Chip packaging method and structure, chip, and design method
Patent Information
- Application Number
- MYPI2025006273
- Authority / Receiving Office
- MY · MY
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-04-10
- Filing Date
- 2024-04-10
- Publication Date
- 2024-10-10
AI Technical Summary
In existing chip packages, the parasitic resistance and capacitance (RC) caused by too long wires is large, causing signal delay and distortion. Increasing the number of chip pins to improve connection capabilities will increase cost and area, making it difficult to meet high integration density and storage bandwidth. needs.
Using face-to-face direct interconnection, signal transmission between the wire redistribution layer (RDL) and the bump array layer reduces wire length, increases the number of I/Os, and provides a high-speed channel for signal transmission through pin wells, reducing the cost of The signal transmission distance and area between chips.
It achieves high-speed signal connection, reduces signal delay and distortion, improves the overall performance and integration of the chip, reduces the signal transmission distance and area between chips, and increases bandwidth.
Smart Images

Figure PI2025006273_DRAWING_001 
Figure PI2025006273_DRAWING_002 
Figure PI2025006273_DRAWING_003
Abstract
Description
Chip packaging method, structure, chip and design method Technical Field
[0001] The present invention relates to the field of chip technology, and in particular to a chip packaging method, structure, chip and design method. Background Art
[0002] Currently, packaging multiple chips has been proven in various fields to be an effective method for improving overall chip performance. However, current chip packaging suffers from complex wiring and long wires, resulting in large parasitic resistance and capacitance (RC), causing signal delays and distortion between chips.
[0003] The existing method of connecting chips is to distribute the chip pins in a one-dimensional pattern around the chip, and then interconnect them through wires. However, this connection method has many problems. For example, referring to the existing chip shown in Figures 4a-4b, Figure 4a is a top view of the chip. Chip 1, chip 2, and chip 3 are on the same plane, and the three are connected horizontally by connecting wires. Figure 4b is a side view of the chip. As can be seen from Figure 4b, the connecting wires 104 of chip 1 and chip 2 first penetrate vertically into the substrate 103 and then connect horizontally inside the substrate 103. This chip connection method has at least the following defects:
[0004] 1) If the number of pins per unit side length of a chip is n, the maximum number of I / Os that can be directly connected from the four sides is 4n. However, when the number of pins is large, the pin spacing will be very small, resulting in wiring congestion and even the need to increase the chip area due to pin limitations.
[0005] 2) Small pin spacing can also cause packaging problems. For example, small spacing can cause pin short circuits, increasing packaging difficulty and cost.
[0006] 3) The length of the connecting wire ranges from a few millimeters to tens of millimeters, with an average of more than 5 millimeters. Since the bandwidth is directly inversely proportional to the resistance and capacitance (RC) of the wire, and RC is proportional to the square of the physical length of the wire, RC will limit the bandwidth.
[0007] Therefore, as the chip design integration density increases and the demand for storage bandwidth increases, the existing connection method will be difficult to meet the design requirements.
[0008] As the world evolves, chip design demands higher integration density and memory bandwidth. For example, a chip needs to connect to multiple DRAM or RF chips, increasing the amount of data transmitted between chips. Existing technologies primarily address this by increasing the number of chip pins. However, simply increasing the number of pins significantly increases chip production costs and even increases chip area, impacting chip integration.
[0009] Summary of the Invention
[0010] The purpose of the embodiments of the present invention is to provide a chip packaging method, structure, chip and design method, so as to at least partially solve the above technical problems.
[0011] To achieve the above-mentioned purpose, an embodiment of the present invention provides a chip packaging method, comprising the following steps performed in sequence: preparing an initial chip structure, the initial chip structure comprising a wire redistribution layer RDL, a chip layer and a bump array layer stacked in sequence, wherein the chip layer comprises at least one first chip unit, the first chip unit comprising a first bare chip and a connection structure distributed on both sides of the first bare chip, the connection structure being capable of realizing signal transmission between the RDL and the bump array layer; providing at least one second chip unit, wherein each second chip unit corresponds to a first chip unit, and each second chip unit comprises a second bare chip; flip-chipping the second bare chip in the second chip unit to the first bare chip in the corresponding first chip unit and the bump array layer on the connection structure in a face-to-face direct interconnection manner; cutting out each corresponding group of first chip units and second chip units as a single chip structure; for the single chip structure, attaching a heat sink on the back of the second bare chip; and preparing a first solder ball on the RDL to obtain a final chip packaging structure.
[0012] Optionally, the preparation of the initial chip structure includes: providing the first bare chip and the connection structure to form the first chip unit; placing the front connection points of the first bare chip and the connection structure downward to be placed on the upper surface of the carrier; performing a molding operation on the first bare chip and the connection structure; preparing the RDL on the back side of the first bare chip and the connection structure, and removing the carrier from the front side; performing a flip-chip mold thinning operation on the first chip unit after removing the carrier to expose the connection points; and preparing the bump array layer on the exposed connection points.
[0013] Optionally, the connection structure is a chip with through silicon vias.
[0014] Optionally, the chip packaging method further includes: performing bottom filling or molding on the first chip unit and the second chip unit after the flip chip is completed and before the cutting.
[0015] Optionally, the chip packaging method also includes: electrically connecting an external device through the first solder ball to realize signal transmission between the first bare chip and the external device based on the first solder ball, and realizing signal transmission between the second bare chip and the external device based on the first solder ball and the connection structure.
[0016] An embodiment of the present invention also provides a chip packaging structure implemented by any of the above-mentioned chip packaging methods, comprising: an initial chip structure, which includes a wire redistribution layer RDL, a chip layer and a bump array layer stacked in sequence, wherein the chip layer includes at least one first chip unit, the first chip unit includes a first bare chip and a connection structure distributed on both sides of the first bare chip, and the connection structure can realize signal transmission between the RDL and the bump array layer; at least one second chip unit, wherein each second chip unit corresponds to a first chip unit, and each second chip unit includes a second bare chip, wherein the second bare chip in the second chip unit is flip-chip mounted to the first bare chip in the corresponding first chip unit and the bump array layer on the connection structure in a face-to-face direct interconnection manner; for a single chip structure including a corresponding group of first chip units and second chip units cut out after the flip-chip mounting is completed, a heat sink mounted on the back of the second bare chip; and a first solder ball prepared on the RDL.
[0017] Optionally, in the initial chip structure: the RDL is located on the back sides of the first bare chip and the connection structure, and the front sides of the first bare chip and the connection structure expose connection points for connecting to the bump array layer.
[0018] Optionally, the connection structure is a chip with through silicon vias.
[0019] Optionally, the first solder ball is electrically connected to an external device to enable signal transmission between the first bare chip and the external device based on the first solder ball, and to enable signal transmission between the second bare chip and the external device based on the first solder ball and the connection structure.
[0020] Through the above technical solution, the embodiment of the present invention realizes a high-speed connection between two bare chips based on the shortest wire through a face-to-face direct interconnection between the second bare chip, the first bare chip and the connection structure.
