indicates a single-row unit with a secondary power supply (SPWR) having placement restrictions on the upper or north edge. Additionally, in other embodiments, the grounding power rail (GND) may be interrupted (with a break or gap), and the secondary power rail (SPWR) may be inserted into the gap space defined by the interrupted grounding rail (GND).
[0034] like Figure 3BAs shown, BPDN architecture 304B can refer to a multi-row cell having a first row cell boundary (CB1) and a second row cell boundary (CB2), wherein the secondary power supply (SPWR) is not restricted in placement at any edge. BPDN architecture 304B can have a primary power rail (PPWR) coupled to the primary power (PPWR) and multiple ground power rails (GND) coupled to ground (GND). Additionally, BPDN architecture 304B can have multiple areas for the device available span (USOD). In various cases, the primary power rail (PPWR) can be interrupted, and the secondary power rail (SPWR) can be inserted into the gap space interval defined by the primary power interruption 308. Furthermore, the primary power rail (PPWR) provides the primary power supply (PPWR), and the secondary power rail (SPWR) provides the secondary power supply (SPWR). Additionally, the primary power rail (PPWR), secondary power rail (SPWR), and / or ground power rail (GND) can be formed from the same or different back metal layers (BML). In other cases, the grounding power rail (GND) may be interrupted (with a break or gap), and the secondary power rail (SPWR) may be inserted into the gap space defined by the interrupted grounding rail.
[0035] Figures 4A-4B Additional diagrams illustrating the BPDN architecture according to various specific implementations described herein are shown. In particular, Figure 4A Figure 400A shows a BPDN architecture 404A with primary power (PPWR) and secondary power (SPWR) in a single row, and Figure 4B Figure 400B shows a BPDN architecture 404B with primary power (PPWR) and secondary power (SPWR) in multiple rows.
[0036] like Figure 4A As shown, the BPDN architecture 404A may have a primary power rail (PPWR) coupled to the primary power (PPWR) and a ground power rail (GND) coupled to ground (GND). The BPDN architecture 404A may refer to a cell architecture having a cell boundary (CB) and multiple regions (including an extension region 414 for the USOD) defined for the device available span (USOD). In some cases, the primary power rail (PPWR) may be interrupted, and the secondary power rail (SPWR) may be inserted into a gap space interval defined by the primary power interruption 408. Additionally, the primary power rail (PPWR) provides the primary power supply (PPWR), and the secondary power rail (SPWR) provides the secondary power supply (SPWR). In various cases, the primary power rail (PPWR), secondary power rail (SPWR), and / or ground power rail (GND) may be formed from the same or different back metallization layers (BML).
[0037] In some specific implementations, Figure 4A This can represent a single-row cell with primary power (PPWR) and secondary power (SPWR) but placement restrictions on the upper or north edge. Additionally, Figure 4A The diagram shows that the USOD may include multiple extended use areas 414 (i.e., extended USODs) that can be enclosed within a cell boundary (CB). Additionally, in other embodiments, the grounding power rail (GND) may be interrupted (having one or more breaks or gaps), and the secondary power rail (SPWR) may be inserted into the gap space defined by the interrupted grounding rail.
[0038] like Figure 4B As shown, BPDN architecture 404B can refer to a multi-row cell having a first row cell boundary (CB1) and a second row cell boundary (CB2), wherein the secondary power supply (SPWR) is not restricted in placement at any edge. BPDN architecture 404B can have a primary power rail (PPWR) coupled to the primary power (PPWR) and multiple ground power rails (GND) coupled to ground (GND). Additionally, BPDN architecture 404B can have multiple areas for the device available span (USOD). In various cases, the primary power rail (PPWR) can be interrupted, and the secondary power rail (SPWR) can be inserted into the gap space interval defined by the primary power interruption 408. Furthermore, the primary power rail (PPWR) provides the primary power supply (PPWR), and the secondary power rail (SPWR) provides the secondary power supply (SPWR). Additionally, the primary power rail (PPWR), secondary power rail (SPWR), and / or ground power rail (GND) can be formed from the same or different back metal layers (BML). In other cases, the grounding power rail (GND) may be interrupted (with one or more breaks or gaps), and the secondary power rail (SPWR) may be inserted into the gap space defined by the interrupted grounding rail.
[0039] Figures 5A-5D Additional diagrams illustrating the BPDN architecture according to the specific implementation described herein are shown. In particular, Figure 5A Figure 500A shows a BPDN architecture 504A with secondary power (SPWR) in a single row. Figure 5B Figure 500B shows a BPDN architecture 504B with primary power (PPWR) and secondary power (SPWR) in a single row. Figure 5C Figure 500C shows a BPDN architecture 504C with primary power (PPWR) and secondary power (SPWR) in multiple rows, and Figure 5DAnother figure 500D shows a BPDN architecture 504D with primary power (PPWR) and secondary power (SPWR) in multiple rows.
[0040] like Figure 5A As shown, BPDN architecture 504A can refer to an area-efficient cell design of BPDN cell architecture. BPDN architecture 504A can include a rear power distribution network (BPDN) with rear power rails providing secondary power (SPWR) and ground (GND). BPDN architecture 504A can have a cell architecture with cell boundaries (CB) and areas defined for the available span of equipment (USOD). In some cases, the secondary power rail (SPWR) provides secondary power supply (SPWR), and the ground power rail (GND) can be coupled to ground (GND). The secondary power rail (SPWR) and / or the ground power rail (GND) can be formed from the same or different back metal layers (BML). In some cases, Figure 5A The BPDN architecture 504A in the text refers to an area-efficient cell design with only a secondary power supply (SPWR) and placement restrictions on the east, west, and north edges.
[0041] like Figure 5B As shown, BPDN architecture 504B can refer to an area-efficient cell design of BPDN cell architecture. BPDN architecture 504B can include a rear-side power distribution network (BPDN) having rear-side power rails providing primary power supply (PPWR) and secondary power supply (SPWR) as well as ground (GND). In some cases, the rear-side power rails can include a first power rail having a first segment (PPWR) providing primary power supply (PPWR), a second segment (SPWR) providing secondary power supply (SPWR), and a power rail break 508 separating the first segment (PPWR) from the second segment (SPWR). The rear-side power rail includes a second power rail (GND), which is continuous and coupled to ground (GND).
