Submersible water-cooled server
The modular design of the immersion water-cooled server solves the problems of complex maintenance and inflexible configuration in existing technologies, enabling rapid plug-and-play and efficient maintenance, and improving the server's configuration flexibility and scalability.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Utility models(China)
- Current Assignee / Owner
- 联想长风科技(北京)有限公司
- Filing Date
- 2025-07-17
- Publication Date
- 2026-07-17
AI Technical Summary
Existing immersion liquid-cooled server hardware components are packaged in the same liquid-cooling cavity, resulting in complex maintenance, high operating threshold, and limiting the flexibility of GPU configuration and the scalability of the server.
The server adopts a modular design, dividing it into a main control module, a GPU acceleration module, and a power supply module. Each module has a cold water chamber, a heat exchange chamber, and a hot water chamber, which are separated by perforated partitions. Coolant flows through each chamber in sequence, enabling quick plug-and-play maintenance and flexible configuration.
It improves maintenance efficiency, reduces operation and maintenance costs, and enhances the configuration flexibility and scalability of servers.
Smart Images
Figure CN224519249U_ABST
Abstract
Description
Technical Field
[0001] This utility model relates to the field of water-cooled server technology, and in particular to immersion water-cooled servers. Background Technology
[0002] With the continuous growth of computing power demands in data centers, traditional air-cooled servers, due to their low heat exchange efficiency, limited computing power density per unit volume, severe noise pollution, and weak environmental protection capabilities, are no longer able to meet the heat dissipation requirements of high-power-density devices. Although hybrid air-liquid cooling solutions improve the heat dissipation capacity of core components through localized water-cooled plates, they still suffer from drawbacks such as fan noise, complex airflow organization, and the risk of liquid leakage. Against this backdrop, immersion liquid cooling technology, with its highly efficient heat exchange, quiet operation, and strong environmental protection characteristics brought about by its fully enclosed liquid cooling cavity design, is gradually becoming an important solution for high-density computing scenarios. However, existing immersion liquid-cooled servers encapsulate all hardware components within the same liquid cooling cavity, resulting in a complex process of draining, disassembling, reassembling, and refilling the entire machine when replacing critical components (such as GPU accelerator cards). This not only leads to long maintenance times and high operational barriers but also limits the flexibility of dynamically adjusting GPU configurations according to different application scenarios. This rigid constraint in structural design increases maintenance costs and restricts the scalability of the server architecture. Summary of the Invention
[0003] To address the shortcomings of existing technologies, this utility model provides an immersion water-cooled server, comprising: a chassis with an inlet and an outlet on the rear panel serving as a unified water inlet and outlet channel; a main control module, which includes a main control cold water chamber, a main control heat exchange chamber, and a main control hot water chamber, wherein the main control cold water chamber, the main control heat exchange chamber, and the main control hot water chamber are separated by perforated partitions, and cold water entering through the inlet flows sequentially through the main control cold water chamber, the main control heat exchange chamber, and the main control hot water chamber, carrying away heat from the components before being discharged through the outlet; and a GPU acceleration module group, which includes an acceleration cold water chamber, an acceleration heat exchange chamber, and an acceleration hot water chamber, wherein the acceleration cold water chamber, the acceleration heat exchange chamber, and the acceleration hot water chamber are separated by perforated partitions, and cold water entering through the inlet flows sequentially through the acceleration cold water chamber, the acceleration heat exchange chamber, and the acceleration hot water chamber, carrying away heat from the components before being discharged through the outlet. This achieves the technical effect of enabling rapid plug-and-play maintenance through modular decoupling design, thereby improving configuration flexibility and maintenance efficiency.
[0004] In one feasible design, the server further includes a power module assembly that is inserted from the rear panel into a corresponding slot in the chassis.
[0005] In one feasible design, a cold water distributor is provided at the rear of the chassis to divide the cold water coming in from the inlet into water channels with a preset number of modules, wherein the preset number of modules is the sum of the number of modules in the main control module, the GPU acceleration module group, and the power supply module group.
[0006] In one feasible design, the cold water distributor divides the cold water coming in from the inlet into water channels of the preset number of modules, and delivers them to the inlets of the main control module, the GPU acceleration module group, and the power module group respectively through water inlet pipes.
[0007] In one feasible design, the hot water output from the outlets of the main control module, the GPU acceleration module group, and the power module group is respectively transported through water cooling pipes to the hot water distributor at the rear of the chassis, and then discharged through the outlet on the rear panel of the chassis.