[0021] An embodiment of the present invention provides a chip, comprising: a substrate; and a first chipset and a second chipset, respectively arranged on opposite sides of the substrate, wherein the first chipset includes at least a first sub-chip, and the second chipset includes at least a second sub-chip; wherein the first sub-chip and the second sub-chip are interconnected by face-to-face pins through the substrate.
[0022] Optionally, the first chipset includes a first sub-chip, and the second chipset includes at least two second sub-chips.
[0023] Optionally, the first chipset includes at least two first sub-chips, and the second chipset includes one second sub-chip.
[0024] Optionally, a connecting wire is provided in the substrate; the first sub-chip and the second sub-chip are interconnected by the face-to-face pins via the connecting wire.
[0025] Optionally, the chip further includes: a second solder ball and a peripheral wire; the peripheral wire is connected to the second sub-chip; and the second solder ball is used to lead out the peripheral wire.
[0026] Optionally, the first sub-chip and the second sub-chip are radio frequency chips or digital chips.
[0027] On the other hand, the present invention also proposes a design method for realizing the interconnection of multiple chips, which includes: setting a substrate; and setting a first chipset and a second chipset on opposite sides of the substrate, respectively, wherein the first chipset includes at least a first sub-chip, and the second chipset includes at least a second sub-chip; wherein the first sub-chip and the second sub-chip are interconnected face-to-face through the substrate.
[0028] Optionally, the first sub-chip and the second sub-chip are interconnected by face-to-face pins through the substrate, including: setting the first sub-chip to the first side of the substrate; setting the second sub-chip to the second side of the substrate opposite to the first side in a flip-chip manner; the distance between the connection points of the first sub-chip and the second sub-chip is less than 1 mm.
[0029] Optionally, the design method further includes: after completing the face-to-face pin interconnection, bottom filling the pins.
[0030] Optionally, the design method further includes: performing vertical bonding on the substrate to prepare peripheral wires of the second sub-chip; performing compression molding on the peripheral wires and the first sub-chip; and setting solder joints to lead out the peripheral wires.
[0031] Optionally, a connecting line is provided in the substrate; the first sub-chip and the second sub-chip are interconnected by the face-to-face pins via the connecting line.
[0032] A chip of the present invention comprises: a substrate; and a first chipset and a second chipset respectively arranged on opposite sides of the substrate, wherein the first chipset comprises at least a first sub-chip, and the second chipset comprises at least a second sub-chip; wherein the first sub-chip and the second sub-chip are interconnected by face-to-face pins through the substrate.
[0033] The present invention uses face-to-face connection to significantly increase the number of I / Os under the same line spacing conditions, reduce the distance between chips, shorten the length of wires, increase bandwidth, and significantly improve the overall performance of the chip.
[0034] An embodiment of the present invention provides a chip, which includes: a base layer; a metal layer arranged on the base layer, the metal layer including at least a surface metal and a bottom metal; a plurality of connection points arranged on the surface metal; and a pin well arranged on the bottom metal and connected to some of the connection points in a one-to-one correspondence, and the pin well includes a plurality of pins connected longitudinally in sequence.
[0035] Optionally, the pin is a perforated structure; the pin well connects the surface metal and the bottom metal.
[0036] Optionally, the length difference between any two pins in the pin well is less than 10%.
[0037] Optionally, the pin well is provided with a driving gate circuit or a receiving gate circuit.
[0038] Optionally, the pin well is perpendicular to the surface metal.
[0039] On the other hand, the present invention also proposes a chip design method, which includes: setting a base layer; setting a metal layer on the base layer, the metal layer including at least a surface metal and a bottom metal; and setting a plurality of connection points on the surface metal; wherein, the bottom metal is provided with a pin well which is perpendicular to the surface metal and is connected to some connection points one-to-one, and the pin well includes a plurality of pins connected longitudinally in sequence.
[0040] Optionally, the pin is a perforated structure; the pin well connects the surface metal and the bottom metal.
[0041] Optionally, the length difference between any two pins in the pin well is less than 10%.
[0042] Optionally, the pin well is provided with a driving gate circuit or a receiving gate circuit.
[0043] Optionally, the pin well is perpendicular to the surface metal.
[0044] The chip of the present invention includes: a base layer; a metal layer arranged on the base layer, the metal layer including at least a surface metal and a bottom metal; a plurality of connection points arranged on the surface metal; and a pin well arranged on the bottom metal and connected to some of the connection points in a one-to-one correspondence, and the pin well includes a plurality of pins connected in sequence longitudinally.
[0045] The present invention provides a high-speed channel for signal transmission by setting a pin well, effectively shortens the distance of signal transmission between chips, reduces the area of the chip, and increases the integration of the chip.
[0046] Other features and advantages of the embodiments of the present invention will be described in detail in the subsequent detailed description. BRIEF DESCRIPTION OF THE DRAWINGS
[0047] One or more embodiments are exemplarily illustrated by pictures in the corresponding drawings. These exemplifications do not constitute limitations on the embodiments. Elements with the same reference numerals in the drawings are represented as similar elements. Unless otherwise stated, the figures in the drawings do not constitute proportional limitations.
[0048] FIG1 is a schematic flow chart of a chip packaging method according to a first embodiment of the present invention;
[0049] 2( a ) to 2 ( h ) are schematic diagrams of various steps of a chip packaging method according to an example of an embodiment of the present invention, and also illustrate an exemplary chip packaging structure according to an embodiment of the present invention;
[0050] FIG3 is a schematic diagram of a process for preparing an initial chip structure in Example 1 of the present invention;
[0051] Figures 4a-4b are schematic diagrams of a chip in the prior art;
[0052] FIG5 is a schematic diagram of a chip of the present invention;
[0053] FIG6 is a schematic diagram of a non-differential signal circuit of the present invention;
[0054] FIG7 is a schematic diagram of an existing differential signal circuit;
[0055] 8a-8g are schematic flow charts of a design method for interconnecting multiple chips according to the present invention;
[0056] 9a-9b are schematic diagrams of a specific embodiment of a chip of the present invention;
[0057] 10a-10f are flowcharts of another design method for interconnecting multiple chips according to the present invention;
[0058] Figures 11a and 11b are schematic diagrams of a chip of the present invention;
[0059] FIG12 is a schematic diagram of a pin well of the present invention;
[0060] FIG13 is a schematic diagram of another pin well of the present invention;
[0061] FIG14 is a schematic diagram of an embodiment of a chip of the present invention;
[0062] 15a and 15b are schematic diagrams of embodiments of the pin well of the present invention;
[0063] FIG16 is a flow chart of a chip design method according to the present invention.