[0042] In some specific implementations, the cell architecture can be a standard cell architecture, and the cell architecture can have a cell boundary (CB) and multiple areas of the device available span (USOD) enclosed within the cell boundary (CB). Interruption 508 can refer to a gap space interval in the first power rail, located between a first segment (PPWR) and a second segment (SPWR). A primary power supply (PPWR) can be provided in the first power domain (PPWR), and a secondary power supply (SPWR) can be provided in a second power domain (SPWR) different from the first power domain (PPWR). Additionally, in some cases, the first power domain (PPWR) can refer to the core power domain (PPWR), and the second power domain (SPWR) can refer to the global core power domain (SPWR), and the second power rail (GND) can refer to a grounded power rail coupled to ground (GND). In some cases, Figure 5B The BPDN architecture 504B in the text refers to an area-efficient cell design with primary and secondary power supplies and placement constraints on the east, west, and north edges.
[0043] like Figure 5C As shown, BPDN architecture 504C can refer to an area-efficient cell design of BPDN cell architecture. In some specific implementations, the rear power rails can be arranged in multiple rows, including a first row (row 1) providing secondary power supply (SPWR) at the inner edge 518 of cell architecture 504C and a second row (row 2) also providing secondary power supply (SPWR) at the inner edge 518 of cell architecture 504C. Multiple rows can include a third row (row 3) providing primary power supply (PPWR) at the outer edge 524 of cell architecture 504C, with the second row positioned between the first row (row 1) and the third row (row 3).
[0044] In some specific implementations, Figure 5C The BPDN architecture 504C can represent an area-efficient cell design with primary power supply (PPWR) and secondary power supply (SPWR) and placement constraints on the east, west and north edges. In addition, the BPDN architecture 504C can be used when the cell architecture has more devices connected to the secondary power supply (SPWR) instead of the primary power supply (PPWR).
[0045] like Figure 5DAs shown, BPDN architecture 504D can refer to an area-efficient cell design of BPDN cell architecture. In some specific implementations, the rear power rails can be arranged in multiple rows, including a first row (row 1) providing primary power supply (PPWR) at the inner edge 518 of cell architecture 504D and a second row providing primary power supply (PPWR) at the inner edge 518 of cell architecture 504D. Multiple rows can include a third row (row 3) providing secondary power supply (SPWR) at the outer edge 524 of cell architecture 504D, with the second row (row 2) positioned between the first row (row 1) and the third row (row 3).
[0046] In some specific implementations, Figure 5D The BPDN architecture 504D in the text can represent an area-efficient cell design with primary power supply (PPWR) and secondary power supply (SPWR) and placement constraints on the east, west and north edges. In addition, the BPDN architecture 504D can be used when the cell architecture has more devices connected to the secondary power supply (SPWR) instead of the primary power supply (PPWR).
[0047] In some specific implementations, cell architectures 504C and 504D may refer to standard cell architectures. In addition, cell architectures 504C and 504D may also have a first cell boundary (CB1), a second cell boundary (CB2), and a third cell boundary (CB3), wherein each cell boundary (CB1, CB2, CB3) has a corresponding USOD area that can be enclosed within each cell boundary (CB1, CB2, CB3).
[0048] Figure 6 Another figure 600 shows the BPDN architecture 604 according to the various specific implementations described herein.
[0049] like Figure 6As shown, BPDN architecture 604 refers to a multi-row cell with a cell architecture having a first row cell boundary (CB1) and a second row cell boundary (CB2), wherein each cell boundary (CB1, CB2) has a corresponding USOD area enclosed within each cell boundary (CB1, CB2). BPDN architecture 604 may include a back-side power distribution network (BPDN) having back-side power rails that provide primary power supply (PPWR), secondary power supply (SPWR), and ground (GND). The back-side power rails may have one or more continuous power rails (PPWR) coupled to the primary power supply (PPWR), and additionally, the back-side power rails may have one or more interrupted power rails (GND), which include a ground segment (GND) coupled to ground (GND), a power segment (SPWR) providing secondary power supply (SPWR), and a power rail interruption 5608 separating the ground segment (GND) from the power segment (SPWR).
[0050] In various specific implementations, the BPDN cell architecture 604 may refer to a standard cell architecture used for power gating design, and the BPDN cell architecture 604 may include multiple USOD regions enclosed within a first cell boundary (CB1) and a second cell boundary (CB2). In some cases, a primary power supply (PPWR) may be provided in a first power domain (PPWR) associated with the core voltage (PPWR), and a secondary power supply (SPWR) may be provided in a second power domain (SPWR) associated with the global core voltage (SPWR), and a ground segment (GND) is coupled to ground (GND). Furthermore, in some cases, Figure 6 This indicates a power gating unit design with both a primary power supply (PPWR) and a secondary power supply (SPWR), where the secondary power supply (SPWR) refers to the always-on global power supply.
[0051] Figures 7A-7E Various diagrams illustrating BPDN architectures according to various specific implementations described herein are shown. In particular, Figure 7A Figure 700A shows a BPDN architecture 704A with a cell type having no power edge (NPE). Figure 7B Figure 700B shows a BPDN architecture 704B with a cell type having extended power edges (EXT), and Figure 7C A specific implementation diagram 700C of the BPDN architecture 704C with NPE and EXT unit types is shown. Furthermore, Figure 7D Figure 700D shows a BPDN architecture 704D with a unit type having power rail gaps 708, and additionally, Figure 7EA specific implementation diagram 700E is shown for the BPDN architecture 704E with NPE unit type and EXT unit type and power rail gap 708.
[0052] In some specific implementations, the NPE element type can refer to the element edge of a cell that does not have a power rail at the rail location associated with its element edge. Conversely, the EXT element type can refer to the element edge of a cell that has an unnamed power rail (UNPR) at the rail location associated with its element edge. (This document references...) Figures 7A-7E These aspects, and various other aspects, related to the cell edge type are described in more detail.