[0008] In one feasible design, the main control module and the GPU acceleration module group are inserted from the front panel into the corresponding slots of the chassis.
[0009] In one feasible design, both the main control module and the GPU acceleration module group have pull-out aids on their front panels for users to insert or remove the modules from the chassis.
[0010] In one feasible design, both the main control module and the GPU acceleration module group have locking screws on their front panels, which can fully lock the modules inserted into the chassis to the chassis.
[0011] In one feasible design, the GPU acceleration module group has a PCIe card slot, and the PCIe card slot is equipped with a GPU acceleration card.
[0012] In one feasible design, the front panel of the GPU acceleration module group is equipped with slot indicator lights to indicate the presence status of the internal PCIe card slots.
[0013] The above description is merely an overview of the technical solution of this utility model. In order to better understand the technical means of this utility model and to implement it in accordance with the contents of the specification, and to make the above and other objects, features and advantages of this utility model more obvious and understandable, specific embodiments of this utility model are given below. Attached Figure Description
[0014] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings:
[0015] Figure 1 This is a schematic diagram of the overall appearance of the immersion water-cooled server provided by this utility model.
[0016] Figure 2 The front view of the immersion water-cooled server provided by this utility model.
[0017] Figure 3 The rear view of the immersion water-cooled server provided by this utility model.
[0018] Figure 4 A schematic diagram of the main control module of the immersion water-cooled server provided by this utility model.
[0019] Figure 5 A stacked diagram of the GPU acceleration module of the immersion water-cooled server provided by this utility model.
[0020] Figure 6 This is a cross-sectional view of the immersion water-cooled server provided by this utility model.
[0021] Figure labeling: GPU acceleration module group 11, main control module 12, front panel 13, chassis 14, rear panel 15, locking screw 21, pull-out aid 22, slot indicator light 23, water inlet 31, water outlet 32, main control cold water chamber 41, main control heat exchange chamber 42, main control hot water chamber 43, perforated partition 44, accelerated cold water chamber 51, accelerated heat exchange chamber 52, accelerated hot water chamber 53, power module group 61, cold water separator 62. Detailed Implementation
[0022] To make the objectives, technical solutions, and advantages of the embodiments of this utility model clearer, the technical solutions of the embodiments of this utility model will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this utility model, not all embodiments. Based on the embodiments of this utility model, all other embodiments obtained by those skilled in the art without creative effort are within the protection scope of this utility model.
[0023] In the description of the embodiments of this utility model, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this utility model, "multiple" means at least two, such as two, three, etc., unless otherwise explicitly specified. In this utility model, unless otherwise explicitly specified and limited, the terms "installed," "connected," "joined," "fixed," etc., should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral part; they can refer to a direct connection or an indirect connection through an intermediate medium; they can refer to the internal communication of two elements or the interaction between two elements, unless otherwise explicitly specified. Those skilled in the art can understand the specific meaning of the above terms in this utility model according to the specific circumstances.
[0024] In this utility model, unless otherwise explicitly specified and limited, "above" or "below" the second feature can mean that the first feature is in direct contact with the second feature, or that the first feature is in indirect contact with the second feature through an intermediate medium. Furthermore, "above," "on top of," and "over" the second feature can mean that the first feature is directly above or diagonally above the second feature, or simply that the first feature is at a higher horizontal level than the second feature. "Below," "below," and "under" the second feature can mean that the first feature is directly below or diagonally below the second feature, or simply that the first feature is at a lower horizontal level than the second feature.
[0025] In the description of this utility model, it should be understood that the terms "inner," "outer," "upper," "bottom," "front," and "rear," etc., indicate the orientation or positional relationship (if any) based on the appendix. Figure 1 The orientations or positional relationships shown are for the convenience of describing this utility model and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this utility model.
[0026] Examples, such as Figure 1 , Figure 2 , Figure 3 , Figure 4 , Figure 5 , Figure 6 As shown, this utility model provides an immersion water-cooled server, comprising:
[0027] The chassis 14 has a rear panel 15 with an inlet 31 and an outlet 32, serving as a unified water inlet and outlet channel. The main control module 12 includes a main control cold water chamber 41, a main control heat exchange chamber 42, and a main control hot water chamber 43. These chambers are separated by a perforated partition 44. Cold water entering through the inlet 31 flows sequentially through these chambers. After carrying away the heat from the components, the water is discharged through the outlet 32; GPU acceleration module group 11, the GPU acceleration module group 11 includes an accelerated cold water chamber 51, an accelerated heat exchange chamber 52 and an accelerated hot water chamber 53, wherein the accelerated cold water chamber 51, the accelerated heat exchange chamber 52 and the accelerated hot water chamber 53 are separated by an open partition 44, and the cold water entering from the inlet 31 flows through the accelerated cold water chamber 51, the accelerated heat exchange chamber 52 and the accelerated hot water chamber 53 in sequence, and after carrying away the heat from the components, it is discharged through the outlet 32.