[0064] Reference Numerals: 110 - First Bare Die; 120 - Connection Structure; 200 - Carrier; 300 - Mold Layer; 400 - RDL; 500 - Bump Array Layer; 600 - Second Bare Die; 700 - Heat Sink; 800 - First Solder Ball. 101 - First Sub-Chip; 102 - Second Sub-Chip; 103 - Substrate; 104 - Connection Wire; 105 - Second Solder Ball. 201 - Base Layer; 202 - Metal Layer; 203 - Connection Point; 204 - Pin Well; 205 - Surface Metal; 206 - Bottom Metal; 207 - Pin. DETAILED DESCRIPTION
[0065] The following describes the specific implementation of the embodiment of the present invention in detail with reference to the accompanying drawings. It should be understood that the specific implementation described herein is only used to illustrate and explain the embodiment of the present invention and is not used to limit the embodiment of the present invention.
[0066] Before introducing the solutions of the embodiments of the present invention, some of the terms involved are introduced so that those skilled in the art can better understand the solutions of the embodiments of the present invention.
[0067] 1) Chip packaging: It refers to the protection of bare chips to prevent them from being damaged by the outside world. Different packaging technologies vary greatly in preparation procedures and processes.
[0068] 2) Flip-chip: This refers to the process of flipping a chip, which means that the chip's connection points (also known as bumps) face downward for operation, and are connected to a substrate, carrier, circuit board, another chip, etc. For example, flip-chip bonding refers to the process of connecting the chip's connection points downward to a substrate, carrier, circuit board, another chip, etc. Among them, a bump is a typical connection point. In an embodiment of the present invention, the flip-chip process, for example, is to form metal solder balls at both ends of the chip using a planar process, attach the chip face-down to the substrate, and use a reflow process to form solder joints between the chip solder balls and the substrate pads, thereby achieving electrical, thermal, and mechanical connections between the chip and the substrate.
[0069] 3) Underfill: This involves applying a dot of epoxy resin glue to the edge of the flip-chip frame. Through the "capillary effect," the glue is drawn to the opposite side of the frame, completing the underfill process. The glue is then cured by heating, resulting in a reliable and stable chip process. The underfill process generally consists of two steps: the initial underfill process and the subsequent curing process.
[0070] 4) Face-to-face direct interconnection: For example, for two chips, one chip is vertically soldered to the flip-chip bump array of another chip without additional substrates, wires, RDLs, etc. to achieve chip interconnection.
[0071] 5) Compression Molding: This refers to the chip compression process, where the chip's connection frame is placed in a mold and a curing material is injected into the mold to compress and form a mold layer structure that protects the chip's connection frame. This process involves compressing the curing material, such as encapsulating the chip or conductor structure with a die head whose die surface is no smaller than the size of the chips in parallel.
[0072] 6) Substrate and carrier: The substrate has electrical properties and has wiring inside so that the bare chip can transmit signals horizontally and vertically through the wiring; the carrier does not have electrical properties and only serves as a mechanical load-bearing device.
[0073] 7) Single chip: refers to a single chip that has been cut and separated (i.e., no longer needs to be cut), which can independently perform specific calculations after packaging. It can integrate multiple bare chips to perform multiple calculations, such as a powerful CPU, GPU and AI chip.
[0074] 8) Face-to-face pin interconnection: This refers to, for example, for two chips, directly connecting the pins of one chip to the flip-chip pin array of another chip, such as directly connecting the pins through connecting wires, which can pass through part of the substrate, and the part of the substrate provides direct face-to-face interconnection of the chips through high-density vertical wires, and the wire length is approximately equal to the substrate thickness.
[0075] Example 1.
[0076] Figure 1 is a schematic flow chart of a chip packaging method according to a first embodiment of the present invention. The chip packaging method includes the following steps S100-S600. Figures 2(a)-2(h) are schematic process diagrams of an example of applying the chip packaging method, including steps s1-s8. Combining Figures 1 and 2(a)-2(h), the chip packaging method includes the following steps S100-S600.
[0077] Step S100: preparing an initial chip structure.
[0078] As shown in Figure 2(e), the initial chip structure includes a wire redistribution layer (RDL) 400, a chip layer and a bump array layer 500 stacked in sequence, wherein the chip layer includes at least one first chip unit, the first chip unit includes a first bare chip 110 and a connection structure 120 distributed on both sides of the first bare chip 110, and the connection structure 120 can realize signal transmission between the RDL400 and the bump array layer 500.
[0079] In a preferred embodiment, as shown in FIG3 , the preparation of the initial chip structure implemented in step S100 may include the following steps S110 - S160 .
[0080] Step S110 : providing the first bare chip and the connection structure to form the first chip unit.
[0081] Step S120, placing the first bare chip and the connection structure with their front connection points facing downward on the upper surface of a carrier;
[0082] Step S130 : performing compression molding on the first bare chip and the connection structure.
[0083] Step S140 , preparing the RDL on the back side of the first bare chip and the connection structure, and performing a carrier removal process on the front side.
[0084] In step S150 , the first chip unit is subjected to a flip-chip thinning process after the carrier is removed to expose the connection points.
[0085] Step S160 , preparing the bump array layer on the exposed connection points.
[0086] 2( a ) to 2 ( e ), the implementation of steps S110 - S150 corresponds to the exemplary steps s1 - s5 .
[0087] Step s1: As shown in FIG. 2( a ), a first chip unit is provided, comprising a first bare chip 110 and a connection structure 120 ; and the first bare chip 110 and the connection structure 120 are placed on the upper surface of the carrier 200 with their front connection points facing downward.
[0088] The connection structure 120 is, for example, a chip with through-silicon vias (TSVs), which enable signal transmission between modules placed on its upper and lower surfaces. As shown in FIG2(e), the TSVs of the chip enable signal transmission between the RDL 400 and the bump array layer 500. Alternatively, the connection structure 120 may be an organic substrate with internal wiring, which can also enable signal transmission between modules on the upper and lower surfaces of the substrate.
[0089] Step s2: As shown in FIG. 2( b ), the first bare chip 110 and the connection structure 120 are subjected to compression molding to obtain a corresponding mold layer structure 300 .
[0090] Step s3: As shown in FIG. 2( c ), the carrier 200 on the front side of the chip unit is removed, and an RDL 400 is prepared on the back side of the first bare chip 110 and the connection structure 120 .
[0091] Step s4: As shown in FIG2(d), a flip-chip mold thinning process is performed (equivalent to rotating FIG2(c) 180 degrees and then performing mold thinning) to expose the connection points of the chip.
[0092] Step s5: As shown in FIG. 2( e ), a bump array layer 500 is formed on the exposed connection points.
[0093] In step S200 , at least one second chip unit is provided, wherein each second chip unit corresponds to one first chip unit and each second chip unit includes a second bare chip.
[0094] In an example, a corresponding group of second chip units and first chip units are determined to belong to a single chip.
[0095] Step S300 , flip-chip bonding the second bare chip in the second chip unit to the first bare chip in the corresponding first chip unit and the bump array layer on the connection structure in a face-to-face direct interconnection manner.
[0096] This step S300 corresponds to the process s6 of the example.
[0097] Step s6: As shown in FIG. 2( f ), the second bare chip 600 is bonded to the first bare chip 110 and the bump array layer 500 on the connection structure 120 in a face-to-face direct interconnection manner.