[0053] like Figure 7A As shown, BPDN architecture 704A can refer to an area-efficient cell design of BPDN cell architecture. BPDN architecture 704A can include NPE cell types that do not have power rails at rail locations associated with the No Power Edge (NPE) edge, as well as ground rails (GND). BPDN architecture 704A can have cell architectures with cell boundaries (CB) and areas defined by the Usable Equipment Span (USOD). In some cases, BPDN architecture 704A may not have power rails and may only have ground power rails (GND). Therefore, Figure 7A The BPDN architecture 704A in the text can represent an area-efficient cell design that does not have a power rail but only a ground rail (GND) set on the south edge (or lower edge) of the cell boundary (CB).
[0054] like Figure 7B As shown, BPDN architecture 704B can refer to an area-efficient cell design of BPDN cell architecture. BPDN architecture 704B may include an EXT cell type with an unnamed power rail (UPPR) at the rail location associated with the Extended Power Edge (EXT) and a ground rail (GND). BPDN architecture 704B may have a cell architecture with a cell boundary (CB) and an area defined by the Usable Equipment Span (USOD). Additionally, in some cases, BPDN architecture 704B may include an unnamed power rail (UPPR) and a ground power rail (GND). Therefore, Figure 7B The BPDN architecture 704B in the text can represent an area-efficient cell design with an unnamed power rail (UPPR) and a ground rail (GND) set on the south edge (or lower edge) of the cell boundary (CB).
[0055] In some specific implementations, Figure 7C This illustrates an exemplary use of Extended Power Units (EXT) and No Power Units (NPE), wherein Figure 7C Figure 700C shows the usage of the BPDN architecture 704C with NPE and EXT cell types.
[0056] like Figure 7C As shown, the BPDN architecture 704C may have a first row (row_1) with multiple cells, which have various cell types, such as, for example, EXT cell type, NPE cell type, and UNPR cell type. These cell types may also include various standard cell types, such as, for example, inverter cell type (INV) and buffer cell type (BUF). In some cases, the first row (row_1) and the second row (row_2) may each include a certain number (N) of cells, such as, for example, 8 cells. Additionally, the BPDN architecture 704C may include a rear power rail, such as, for example, a segmented power rail disposed between the lower ground power rail (GND) and the upper ground power rail (GND). In some cases, the segmented power rail may have UNPR segments, PPWR segments, no-power (NP) segments, and SPWR segments, which are disposed in corresponding cells associated with each row (row_1, row_2).
[0057] like Figure 7D As shown, BPDN architecture 704D can refer to an area-efficient cell design of BPDN cell architecture. BPDN architecture 704D may include non-powered (NP) rail segments without power rails between UNPR rail segments and grounded rails (GND). BPDN architecture 704D may have a cell architecture with cell boundaries (CB) and areas defined by usable span for equipment (USOD). In addition, BPDN architecture 704D may include UNPR rails, wherein power rail gaps 708 are provided between UNPR segments of the UNPR rails. Therefore, Figure 7D The BPDN architecture 704D can represent an area-efficient cell design with PR rail segments, power rail gaps 708, and grounding rails (GND). In some specific implementations, the BPDN architecture 704D can represent cells with extended power edges (EXT) and no power edges (NPE), which can be provided as a combination of extended power cells and no power cells.
[0058] In some specific implementations, Figure 7D express Figure 7D Exemplary use of the power gap unit in, wherein Figure 7D Figure 700D shows the use of the BPDN architecture 704D with power gap units, as well as NP units and EXT units.
[0059] like Figure 7DAs shown, the BPDN architecture 704D may have a first row (row_1) with multiple cells, which have various cell types, such as, for example, power gap cells, EXT cells, NP cells, and UNPR cells. The multiple cells may also include various standard cells, such as, for example, inverter cells (INV), buffer cells (BUF), and logic cells (e.g., NAND cells).
[0060] In some specific implementations, BPDN architecture 704D may represent a multi-cell architecture with multiple cell structures of multiple cell types, wherein each cell structure may have a rear power distribution network (BPDN), wherein the rear power rails are arranged in multiple cell types, and wherein each cell structure may have a ground edge (GND) and a configuration edge (e.g., SPWR, UNPR, power gap, PPWR), wherein the configuration edge may provide a no-power edge (NPE), an extended power edge (EXT), and / or an unnamed power edge (UNPR) with at least one interruption (i.e., power rail gap 708).
[0061] In some specific implementations, the first row (row_1) and the second row (row_2) may each include a certain number (N) of cells, such as, for example, four cells. For example, see reference... Figure 7E As shown, the first row (row_1) may include a buffer unit (BUF) as the first unit, a power gap unit as the second unit, an inverter unit (INV) as the third unit, and another inverter unit (INV) as the fourth unit. Additionally, the second row (row_2) may include a buffer unit (BUF) as the first unit, a power-free unit (NP) as the second unit, another power-free unit (NP) as the third unit, and a logic unit (NAND) as the fourth unit. In some cases, each unit structure in the BPDN architecture 704E may be coupled to a rear-side power distribution network (BPDN), where the rear-side power rails are arranged in multiple unit types to provide at least one of primary power supply (PPWR), secondary power supply (SPWR), ground (GND), or power-free (NP).
[0062] Referring to the cell types in BPDN architecture 704E, a first cell type may have a ground edge coupled to ground (GND) and a configuration edge that can provide a non-powered edge (NPE). Additionally, a second cell type may have a ground edge coupled to ground (GND) and a configuration edge that can provide an extended power edge (EXT) coupled to either the primary power supply (PPWR) or the secondary power supply (SPWR). Furthermore, a third cell type may have a ground edge coupled to ground (GND) and a configuration edge that can provide a power edge coupled to either the primary power supply (PPWR) or the secondary power supply (SPWR), wherein this power edge has a power rail gap 708 (or interruption), and another non-powered edge (NPE) is inserted into the gap space provided by the power rail gap 708 (or interruption).
[0063] As described herein, primary power supply can be provided with reference to a first power domain associated with the core voltage (PPWR), and secondary power supply can be provided with reference to a second power domain associated with the global core voltage (SPWR). Additionally, the ground edge can be coupled to ground (GND).
[0064] In some implementations, the secondary power / grounding back rail may reach the east / west edge of the unit, one edge of which may have a standard power rail (e.g., when the unit requires power from the primary power supply and if the unit requires power from the primary power supply), and the other edge may point to a secondary power rail. Additionally, the reaching edges may have assigned edge types depending on which power / ground rail reaches the unit edge. Rules may exist regarding how these edge types are configured to be adjacent during placement and wiring. Some identical edge types may be adjacent with zero spacing, and different edge types may be adjacent with an "n" spacing, where "n" is determined by the design rules of the various process technologies used in the physical design.
[0065] Some units have different power supply rails with adjacent edges that maintain rail spacing between unit edges, and power gap units can be designed with special edge types not found in other standard units. Power gap units can be configured to be adjacent to any other edge type with zero spacing and ensure that gaps created by placement and routing (PNR) tools between unit edges of the same edge type are filled. In addition to power gap units, each edge type may require special filler units.
[0066] Furthermore, instead of building area overhead within each cell with secondary power supply, area overhead can be applied at the PNR level. This allows the EDA tool to perform placement based on edge type and adjacency rules, as well as special cells created for this purpose. This concept saves area when the EDA tool places multiple cells with the same secondary power supply into rows that are closer together with zero spacing, and the area overhead can be applied only to the ends of the chain. Similarly, some cells with a standard power rail on one edge and a secondary power rail on another edge can be placed in the chain with zero spacing by alternately flipping the cells on their y-axis. This also reduces the area overhead from each cell to the end of the chain, and no additional area overhead is applied if the chain ends at the edge of a standard rail.
[0067] Figure 8 A system 800 for providing back-side power distribution technology by physical design, according to a specific implementation described herein, is shown.
[0068] refer to Figure 8 System 800 may be associated with a computer or computing device 804, which is implemented as a dedicated machine for providing back-end power distribution technology in a physical design, as described herein. In some cases, computing device 804 may refer to a device having any standard elements and / or components, including at least one processor 810, memory 812 (e.g., a non-transitory computer-readable storage medium), one or more databases 840, a power supply, peripheral devices, and other components that may not be present in the standard configuration. Figure 8 Various other computing elements and / or components are specifically illustrated herein. The computing device 804 may include instructions recorded or stored on a non-transitory computer-readable medium 812, which may be executed by at least one processor 810. The computing device 804 may be associated with a display device 850 (e.g., a monitor or other display) that can be used to provide a user interface (UI) 852, such as a graphical user interface (GUI). In some cases, the UI 852 may be used to receive various parameters and / or preferences from a user for managing, operating, and / or controlling the computing device 804. Therefore, the computing device 804 may include a display device 850 for providing various types of output to a user, and additionally, the display device 850 may include a UI 852 for receiving various types of input from a user.
[0069] For reference Figure 8 As shown, computing device 804 may include BPDN manager 820, which can be configured to enable processor 810 to implement the features described herein. Figures 1 to 7EThe described back-side power distribution scheme and techniques include methods related to implementing various associated integrated circuits (ICs) in the physical design. The BPDN manager 820 can be implemented in hardware and / or software. For example, if implemented in software, the BPDN manager 820 can be stored in memory 812 and / or database 840. Alternatively, in some cases, if implemented in hardware, the BPDN manager 820 and related components can refer to a separate processing unit configured to interact with processor 810.
[0070] In some cases, the BPDN manager 820 can be configured to cause at least one processor 810 to perform various operations, as referenced herein. Figures 1-7E The back-side power distribution scheme and technology described herein provide. Memory 812 may include instructions stored thereon that, when executed by processor 810, cause processor 810 to perform one or more of the following operations.
[0071] For example, the BPDN manager 820 can be configured to cause at least one processor 810 to perform operations associated with a cell architecture providing a back-side power distribution network having a back-side power rail with a primary power supply (PPWR). Additionally, at least one back-side power rail can be interrupted, and a secondary power supply (SPWR) is inserted into the gap space interval of the interruption.
[0072] The BPDN manager 820 can be configured to cause at least one processor 810 to perform operations associated with providing a back-side power distribution network having back-side power rails providing primary power supply (PPWR), secondary power supply (SPWR), and ground (GND). Additionally, the back-side power rails may include a first power rail having a first segment providing primary power supply (PPWR), a second segment providing secondary power supply (SPWR), and an interrupt separating the first and second segments. Furthermore, the back-side power rails may include a second power rail that is continuous and coupled to ground (GND).
[0073] The BPDN manager 820 can be configured to cause at least one processor 810 to perform operations associated with providing a back-side power distribution network having back-side power rails that provide primary power supply (PPWR), secondary power supply (SPWR), and ground (GND). Additionally, the back-side power rails may include one or more continuous power rails coupled to the primary power supply (PPWR), and the back-side power rails may include one or more interrupted power rails having a ground segment coupled to ground (GND), a power segment providing secondary power supply (SPWR), and one or more interruptions separating the ground segment from the power segment.
[0074] The BPDN manager 820 can be configured to cause at least one processor 810 to perform operations associated with a multi-cell architecture providing multiple cell structures with multiple cell types. In some cases, each cell structure has a rear power distribution network, wherein the rear power rails are arranged in multiple cell types, and each cell structure has a ground edge and configuration edges that provide no power edge, extended power edge, or power edge with interruption.
[0075] According to the reference Figures 1-7E The various specific implementations described herein, and any one or more of these operations performed by the BPDN manager 820, may be modified, altered, or changed to provide, for example... Figures 1-7E Various specific implementation schemes are shown. Furthermore, the back-side power distribution schemes and techniques described herein can be used to provide various integrated circuits in the form of logic blocks or modules having a physical structure included in the placement and wiring environment of electronic design automation (EDA) tools used in relevant EDA systems.
[0076] Additionally, refer to Figure 8 The computing device 804 may include a simulator 824 configured to enable the processor 810 to generate one or more simulations of integrated circuits. Simulator 824 may be referred to as a simulation component and / or module that can be implemented in hardware and / or software. If implemented in software, simulator 824 may be recorded or stored in memory 812 or database 840. If implemented in hardware, simulator 820 may refer to a separate processing component configured to interact with processor 810. In some cases, simulator 824 may refer to a SPICE simulator configured to generate SPICE simulations of integrated circuits. Generally, SPICE is an acronym for integrated circuit-focused simulation program, which stands for open-source analog electronic circuit simulator. Additionally, SPICE may refer to a general-purpose software program used by the semiconductor industry to check the integrity of integrated circuit designs and predict their behavior. Therefore, in some specific implementations, BPDN manager 820 may be configured to interact with simulator 824 to generate various timing data and information based on one or more simulations of the integrated circuit (including, for example, SPICE simulations) that can be used to analyze the performance characteristics of the integrated circuit (including timing data of the integrated circuit). Additionally, the BPDN Manager 820 can be configured to use simulations of the integrated circuit (e.g., SPICE simulations) to evaluate its operational behavior and conditions.