[0098] The definition of face-to-face direct interconnection has been given above. In conjunction with Figure 2(f), it should be pointed out that face-to-face direct interconnection has the following two characteristics:
[0099] 1) The second bare chip 600 is directly connected to the first bare chip 110 and the connection structure 120 without using RDL, for example, by wire bonding;
[0100] 2) The "face-to-face" connection structure between the second bare chip 600 and the first bare chip 110 is manifested in that the wires used for welding between the two are vertical. Based on the theory that "the straight line between two points is the shortest", it is easy to know that the wires currently connecting the corresponding connection points of the two bare chips are the shortest, for example, less than 1 mm.
[0101] Step S400 , after the flip-chip bonding is completed, cutting out each corresponding group of first chip units and second chip units to form a single chip structure.
[0102] Here, step S300 forms a chip package structure prepared on a large wafer or panel. A single chip structure is required in the application, and then a solution for obtaining the single chip structure through a cutting process is proposed in step S400.
[0103] The cutting process is a conventional technical means for those skilled in the art and will not be described in detail here.
[0104] Step S500: For the single chip structure, a heat sink is mounted on the back side of the second bare chip.
[0105] Corresponding step s7 : As shown in FIG. 2( g ), a heat sink 700 is added to the back surface of the second bare chip 600 .
[0106] Step S600 , preparing a first solder ball on the RDL to obtain a final chip packaging structure.
[0107] Corresponding step s8: as shown in FIG2(h), a first solder ball 800 is prepared on the RDL 400 to obtain a final chip packaging structure.
[0108] In an example, an external device can be electrically connected via the first solder balls 800 to enable signal transmission between the first bare chip 110 and the external device via the first solder balls 800, and signal transmission between the second bare chip 600 and the external device via the first solder balls 800 and the connection structure 120. For example, if the external device is a power supply, power can be ensured to the first bare chip 110 and the second bare chip 600.
[0109] In summary, the chip packaging method according to the embodiment of the present invention has at least the following two advantages.
[0110] 1) Through direct face-to-face interconnection between the second die, the first die, and the connection structure, a connection between the two die is achieved using the shortest possible wire. Based on the principle that "the longer the wire, the greater the RC," it is clear that shortening the wire significantly reduces RC, thereby reducing signal delay and distortion, increasing the bandwidth of signal transmission between the two die, and achieving a high-speed connection between the two die. In this example, this high-speed connection can be applied to integrated chips combining DRAM and CPU / GPU, or to integrated chip structures combining RF chips and digital chips in communication systems, significantly improving the overall performance of the integrated chip.
[0111] 2) Through the embodiment of the present invention, a single chip structure can be obtained by cutting, and a heat sink and a first solder ball can be further added. The addition of the heat sink ensures the normal heat dissipation of high-power chips such as CPU and GPU, while the addition of the first solder ball ensures the signal transmission between the bare chip and the external device.
[0112] Example 2.
[0113] A second embodiment of the present invention provides a chip packaging structure, as shown in FIG2(a) to FIG2(h), which is prepared by the method of the first embodiment, and the chip packaging structure includes: an initial chip structure, the initial chip structure including a wire redistribution layer RDL400, a chip layer and a bump array layer 500 stacked in sequence, wherein the chip layer includes at least one first chip unit, the first chip unit includes a first bare chip 110 and a connection structure 120 distributed on both sides of the first bare chip 110, the connection structure 120 can realize signal transmission between the RDL400 and the bump array layer 500; at least A second chip unit, wherein each second chip unit corresponds to a first chip unit, and each second chip unit includes a second bare chip 600, wherein the second bare chip 600 in the second chip unit is flip-chip mounted to the first bare chip 110 in the corresponding first chip unit and the bump array layer 500 on the connection structure 120 in a face-to-face direct interconnection manner; for a single chip structure including a corresponding group of first chip units and second chip units cut out after the flip-chip mounting is completed, a heat sink 700 is mounted on the back side of the second bare chip 600; and a first solder ball 800 prepared on the RDL400.
[0114] In a preferred embodiment, in the initial chip structure, the RDL 400 is located on the backside of the first bare chip 110 and the connection structure 120 , and the front sides of the first bare chip 110 and the connection structure 120 expose connection points for connecting to the bump array layer 500 .
[0115] In a preferred embodiment, the first solder ball 800 is electrically connected to an external device to enable signal transmission between the first bare chip 110 and the external device based on the first solder ball 800, and to enable signal transmission between the second bare chip 600 and the external device based on the first solder ball 800 and the connection structure 120.
[0116] More implementation details and effects of the chip packaging structure can be found in the aforementioned embodiment 1 of the chip packaging method, and will not be described in detail here.
[0117] Example 3.
[0118] Figure 5 is a schematic diagram of a chip of the present invention. As shown in Figure 5, a chip of the present invention includes: a substrate 103; and a first chipset and a second chipset respectively arranged on opposite sides of the substrate 103, wherein the first chipset includes at least a first sub-chip 101, and the second chipset includes at least a second sub-chip 102; wherein the first sub-chip 101 and the second sub-chip 102 are interconnected face-to-face through the substrate.
[0119] The substrate 103 provides electrical connection, protection, support, heat dissipation, and assembly for the chips, enabling multi-pin, reduced package size, improved electrical performance and heat dissipation, ultra-high density, or multi-chip modularization. The substrate at the interface between the chips allows for direct face-to-face interconnection via high-density vertical conductors.
[0120] In a preferred embodiment, the chip connection method of the embodiment of the present invention can also omit the substrate 103 and directly connect the pins of the two groups of chips face to face. After omitting the substrate 103, the distance between the chips is shorter and the bandwidth is higher.
[0121] According to a specific embodiment, the first chipset includes a first sub-chip 101, and the second chipset includes at least two second sub-chips 102. As shown in FIG9a, the first sub-chip 101 is chip 1, and the at least two second sub-chips 102 include chip 2, chip 3, chip 4 and chip 5.
[0122] According to another specific embodiment, the first chipset includes at least two first sub-chips 101 , the second chipset includes one second sub-chip 102 , the at least two first sub-chips 101 include chip 2 and chip 3 , and the second sub-chip 102 is chip 1 .
[0123] It should be noted that, for these two specific implementations, the specific processes of the two will be described in detail below in combination with Figures 8a to 8g and Figures 10a to 10f, and will not be repeated here.
[0124] In a preferred embodiment, as shown in FIG5 , the first sub-chip 101 and the second sub-chip 102 are interconnected face-to-face via connecting wires 104. The connecting wires 104 may be conductive wires, and are primarily distributed vertically. The lengths of the connecting wires 104 face-to-face between the first sub-chip 101 and the second sub-chip 102 are similar, and preferably, the difference in length is less than 10%.