[0077] In some cases, computing device 804 may include one or more databases 840 configured to store and / or record various data and information related to the implementation of conversion rate characterization schemes and techniques in a physical design. In various cases, database 840 may be configured to store and / or record data and information related to integrated circuits, operating conditions, operating behavior, and / or timing data. Additionally, database 840 may be configured to store data and information related to integrated circuits and timing data from reference simulation data (e.g., SPICE simulation data).
[0078] The various approaches and techniques described herein provide a range of methods for electronic design automation (EDA) that enable the construction, design, and / or provision of area-efficient standard cell blocks to support back-side power distribution networks (BPDNs) using an array of area-efficient standard cells. For example, referring to the physical design of semiconductor integrated circuits, automation tools can be used at various stages of the physical design process, including logic synthesis where the logic design of the circuit, standard cell library information, and various other parameters are provided as input to generate a gate-level netlist. Additionally, placement and routing (PNR) techniques are used, where a gate-level netlist from logic synthesis, standard cell libraries, and / or various other parameters are provided as input to generate a physical layout design for integrated circuits and / or their components.
[0079] In some specific implementations, logic synthesis and PNR tools can be limited when seeking area efficiency for standard cells with any electrically isolated structures such as NWELL and / or power / ground rails, because standard cells are designed with area overhead built into each cell that requires such electrically isolated elements, and each instance of these cells results in area overhead. PNR tools typically require each standard cell to follow specific boundary conditions when adjacent to each other. To overcome some of these limitations, edge types for cell edges can be assigned at the cell level, and additionally, design rules can be provided for different edge types to be adjacent to each other or to maintain a minimum space between them. In some cases, even conventional edge type approaches may not provide area-efficient placement solutions when electrically isolated structures are present in the cell.
[0080] Therefore, the various solutions and techniques described herein achieve area-efficient synthesis and PNR processes by using specially designed standard cell layouts and methods when standard cells require electrically isolated structures (such as, for example, NWELLs with different potentials, power rails coupled to secondary power supplies that differ from the primary power supply in the physical cell design). Thus, the various solutions and techniques described herein are primarily applicable to semiconductor processes supporting back-side power distribution networks (BPDNs) with back-side power rails that are coupled to transistors on the front side via embedded vias to interrupt the primary power rail and / or ground rail for inserting secondary power rail segments and / or ground rail segments. Furthermore, the various solutions and techniques described herein can therefore be applied to semiconductor processes with front rails for inserting secondary rail segments by interrupting the primary rail and for isolating NWELLs with different potentials in an area-efficient manner. Additionally, in some cases, different potentials may refer to voltage values and / or differences between different power / voltage sources.
[0081] Figure 9 A composition and placement and routing (PNR) process 900 and related techniques are illustrated according to a specific implementation described herein. In various implementations, the composition and placement and routing (PNR) process 900 refers to a method for utilizing a physical unit layout structure that can be used to construct, generate, and / or provide design blocks.
[0082] It should be understood that even though process 900 specifies a particular order of operations, in some cases, the specific parts of the operations may be performed in a different order and on different systems. In other cases, additional operations and / or steps may be added to process 900 and / or omitted from the method. Furthermore, process 900 may be implemented in hardware and / or software. If implemented in hardware, process 900 may be implemented using various components and / or circuits, as referenced herein. Figures 1-8 As described herein. Additionally, if implemented in software, process 900 can be implemented as a program and / or software instruction process configured for synthesis and PNR technologies, as described herein. Furthermore, if implemented in software, the instructions related to implementing process 900 can be stored and / or recorded in memory and / or a database. For example, various types of computing devices having processors and memory, such as… Figure 8 The computer system 800 can be configured to execute process 900.
[0083] In various specific implementations, process 900 can refer to a method of designing, providing, constructing, fabricating, and / or manufacturing a power rail architecture as an integrated system, device, and / or circuit, which may involve using various IC circuit components described herein to implement various power distribution schemes and techniques associated with them. In some specific implementations, the power rail architecture may be integrated with various computing circuits and other related components on a single chip, and the power distribution circuitry may be implemented in some embedded systems for automotive, electronics, mobile, server, and Internet of Things (IoT) applications (including remote sensor nodes).
[0084] At block 910, process 900 can provide the logical definition of a design block for the RTL in a digital circuit design. In some implementations, the register transfer level (RTL) may refer to the physical layout design that characterizes the patterns and behavior of a digital circuit through the movement of data signals along digital signal paths between circuit components, and the RTL may have information related to the logical operations performed on the data signals. At block 914, process 900 can obtain, provide, and / or generate design constraints for the design block of the RTL in the digital circuit design. Additionally, at block 918, process 900 can access, obtain, and / or utilize information stored in one or more standard cell libraries.
[0085] At box 924, process 900 can interact with a synthesis tool configured to access, obtain, and / or utilize logical definitions from design blocks in box 910, design constraints from box 914, and / or information stored in one or more standard cell libraries in box 918. In some embodiments, process 900 is configured to interact with the synthesis tool to synthesize standard cells for a physical design of a power supply network with power supply rails. The power supply network may refer to different power supplies selectively coupled to each standard cell via power supply rails associated with each of the different power supplies. In various embodiments, synthesizing standard cells may refer to power edge type-aware synthesis, and the synthesis tool may be configured to process information provided in one or more standard cell libraries and related to various design constraints. The synthesis tool may be configured to synthesize standard cells with electrically isolated structures having area as a constraint, and the synthesis tool may be configured to group the synthesized standard cells together at a global level to achieve an area-efficient design for use by placement and wiring (PNR) tools.
[0086] At block 928, process 900 may interact with a Placement and Routing (PNR) tool configured to access, obtain, and / or utilize the synthesized standard cells from the physical design of block 924 and information stored in one or more standard cell libraries from block 918. In some embodiments, process 900 is configured to employ the Placement and Routing (PNR) tool to define the edge type of each standard cell in the physical design based on the power supply network and power supply rail that touch at least one edge of each standard cell in the standard cell. At least one edge of each standard cell refers to at least one of the upper edge, the lower edge opposite the upper edge, the right edge, and the left edge opposite the right edge. Additionally, in some embodiments, process 900 may be configured to employ the PNR tool to reduce the number of interruptions in the power supply rail by grouping cells with similar edge types together and to position standard cells with similar edge types adjacent to each other by using special cells defined in one or more standard cell libraries, while maintaining the minimum length rule of isolated power rail segments. In addition, process 900 can be configured to employ PNR tools to maintain the gap between the power supply rails of the standard unit when connected to different power supply networks.
[0087] In various specific implementations, a special unit may refer to and / or be defined as a powerless unit of a first edge type that defines a powerless edge type within the edge types, and additionally, a powerless unit may be configured not to provide an available power connection to a power supply rail of the power supply network. Furthermore, a special unit may refer to and / or be defined as an extended power unit of a second edge type that defines an extended power edge type within the edge types, and additionally, an extended power unit may be configured to provide an extended power supply rail that can be selectively connected to a power supply rail of an adjacent standard unit. Moreover, a special unit may refer to and / or be defined as a power gap unit of a third edge type that defines a power gap edge type within the edge types, and additionally, a power gap unit may be configured to provide an interrupted power rail with a gap space interval that allows different segments of the interrupted power rail of the power gap unit to be selectively connected to different power supply networks.
[0088] At box 934, process 900 can provide a physical layout of design blocks having standard cells synthesized by a synthesis tool and placed by a PNR tool. For example, refer to Figures 3A-7EThe physical layout of the design blocks may refer to a device with a cell architecture having a rear power rail. In this case, the device may include a stacked array of standard cells arranged to include an upper row, one or more inner rows, and a lower row, wherein each standard cell has a cell architecture with a configurable power supply rail, and wherein each of the multiple rows has multiple rail positions for the configurable power supply rail. The configurable power supply rail may include an upper rail position for an upper configurable power supply rail shared between upper adjacent standard cells in one or more inner rows, or located at the upper outer edge of the array for the upper row. Additionally, the configurable power supply rail may include a lower rail position for a lower configurable power supply rail shared between lower adjacent standard cells in the lower adjacent row, or located at the lower outer edge of the array for the lower row.
[0089] In some implementations, configurable power supply rails can be alternately coupled to a primary power supply and / or a primary ground supply. Additionally, at least one other configurable power supply rail for each standard unit can be configured as a removed rail, a single-segment rail, and / or a multi-segment rail. A removed rail may refer to one or more rail locations where no configurable supply rail is present, and a single-segment rail refers to one or more rail locations where a single-segment supply rail is coupled to a primary power supply, coupled to a secondary power supply, or not connected to any power supply. A multi-segment rail may refer to multiple segments spaced apart by gaps to provide interruptions between different rail segments coupled to a primary power supply, coupled to a secondary power supply, or not connected to any power supply. Furthermore, referring to a multi-row, single-segment configuration, at least one configurable power supply rail for each standard unit may be coupled to a primary power supply or a secondary power supply, or not connected to any power supply. Moreover, referring to a multi-row, multi-segment configuration, at least one configurable power supply rail for each standard unit may be coupled to a primary power supply or a secondary power supply, or not connected to any power supply.
[0090] In some cases, refer to Figures 7A-7EThe physical layout of the design blocks may refer to a device with special cells to complement an area-efficient cell design. In this case, the device may include a multi-cell architecture with multiple cell structures of various cell types, wherein each cell structure has a rear power distribution network, wherein the rear power rails are arranged in multiple cell types, and wherein each cell structure may include a ground edge and / or a configuration edge providing a power-free edge, an extended power edge, or a power edge with interruptions. Additionally, each cell structure may include a rear power distribution network, wherein the rear power rails are arranged in multiple cell types to provide at least one of a primary power supply, a secondary power supply, or ground. The multiple cell types may have a first cell type with a ground edge coupled to ground and a configuration edge providing a power-free edge. The multiple cell types may have a second cell type with a ground edge coupled to ground and a configuration edge providing an extended power edge coupled to either the primary or secondary power supply. Multiple unit types may include a third unit type with a grounded edge coupled to ground and a power edge providing coupling to a primary or secondary power supply, and the power edge may have an interruption, wherein another non-powered edge is inserted in the gap space interval provided by the interruption.
[0091] Figure 10 Figure 1000 illustrates a standard cell design placement 1004 for physical cell design according to various specific implementations described herein. In some cases, the standard cell design placement 1004 may refer to a set of area-efficient cells in a physical layout design that characterizes, defines, and / or constructs area-efficient standard cell blocks to thereby support a back-side power distribution network (BPDN). Additionally, Figure 10 A diagram is provided showing the standard cells placed in the physical layout design.
[0092] like Figure 10As shown, the standard cell design placement 1004 may include a number (N) of standard cells (e.g., cell_1, cell_2, ..., cell_9) arranged in a number (N) of rows (e.g., row_1, row_2, ..., row_5). Additionally, the standard cell design placement 1004 may include various power supply rails (e.g., Vdd / Vddx and / or Vss / Vssx) arranged along the length of each row (e.g., row_1, row_2, ..., row_5) between the standard cells (e.g., cell_1, cell_2, ..., cell_9). In some specific implementations, the first row (row_1) may not have any cells arranged therein, and the second row (row_2) may have multiple cells (3 cells: cell_1, cell_2, cell_3) arranged adjacent to each other across the second row (row_2). Additionally, the third row (row_3) may have multiple cells (2 cells: cell_5, cell_6) arranged adjacent to each other across the third row (row_3). Similarly, the fourth row (row_4) may have multiple cells (4 cells: cell_6, cell_7, cell_8, cell_9) arranged adjacent to each other across the fourth row (row_4). The fifth row (row_5) may not have any cells within it.