[0125] Specifically, referring to Figure 5, the second sub-chip 102 is embedded on the substrate 103 through a flip-chip process, and is located on the opposite side of the pins of the first sub-chip 101, so that the pins of the first sub-chip 101 directly face the pins of the second sub-chip 102; some pins of the first sub-chip 101 and some pins of the second sub-chip 102 correspond one to one, and are directly connected through connecting wires 104; it is easy to know that since the face-to-face interconnection of the two sub-chips is a vertical connection method relative to the substrate, the length of the connecting wire 104 is the shortest achievable straight-line distance between the pins of the two chips on opposite sides, and the length difference between the connecting wires 104 is also distributed in the minimum achievable form.
[0126] The connecting line 104 includes a signal line and a signal reference line. The signal reference line is used to compare changes in signals on the signal line. The voltage of the signal reference line is fixed and does not change over time. The line is shared by all signal lines. The length difference between the signal lines and between the signal lines and the signal reference line is less than 10%.
[0127] Through the chip connection method of the embodiment of the present invention, the length of the connection line between the first sub-chip 101 and the second sub-chip 102 can be less than 1mm. In addition, the pins of the chip of the present invention adopt a vertical chip connection method, which is a two-dimensional distribution of pins relative to the one-dimensional distribution of pins in Figure 4a. The present invention uses face-to-face connection. Similarly, if the number of pins per unit side length of the chip is n, the available area of the I / O directly connected to the chip is the square of n, which greatly increases the number of I / Os between chips under the same line spacing conditions, and the larger the chip area, the more obvious the advantage. Moreover, by reducing the distance between chip pins through face-to-face connection, the length of the connecting wire 104 used for pin connection is reduced (reducing the connecting wire to less than 1 mm), greatly shortening the wiring length, achieving high-speed connection, effectively reducing RC, and thus improving bandwidth.
[0128] The present invention also proposes that, based on the aforementioned chip connection method, non-differential signal transmission can be implemented between the first sub-chip 101 and the second sub-chip 102, and bidirectional signal transmission is adopted between the two chips. Thus, the pins of the first sub-chip 101 and the second sub-chip 102 can provide bidirectional input and output (I / O), doubling the number of channels. Combined with the increased bandwidth due to the reduced wire length, the overall data bandwidth can be increased by more than 10 times. This increase in bandwidth can reduce chip wiring congestion, allowing for the use of larger wire spacing, which in turn reduces signal interference. Although non-differential signal transmission and bidirectional transmission methods will result in some loss of signal quality, the increased bandwidth will far outweigh this loss in chip performance. This high-speed connection can be applied to DRAM and CPU / GPU, or to the connection between RF chips and digital chips in communication systems, significantly improving the overall performance of the chip. Figure 7 is a schematic diagram of an existing differential signal circuit. As shown in Figure 7, existing chips have large chip spacing, long inter-chip connections, and large RC, which can cause high-frequency signal distortion and reduce the signal-to-noise ratio. Therefore, existing chips can only transmit signals in a differential manner. A differential signal circuit is used between pins. Each I / O must be connected to a pin, that is, two RC wires. This requires more wires and a smaller wire distance in wiring. Longer wires can easily cause signal congestion.
[0129] Figure 6 is a schematic diagram of a non-differential signal circuit that can be implemented using the chip connection method of an embodiment of the present invention. As shown in Figure 6, the present invention uses non-differential signals with bidirectional signal transmission for communication. This method benefits from short inter-chip connections, reduced RC, and maintained signal-to-noise ratio of high-frequency signals. Therefore, a single-pin non-differential signal transmission method can achieve sufficient functions.
[0130] Continuing with Figure 6, this circuit requires only one RC element for signal transmission. This reduction significantly reduces the amount of wiring, not only increasing communication bandwidth but also reducing wiring congestion. Therefore, non-differential signals can be used for communication signals, allowing both sides of the circuit to input and output signals. The existing circuit shown in Figure 4 requires two RC elements to achieve the same functionality. The chip proposed in this invention can increase the number of signal channels per unit area, thereby improving signal bandwidth, or reduce the spacing between wires by using fewer pins, thereby reducing interference between signals.
[0131] The present invention also proposes a design method for realizing interconnection of multiple chips, which includes: setting a substrate 103; and setting a first chipset and a second chipset on opposite sides of the substrate 103, respectively, wherein the first chipset includes at least a first sub-chip 101, and the second chipset includes at least a second sub-chip 102; wherein the first sub-chip 101 and the second sub-chip 102 are interconnected face-to-face through the substrate 103.
[0132] According to a specific embodiment, Figures 8a-8g are schematic diagrams of a chip design method according to an embodiment of the present invention, wherein the face-to-face interconnection of two sub-chips is used as an example. As shown in Figure 8a, the entire process includes steps S801-S806, and each step corresponds to Figures 8b-8g:
[0133] Step S801, as shown in FIG8b, first prepare a substrate 103 having connection points on both the upper and lower surfaces; a connection line 104 vertically connecting the upper and lower connection points is provided in the substrate;
[0134] Step S802 , flip-chip-connecting the first sub-chip 101 on the first surface of the substrate 103 , and performing underfilling on the connection interface to protect the solder joints, thereby obtaining the structure of FIG8 c ;
[0135] Step S803 , for the structure of FIG8 c (similar to flipping 180 degrees), flip-chip the second sub-chip 102 on the second surface of the substrate 103 , and perform underfill on the connection interface to protect the solder joints, thereby obtaining the structure of FIG8 d ;
[0136] Step S804, as shown in FIG8e, vertical bonding is performed on the second surface of the substrate in FIG8d to establish peripheral wires, wherein the peripheral wires are used to lead out some pins of the first sub-chip 101;
[0137] Step S805, as shown in FIG8f, based on FIG8e, the second sub-chip 102 and the connecting wires are molded, and the peripheral wires and the first sub-core 101 can also be molded to protect the vertical wires;
[0138] In step S806 , as shown in FIG8 g , the die of FIG8 f is thinned so that the die layer exposes the connection points of the connection wires, and second solder balls 105 are added to the connection points to obtain the final chip.
[0139] The RC of the chip designed using the present application can be reduced by 25 times, and this method can increase the I / O density while maintaining a larger line spacing and reduce interference between channels, thereby significantly improving the overall performance of the chip.
[0140] Compared to the two-chip interconnection example shown in Figures 8a-8g, Figures 9a-9b illustrate a chip structure formed by interconnecting more chips in an embodiment of the present invention, with Figure 9a being a top view and Figure 9b being a side view. As shown in Figures 9a-9b, the chip comprises a substrate 103 and five sub-chips, including chip 1, chip 2, chip 3, chip 4, and chip 5, respectively disposed on the substrate 103. Specifically, chip 1 is located on one side of the substrate 103, while chip 2, chip 3, chip 4, and chip 5 are located on the other side of the substrate 103. Chip 1 is connected to chips 2, chip 3, chip 4, and chip 5 through the substrate 103 through face-to-face pin connections. Chip 1 is flip-chip embedded on one side of the substrate 103.