[0093] refer to Figures 9-10 The various schemes and techniques described herein provide computer-implemented algorithms for the area-efficient synthesis and placement of standard cells with electrically isolated structures, where the electrically isolated structures can refer to power rails and NWELLs. The schemes and techniques described herein can be used in semiconductor manufacturing processes that allow back-side power delivery using back-side power rails and / or back-side signal rails. In some specific implementations, the schemes and techniques described herein employ synthesis tools to synthesize RTLs using a library of standard cells with specially designed cells that do not carry the area overhead for electrically isolated structures, such as, for example, secondary supply rails in a process supporting a BPDN with back-side power rails, or isolated NWELLs, or secondary supply rails in a process supporting front-side power supply rails. Furthermore, the synthesis tools can handle the information provided in the standard cell library along with design constraints, and can treat standard cells with electrically isolated structures and their area impact as constraints. Moreover, these synthesized standard cells, when combined together, can contribute to achieving an overall area-efficient design.
[0094] In some implementations, during global placement, Placement and Routing (PNR) tools can be used to define the edge type on the fly for each cell that has been synthesized for placement in the design, based on the electrical isolation structure. This design may refer to the power supply network that the power rail touches for a given edge of the cell, and rail connectivity can be assigned to the correct power supply network. Global placement may also consider grouping cells with the same edge type together as much as possible to minimize rail breaks and reduce the area overhead associated with them. In various implementations, the edge of a cell may refer to East / West / North / South (or Left / Right / Top / Bottom), and the power supply network refers to the power supply name of the design, not the power rail name of the standard cell. In some cases, a standard cell may have a Vdd rail, but the Vdd rail of the same cell in different instances (PNR blocks) of the design may be coupled to one or more Vddx rails of the design (e.g., Vdd1 or Vdd2). Therefore, each instance of the same standard cell in the design may have different edge types provided on the fly, because in the design, the Vdd rail of the standard cell may be coupled to Vdd1 or Vdd2.
[0095] In some specific implementations, during standard cell placement legitimization, Placement and Routing (PNR) tools can be used to obtain standard cells with the same edge type to be placed as close to each other as possible in groups, while maintaining minimum length rules for isolated power rail segments by using special cells (no-power cells, extended power cells, and / or power gap cells) from the standard cell library. Proximity of identical rails is an important feature during global placement and placement legitimization. During standard cell placement legitimization, Placement and Routing (PNR) tools can be used to place standard cells in appropriate orientations based on available primary supply rail segments in the cell. Alternatively, the PNR tools can be configured to accept mappings from secondary supplies to primary supplies, and any secondary supply mapped to a primary supply can be placed at the primary supply location in the design. Additionally, during placement legitimization, the PNR tools can be used to maintain edge type rules and to provide necessary clearance between power rails in cells where power rails are coupled to different power supply networks.
[0096] It should be anticipated that the subject matter set forth in the claims is not limited to the specific embodiments and illustrations provided herein, but includes modifications of those embodiments according to the claims, including portions of the embodiments and combinations of various elements from different embodiments. It should be understood that in the development of any such embodiment, as in any engineering or design project, many embodiment-specific decisions may be made to achieve the developer's specific objectives, such as conforming to system-related and / or business-related constraints, which may vary between different embodiments. Furthermore, it should be understood that such development work can be complex and time-consuming; however, it remains a routine task of design, manufacture, and / or production for those skilled in the art who benefit from the various solutions and techniques described herein with reference to this disclosure.
[0097] This document describes various specific implementations of the method. The method may include synthesizing standard cells of a physical design having a power supply network with power supply rails. The method may include employing placement and wiring tools to define the edge type of each standard cell in the physical design based on the power supply network and power supply rails that touch at least one edge of each standard cell in the standard cell.
[0098] This document describes various specific implementations of a device having a stacked array of standard cells arranged to include an upper row, one or more inner rows, and a lower row. Additionally, each standard cell may have a cell architecture with a configurable power supply rail, and each of the multiple rows may have multiple rail positions for the configurable power supply rail. The multiple rail positions may include an upper rail position for an upper configurable power supply rail shared between upper adjacent standard cells in one or more inner rows, or located at the upper outer edge of the array for the upper row. The multiple rail positions may include a lower rail position for a lower configurable power supply rail shared between lower adjacent standard cells in the lower adjacent row, or located at the lower outer edge of the array for the lower row.
[0099] This document describes various specific implementations of a device with a multi-cell architecture, featuring multiple cell structures of various cell types. Additionally, each cell structure may have a rear-side power distribution network, where rear-side power rails are arranged in various cell types. Furthermore, each cell structure may have a grounded edge and configured edges providing no power edge, extended power edge, and / or power edge with interruption.
[0100] Various specific implementations have been referenced in detail, examples of which are shown in the accompanying drawings and illustrations. Numerous specific details are set forth in the following detailed description provided herein to provide a thorough understanding of the disclosure. However, various aspects of the disclosure provided herein can be practiced without these specific details and / or features. In some cases, various well-known methods, procedures, systems, components, circuits, and / or networks have not been described in detail so as not to unnecessarily obscure the details of the implementation.
[0101] It should also be understood that while the terms "first," "second," etc., may be used herein to describe various elements, these elements should not be limited by these terms. These terms are used only to distinguish one element from another. 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. Both the first and second elements are elements, but they are not considered the same element.
[0102] The terminology used in the description of this disclosure provided herein is for the purpose of describing particular specific embodiments and is not intended to limit the scope of the disclosure provided herein. As used in the disclosure provided herein and in the description of the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. As used herein, the term “and / or” means and covers any and all possible combinations of one or more of the associated listed items. When used in this specification, the terms “comprising,” “including,” and / or “containing” specify the presence of the stated features, integers, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and / or groups thereof.
[0103] As used herein, depending on the context, the term "if" can be interpreted as meaning "when," "at," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if it is determined..." or "if [the condition or event] is detected" can be interpreted as meaning "in response to determination..." or "in response to detection of [the condition or event]." The terms "up" and "down"; "upper part" and "lower part"; "upward" and "downward"; "below" and "above"; and other similar terms indicating the relative position above or below a given point or element may be used in conjunction with some specific implementations of the various techniques described herein.
[0104] While the foregoing relates to specific implementations of the various techniques described herein, other and additional specific implementations can be conceived based on the disclosure herein, which can be defined by the appended claims. Although the subject matter has been described in language specific to structural features and / or methodological actions, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are disclosed as exemplary forms for implementing the claims.