[0141] The chip 1 corresponds to the pins of chip 2, chip 3, chip 4, and chip 5 respectively, and can be directly connected through a connecting wire 104. It is easy to know that the length of the connecting wire 104 is the shortest straight-line distance between the two pins in one-to-one correspondence, and the spacing between the chip 1 and the chip 2, chip 3, chip 4, and chip 5 is less than 1 mm. Preferably, the length difference between any two connecting wires 104 is less than 10%.
[0142] Compared with the design process for two chips in Figures 8a to 8g, Figures 10a to 10f are flow charts of the design process for more chips in the chip design method according to an embodiment of the present invention. As shown in Figures 10a to 10f, the chip design method includes: as shown in Figure 10a, first preparing a substrate 103, wherein the substrate 103 is provided with a connecting wire 104; flip-chip connecting the chip 2 and the chip 3 on the first surface of the substrate 103, and performing bottom filling to protect the welding point on the connection interface to obtain the structure of Figure 10b; for the structure of Figure 10b (similar to flipping 180 degrees), on the substrate The second surface of the board 103 is flip-chip mounted on chip 1, and the connection interface is underfilled to protect the solder joints, resulting in the structure of Figure 10c. As shown in Figure 10d, vertical wire bonding is performed on the substrate of Figure 10c to establish peripheral connecting wires, which are used to lead out some pins of chip 2 and chip 3. As shown in Figure 10e, based on Figure 10d, the chip 1 and the connecting wires are molded to protect the vertical wires. As shown in Figure 10f, the mold of Figure 10e is thinned so that the mold layer exposes the connection points of the connecting wires, and a second solder ball is added to the connection points to obtain the final chip. This method uses face-to-face connections between multiple chips, so that the connection points between the chips overlap in the vertical direction. The connecting wires 104 are mainly concentrated in the vertical direction. Under the same line spacing conditions, the number of I / Os is greatly increased, the distance between chips is reduced, the line length of the wires is reduced, the bandwidth is increased, and the overall performance of the chip is greatly improved.
[0143] Example 4.
[0144] Figures 11a and 11b are schematic diagrams of a chip according to the present invention, and Figure 12 is a schematic diagram of a pin well according to the present invention. As shown in Figures 11 and 12, the position of the pin well is determined by the design. The chip according to the present invention includes: a base layer 201; a metal layer 202 disposed on the base layer 201, the metal layer 202 comprising at least a surface metal layer 205 and a bottom metal layer 206; a plurality of connection points 203 disposed on the surface metal layer 205; and a pin well 204 disposed on the bottom metal layer 206 and connected to some of the connection points 203 in a one-to-one correspondence. The pin well 204 includes a plurality of pins 207 connected in series longitudinally, and the material of the pins 207 is preferably metal.
[0145] Specifically, the base layer 201 can be made of silicon or other semiconductor materials and is primarily used to construct transistors in the bottom-layer circuit, such as driver gates and receiver gates. The metal layer 202 can be made of different metals, such as copper, aluminum, or titanium, depending on the transistor design and signal transmission requirements. The surface metal 205 is a multilayer structure composed of copper, aluminum, titanium, or other metals. The bottom metal 206 is a multilayer metal structure that matches different transistor designs. The surface metal 205 is connected to the bottom metal 206, which is in turn connected to the base layer 201. The metal layer 202 is used to provide signal transmission, transistor power supply, etc. The surface metal 205 is used to provide connections to other chips, and the bottom metal 206 is used for the design of different gate circuits. The surface metal 205 of the chip is provided with multiple connection points 203, made of copper, nickel, tin, silver, or other metals, for connecting to other chips or interposers.
[0146] The chip is also provided with at least one pin well 204, which is disposed on the bottom metal 206 and is connected to some connection points 203 in a one-to-one correspondence. Each pin well 204 includes a plurality of pins 207 that are sequentially connected longitudinally. The pin wells 204 can be placed in appropriate locations according to specific needs. As shown in FIG11a , the pin well 204 of the chip is disposed in the central portion of the chip. As shown in FIG11b , the pin well 204 of the chip is disposed on one side of the chip. As shown in FIG11a and FIG11b , the wiring of the chip in the pin well 204 is arranged in a vertical direction, such as a downward straight line wiring or a downward staircase distribution. The peripheral wiring of the chip needs to bypass the pin wells on each metal layer, such as a Z-shaped horizontal layer change distribution.
[0147] According to a specific embodiment, the pins 207 are designed with multiple layers of stacked vias. Each metal layer is connected via a via. The connection can be vertical, connecting the bottom metal layer 206 to the surface connection point with the shortest wire distance. This connection can achieve the shortest wire distance and the maximum bandwidth. Alternatively, staggered vertical connections can be used, which will increase the wire length slightly but can reduce pin stress and improve reliability (as shown in Figure 13).
[0148] As shown in Figure 14, chip 1 and chip 2 are vertically connected through the pin well 204. In addition, the specific connection method of the pin well 204 is shown in Figure 15a or Figure 15b, where Figure 15a shows the structure after the two chips are directly connected through two pin wells. This connection method makes the distance between the driving gate and the receiving gate less than 0.1mm. Figure 15b shows that the two chips are first connected through a substrate or an interposer and then through two pin wells 204. The pin well 204 is perpendicular to the surface metal 205, and the signal is transmitted between multiple chips through the pin well 204. If a substrate is used, the total length from the driving gate to the receiving gate can be less than 1mm. The total length of the pins 207 in the same pin well 204 is very close, which ensures a high degree of synchronization during signal transmission in the chip and reduces the signal-to-noise ratio loss caused by the difference between the signal line and the reference line.
[0149] A driver gate or receiver gate is located at the bottom of the pin well 204, as shown in Figure 15a. The driver gate is a circuit that transmits and amplifies signals, while the receiver gate is a circuit that receives and amplifies signals. The driver gate / receiver gate is implemented in the bottom metal layer 206. Connecting wires and metal are provided in the pin well 204 to connect the driver gate / receiver gate to the surface metal layer 205.
[0150] The chip of the present invention can effectively reduce the distance between the driving gate and the receiving gate in the chip by setting the pin well 204, thereby reducing the resistance and capacitance (RC) of the chip pins, which not only improves the bandwidth and speed of signal transmission, but also reduces data signal attenuation, thereby achieving high-fidelity chip communication.
[0151] The present invention also provides a chip design method, as shown in FIG16 , the design method includes:
[0152] Step S1601: setting a base layer 201;
[0153] Step S1602: providing a metal layer 202 on the base layer 201, wherein the metal layer 202 at least includes a surface metal 205 and a bottom metal 206;
[0154] Step S1603: setting a plurality of connection points 203 on the surface metal 205;
[0155] Step S1604: Pin wells 204 are provided on the bottom metal layer 206, perpendicular to the surface metal layer 205 and connected to some of the connection points 203 in a one-to-one correspondence. The pin wells 204 include a plurality of longitudinally connected pins 207. The specific fabrication process includes dielectric layer photolithography, sputtering, and resist coating; metal layer photolithography, electroplating, resist stripping, and chemical mechanical polishing.