Claims
1. A method for physical design of integrated circuits, comprising: Synthesize standard units for the physical design of a power supply network with power supply rails, wherein the synthesis of these standard units refers to power edge type sensing synthesis; and Placement and wiring tools are employed to define the edge type of each standard cell in the physical design based on the power supply network and the power supply rail that touch at least one edge of each standard cell in the standard cell.
2. The method of claim 1, wherein the at least one edge of each standard unit refers to at least one of an upper edge, a lower edge opposite to the upper edge, a right edge, and a left edge opposite to the right edge.
3. The method of claim 1, wherein the power supply network refers to different power supplies, which are selectively coupled to each standard unit via a power supply rail associated with each of the different power supplies.
4. The method according to claim 1, further comprising: A synthesis tool is provided, which is configured to process information provided in a standard cell library and related to design constraints.
5. The method according to claim 4, wherein: The synthesis tool synthesizes the standard cell with an electrically isolated structure having an influence area as a constraint, and The synthesis tool groups the synthesized standard cells together at the global level to achieve an area-efficient design for use by the placement and wiring tools.
6. The method according to claim 1, further comprising: The placement and wiring tools are used to reduce the number of interruptions in the power supply rails by grouping units with the same edge type and to position standard units with the same edge type adjacent to each other by using special units from the standard unit library, while maintaining the minimum length rule for isolated power rail segments. The special unit refers to a powerless unit having a first edge type that defines a powerless edge type in the edge types, wherein the powerless unit is configured not to provide an available power connection to the power supply rail of the power supply network.
7. The method according to claim 1, further comprising: The placement and wiring tools are used to reduce the number of interruptions in the power supply rails by grouping units with the same edge type and to position standard units with the same edge type adjacent to each other by using special units from the standard unit library, while maintaining the minimum length rule for isolated power rail segments. The special unit refers to an extended power unit having a second edge type that defines an extended power edge type in the edge type, wherein the extended power unit is configured to provide an extended power supply rail that can be selectively connected to an adjacent standard unit.
8. The method according to claim 1, further comprising: The placement and wiring tools are used to reduce the number of interruptions in the power supply rails by grouping units with the same edge type and to position standard units with the same edge type adjacent to each other by using special units from the standard unit library, while maintaining the minimum length rule for isolated power rail segments. The special unit refers to a power gap unit of a third edge type that defines a power gap edge type in the edge types, wherein the power gap unit is configured to provide interrupted power rails with gap space intervals that allow different segments of the interrupted power rails of the power gap unit to be selectively connected to different power supply networks.
9. The method according to claim 1, further comprising: The placement and wiring tools are used to maintain the gap between the power supply rails of the standard unit when connecting to different power supply networks.
10. An integrated circuit device, comprising: The rear-side power distribution network (BPDN) has a rear-side power rail. as well as A standard cell array, arranged as a stack of multiple rows including an upper row, one or more inner rows, and a lower row, wherein each standard cell has a cell architecture with a configurable power supply rail, wherein the configurable power supply rail is a rear power rail of the BPDN, and wherein each of the multiple rows has multiple rail positions for the configurable power supply rail, the multiple rail positions including: The upper rail position of the configurable power supply rail, which is shared among the upper adjacent standard cells in the upper adjacent rows of one or more inner rows, or located at the upper outer edge of the array for the upper row, and The lower rail position of the configurable power supply rail is shared between the lower adjacent standard cells in the lower adjacent row, or is located at the lower outer edge of the array for the lower row.
11. The integrated circuit device of claim 10, wherein the configurable power supply rail is alternately configured to be coupled to a primary power supply and a primary ground supply.
12. The integrated circuit device according to claim 10, wherein: At least one additional configurable power supply rail for each standard unit is configured as a removal rail, a single-segment rail, or a multi-segment rail.
13. The integrated circuit device according to claim 12, wherein: The removal of a rail refers to the absence of a configurable supply rail at one or more of the plurality of rail positions. The single-segment rail refers to a single supply rail located at one or more of the plurality of rail positions that is coupled to a primary power supply, coupled to a secondary power supply, or not connected to any power supply. The multi-segment rail refers to multiple segments separated by gaps to provide interruptions between different rail segments coupled to a primary power supply, coupled to a secondary power supply, or not connected to any power supply.
14. The integrated circuit device according to claim 10, wherein: Referring to multi-row, single-segment configurations, each standard unit may have at least one configurable power supply rail coupled to a primary or secondary power supply, or not connected to any power supply. Referring to multi-row, multi-segment configurations, at least one configurable power supply rail of each standard unit may be coupled to a primary power supply or a secondary power supply, or may not be connected to any power supply.
15. An integrated circuit device, comprising: Multi-unit architecture, wherein the multi-unit architecture has multiple unit structures of various unit types, Each unit structure has a rear power distribution network, wherein the rear power rails are arranged in the various unit types, and Each unit structure has a grounded edge and configuration edges that provide a power-free edge, an extended power edge, or a power edge with interruption.
16. The integrated circuit device according to claim 15, wherein: Each unit structure includes the rear power distribution network, wherein the rear power rails are arranged in the various unit types to provide at least one of primary power supply, secondary power supply, and grounding.
17. The integrated circuit device of claim 16, wherein the plurality of cell types includes: A first unit type, the first unit type having a grounded edge coupled to ground and a configuration edge providing the power-free edge.
18. The integrated circuit device of claim 16, wherein the plurality of cell types includes: The second unit type has a grounding edge coupled to ground and a configuration edge that provides an extended power edge coupled to the primary power supply or the secondary power supply.
19. The integrated circuit device of claim 16, wherein the plurality of cell types includes: The third unit type has a grounding edge coupled to ground and a configuration edge providing a power edge coupled to the primary power supply or the secondary power supply, wherein the power edge has an interruption, wherein another non-power edge is inserted in the gap space interval provided by the interruption.
Citation Information
Patent Citations
Integrated circuit, method of generating a layout of an integrated circuit using standard cells, and a standard cell library providing such standard cells
US20120286858A1
Integrated circuit including multi-height standard cell and method of designing the same
US20190393205A1
Semiconductor device including standard cells with combined active region
US20210224458A1