[0156] The chip of the present invention includes: a base layer 201; a metal layer 202 disposed on the base layer 201, the metal layer 202 comprising at least a surface metal layer 205 and a bottom metal layer 206; a plurality of connection points 203 disposed on the surface metal layer 205; and pin wells 204 disposed on the bottom metal layer 206 and connected to some of the connection points 203 in a one-to-one correspondence. The pin wells 204 include a plurality of pins 207 connected in series longitudinally. By providing the pin wells 204, the present invention provides a high-speed signal transmission channel, effectively reducing the distance between pins within the chip, reducing the chip area, and increasing the chip's integration density.
[0157] The above describes in detail the optional implementation methods of the embodiments of the present invention in conjunction with the accompanying drawings. However, the embodiments of the present invention are not limited to the specific details in the above implementation methods. Within the technical concept of the embodiments of the present invention, various simple modifications can be made to the technical solutions of the embodiments of the present invention, and these simple modifications all fall within the scope of protection of the embodiments of the present invention.
[0158] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any appropriate manner without contradiction. To avoid unnecessary repetition, the embodiments of the present invention will not further describe various possible combinations.
[0159] It should also be noted that the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, method, commodity, or apparatus that includes a series of elements includes not only those elements but also other elements not explicitly listed, or includes elements inherent to such process, method, commodity, or apparatus. In the absence of further limitations, an element defined by the phrase "comprises a ..." does not exclude the presence of other identical elements in the process, method, commodity, or apparatus that includes the element.
[0160] The above are merely embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application may have various changes and variations. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principles of the present application should all be included within the scope of the claims of the present application.
Claims
1. A chip packaging method, characterized in that: It includes the following steps which are performed in sequence: An initial chip structure is prepared, the initial chip structure comprising a wire redistribution layer (RDL) (400), a chip layer and a bump array layer (500) stacked in sequence, wherein the chip layer comprises at least one first chip unit, the first chip unit comprises a first bare chip (110) and a connection structure (120) distributed on both sides of the first bare chip (110) to define a space capable of accommodating the first bare chip (110), the connection structure (120) being capable of realizing signal transmission between the RDL (400) and the bump array layer (500) disposed on the upper and lower surfaces thereof; Providing at least one second chip unit, wherein each second chip unit corresponds to one first chip unit, and each second chip unit includes a second bare chip (600); Flip-chipping the second bare chip (600) in the second chip unit to the first bare chip (110) in the corresponding first chip unit and the bump array layer (500) on the connection structure (120) in a face-to-face direct interconnection manner, so that the wires between the corresponding connection points are as short as possible; Cutting out each corresponding group of first chip units and second chip units to form a single chip structure; For the single chip structure, a heat sink (700) is mounted on the back side of the second bare chip (600); as well as Preparing a first solder ball (800) on the RDL (400) to obtain a final chip packaging structure; The second bare chip (600) in the second chip unit is flip-chip mounted to the first bare chip (110) in the corresponding first chip unit and the bump array layer (500) on the connection structure (120) in a face-to-face direct interconnection manner, comprising: The second bare chip (600) overlaps with the first bare chip (110) in a vertical direction, the second bare chip (600) and the first bare chip (110) contact opposite surfaces of the RDL (400), and are directly connected to the first bare chip (110) without passing through other RDLs after being welded by vertical wires; The second bare chip (600) overlaps with the first bare chip (110) in a vertical direction, and the second bare chip (600) is directly connected to the RDL (400) through a connection structure (120) without passing through other RDLs.
2. The chip packaging method according to claim 1, characterized in that: The preparation of the initial chip structure comprises: Providing the first bare chip (110) and the connection structure (120) to form the first chip unit; Place the first bare chip (110) and the connection structure (120) with their front connection points facing downward on the upper surface of the carrier (200); Performing compression molding on the first bare chip (110) and the connection structure (120); The RDL (400) is prepared on the back side of the first bare chip (110) and the connection structure (120), and a substrate removal process is performed on the front side; Performing flip chip thinning on the first chip unit after the carrier is removed to expose the connection points; and The bump array layer (500) is prepared on the exposed connection points.
3. The chip packaging method according to claim 1, characterized in that: The connection structure (120) is a chip with through silicon vias.
4. The chip packaging method according to claim 1, characterized in that: The chip packaging method further comprises: After the flip chip is completed and before the dicing, the first chip unit and the second chip unit are subjected to underfilling or molding.
5. The chip packaging method according to claim 1, characterized in that: The chip packaging method also includes: An external device is electrically connected via the first solder ball (800) to achieve signal transmission between the first bare chip (110) and the external device based on the first solder ball (800), and signal transmission between the second bare chip (600) and the external device based on the first solder ball (800) and the connection structure (120).
6. A chip packaging structure implemented by the chip packaging method according to any one of claims 1 to 5, characterized in that: include: The initial chip structure includes a wire redistribution layer RDL (400), a chip layer and a convex A bump array layer (500), wherein the chip layer comprises at least one first chip unit, the first chip unit comprising a first bare chip (110) and a connection structure (120) distributed on both sides of the first bare chip (110) to define a space capable of accommodating the first bare chip (110), the connection structure (120) being capable of realizing signal transmission between the RDL (400) disposed on the upper and lower surfaces thereof and the bump array layer (500); At least one second chip unit, wherein each second chip unit corresponds to a first chip unit, and each second chip unit comprises a second bare chip (600), wherein the second bare chip (600) in the second chip unit is flip-chip mounted to the first bare chip (110) in the corresponding first chip unit and the bump array layer on the connection structure (120) in a face-to-face direct interconnection manner, wherein the face-to-face direct interconnection makes the wires between the corresponding connection points as short as possible; For a single chip structure including a corresponding set of first chip units and second chip units cut out after the flip chip is completed, a heat sink (700) is mounted on the back side of the second bare chip (600); and A first solder ball (800) is fabricated on the RDL (400); The second bare chip (600) in the second chip unit is flip-chip mounted to the first bare chip (110) in the corresponding first chip unit and the bump array layer (500) on the connection structure (120) in a face-to-face direct interconnection manner, comprising: The second bare chip (600) overlaps with the first bare chip (110) in a vertical direction, the second bare chip (600) and the first bare chip (110) contact opposite surfaces of the RDL (400), and are directly connected to the first bare chip (110) without passing through other RDLs after being welded by vertical wires; The second bare chip (600) overlaps with the first bare chip (110) in a vertical direction, and the second bare chip (600) is directly connected to the RDL (400) through a connection structure (120) without passing through other RDLs.
7. The chip packaging structure according to claim 6, characterized in that: In the initial chip structure: The RDL (400) is located on the backside of the first bare chip (110) and the connection structure (120), and the front sides of the first bare chip (110) and the connection structure (120) expose connection points for connecting the bump array layer (500).
8. The chip packaging structure according to claim 6, characterized in that: The first solder ball (800) is electrically connected to an external device to achieve signal transmission between the first bare chip (110) and the external device based on the first solder ball (800), and to achieve signal transmission between the second bare chip (600) and the external device based on the first solder ball (800) and the connection structure (120).
9. A chip, characterized in that: The chip includes: a substrate (103); and A first chip group and a second chip group respectively arranged on opposite surfaces of the substrate (103), the first chip group comprising at least a first sub-chip (101), and the second chip group comprising at least a second sub-chip (102); The first sub-chip (101) and the second sub-chip (102) are interconnected face-to-face through pins through the substrate (103); The substrate (103) is provided with a connecting line (104); The first sub-chip (101) and the second sub-chip (102) are interconnected by pins face to face via a connecting wire (104); The length of the connection line between the first sub-chip (101) and the second sub-chip (102) is less than 1 mm, and the length difference between any two connection lines (104) is less than 10%; The first sub-chip (101) and the second sub-chip (102) implement bidirectional signal transmission between the first sub-chip (101) and the second sub-chip (102) through a non-differential signal circuit; The non-differential signal circuit includes: a unique RC, one end of which is simultaneously connected to the outputs of the first group of Tx channels and the first group of DTx channels, and the outputs of the first group of DRx and the first group of Rx; the other end of the unique RC is simultaneously connected to the outputs of the second group of Tx channels and the second group of DTx channels, and the outputs of the second group of DRx and the second group of Rx.
10. The chip according to claim 9, characterized in that: The first chipset includes a first sub-chip (101), and the second chipset includes at least two second sub-chips (102).
11. The chip according to claim 9, characterized in that: The first chipset includes at least two first sub-chips (101), and the second chipset includes one second sub-chip (102).
12. The chip according to claim 9, characterized in that: The chip also includes: a second solder ball (105) and a peripheral wire; The peripheral wire is connected to the second sub-chip (102); The second solder ball (105) is used to lead out peripheral wires.
13. The chip according to claim 9, characterized in that: The first sub-chip (101) and the second sub-chip (102) are radio frequency chips or digital chips.
14. A design method for interconnecting multiple chips, characterized in that: The design approach includes: Providing a substrate (103); and A first chip group and a second chip group are respectively arranged on opposite surfaces of the substrate (103), wherein the first chip group includes at least a first sub-chip (101), and the second chip group includes at least a second sub-chip (102); The first sub-chip (101) and the second sub-chip (102) are interconnected face-to-face through pins through the substrate (103); A connecting line (104) is provided in the substrate (103); The first sub-chip (101) and the second sub-chip (102) are interconnected by pins face to face via a connecting wire (104); The length of the connection line between the first sub-chip (101) and the second sub-chip (102) is less than 1 mm, and the length difference between any two connection lines (104) is less than 10%; The first sub-chip (101) and the second sub-chip (102) implement bidirectional signal transmission between the first sub-chip (101) and the second sub-chip (102) through a non-differential signal circuit; The non-differential signal circuit includes: a unique RC, one end of which is simultaneously connected to the outputs of the first group of Tx channels and the first group of DTx channels, and the outputs of the first group of DRx and the first group of Rx; the other end of the unique RC is simultaneously connected to the outputs of the second group of Tx channels and the second group of DTx channels, and the outputs of the second group of DRx and the second group of Rx.
15. The design method according to claim 14, characterized in that: The first sub-chip (101) and the second sub-chip (102) are interconnected face-to-face through pins of the substrate (103), comprising: Arranging a first sub-chip (101) on a first surface of the substrate (103); The second sub-chip (102) is arranged on a second surface of the substrate (103) opposite to the first surface in a flip-chip manner.
16. The design method according to claim 14, characterized in that: The design method further comprises: After completing the face-to-face pin interconnection, the pins are underfilled.
17. The design method according to claim 14, characterized in that: The design method further comprises: Performing vertical bonding on the second surface of the substrate (103) to prepare peripheral wires of the second sub-chip (102); Performing compression molding on the peripheral wires and the first sub-chip (101); A second solder ball (105) is provided to lead out a peripheral wire.
18. A chip, characterized in that: The chip includes: Basal layer (201); A metal layer (202) disposed on the base layer (201), the metal layer (202) comprising at least a surface metal (205) and a bottom metal (206); A plurality of connection points (203) disposed on the surface metal (205); and A pin well (204) is disposed on the bottom metal (206) and is connected to some of the connection points (203) in a one-to-one correspondence and is used to provide a signal transmission channel, and the pin well (204) includes a plurality of pins (207) that are connected in sequence longitudinally; The pin well (204) is provided with a driving gate circuit or a receiving gate circuit; The pin well (204) is provided with a connecting wire and metal for connecting the surface metal (205) with a driving gate circuit or a receiving gate circuit; The chip comprises a plurality of pin wells (204), and the total lengths of the pins (207) in the pin wells (204) are close, so that chip signal transmission is synchronized.
19. The chip according to claim 18, characterized in that: The tube pin (207) is a perforated structure; The pin well (204) is connected to the surface metal (205) and the bottom metal (206); The tube pin (207) is designed with multi-layer stacked holes; The connection of the pin (207) is vertically connected to connect the bottom metal (206) and the surface connection point with the shortest wire distance. or, The connection of the pins (207) is staggered and vertically connected.
20. The chip according to claim 18, characterized in that The length difference between any two pins (207) in the pin well (204) is less than 10%.
21. The chip according to claim 18, characterized in that The pin well (204) is perpendicular to the surface metal (205).
22. A chip design method, characterized in that: The design approach includes: Setting a base layer (201); A metal layer (202) is provided on the base layer (201), wherein the metal layer (202) comprises at least a surface metal (205) and a bottom metal (206); and Arranging a plurality of connection points (203) on the surface metal (205); The bottom metal (206) is provided with a pin well (204) which is connected to some of the connection points (203) in a one-to-one correspondence and is used to provide a signal transmission channel, and the pin well (204) includes a plurality of pins (207) connected in sequence longitudinally; The pin well (204) is provided with a driving gate circuit or a receiving gate circuit; The pin well (204) is provided with a connecting wire and metal for connecting the surface metal (205) with a driving gate circuit or a receiving gate circuit; The chip comprises a plurality of pin wells (204), and the total lengths of the pins (207) in the pin wells (204) are close, so that chip signal transmission is synchronized.
23. The design method according to claim 22, characterized in that: The tube pin (207) is a perforated structure; The pin well (204) is connected to the surface metal (205) and the bottom metal (206); The tube pin (207) is designed with multi-layer stacked holes; The connection of the pin (207) is vertically connected to connect the bottom metal (206) and the surface connection point with the shortest wire distance. or, The connection of the pins (207) is staggered and vertically connected.
24. The design method according to claim 22, characterized in that: The length difference between any two pins (207) in the pin well (204) is less than 10%.
25. The design method according to claim 22, characterized in that: The pin well (204) is perpendicular to the surface metal (205).