[0028] In this utility model, the immersion water-cooled server adopts a modular design. The complete immersion water-cooled server consists of a chassis 14, a main control module 12, a GPU acceleration module, and a power supply module 61.
[0029] For the chassis 14 of the immersion water-cooled server, the rear panel 15 is provided with a water inlet 31 and a water outlet 32. The water inlet 31 and the water outlet 32 form a unified water inlet and outlet channel to ensure that the coolant can effectively enter and flow out of the chassis 14, thus ensuring the efficient heat dissipation function of the immersion water-cooled server.
[0030] The main control module 12 has three functional areas: a main control cold water chamber 41, a main control heat exchange chamber 42, and a main control hot water chamber 43. These three chambers are physically isolated by a perforated partition 44. Coolant (such as cold water) enters through the inlet 31 on the rear panel 15 of the chassis 14, first flowing through the inlet pipe into the main control cold water chamber 41, which is used to cool the main electronic components (such as RAID cards). After passing through the main control cold water chamber 41, the coolant carries away the heat from the components, then passes through the perforated partition 44 into the main control heat exchange chamber 42, where the coolant further absorbs heat from core components (such as the CPU and memory). Finally, the coolant flows through the perforated partition 44 into the main control hot water chamber 43, carrying away the remaining heat, and then exits through the outlet 32, ensuring that the coolant can circulate in different chambers and efficiently carry away heat, thereby maintaining the stable operating temperature of the server.
[0031] The GPU acceleration module group 11 is also equipped with an accelerated cooling water chamber 51, an accelerated heat exchange chamber 52, and an accelerated hot water chamber 53, similar to the main control cooling water chamber 41, the main control heat exchange chamber 42, and the main control hot water chamber 43. The three chambers are also separated by a perforated partition 44 to ensure that the liquid flows through different chambers and fully utilizes the heat exchange efficiency. After the coolant enters from the inlet 31, it first flows through the accelerated cooling water chamber 51 to cool the core components such as the high-performance accelerator card (such as the GPU), then enters the accelerated heat exchange chamber 52 to remove the heat, and finally completes the final heat exchange through the accelerated hot water chamber 53 before being discharged through the outlet 32.
[0032] This design isolates different cooling chambers and effectively guides the coolant flow path using perforated baffles 44, maximizing heat exchange efficiency within each module while ensuring the overall thermal management capabilities of the server. The coolant flow from inlet 31 to outlet 32 effectively reduces the overall server operating temperature, ensuring efficient server operation. Furthermore, the modular design makes hardware maintenance and replacement more flexible; users can easily update or replace the GPU acceleration module and main control module 12 without complex disassembly and coolant replacement, improving maintenance efficiency.
[0033] Furthermore, the server also includes a power module assembly 61, which is inserted from the rear panel 15 into the corresponding slot of the chassis 14.
[0034] In this invention, the immersion water-cooled server also includes a power module assembly 61, which is inserted into a corresponding slot in the chassis 14 via the rear panel 15. The design of the power module assembly 61 makes the power supply of the entire server more modular and flexible, allowing users to easily plug and unplug power modules for maintenance, replacement, or upgrades. In this design, the connection between the power module assembly 61 and the chassis 14 is achieved through precisely matched slots, which not only ensures the stable installation of the power module but also avoids power supply loosening or poor contact problems caused by external impacts or improper operation.
[0035] Furthermore, a cold water distributor 62 is provided at the rear of the chassis 14 to divide the cold water coming in from the water inlet 31 into water channels with a preset number of modules, wherein the preset number of modules is the sum of the number of modules in the main control module 12, the GPU acceleration module group 11 and the power module group 61.
[0036] In this invention, a cooling water distributor 62 is provided at the rear of the chassis 14 of the immersion water-cooled server. The main function of the cooling water distributor 62 is to distribute the cooled water entering through the inlet 31 into multiple water paths to meet the heat dissipation needs of different modules. The cooling water distributor 62 divides the cooled water into multiple independent water paths according to a preset number of modules, ensuring that each module receives the required coolant flow. The preset number of modules includes the total number of modules in the main control module 12, the GPU acceleration module group 11, and the power supply module group 61. Typically, the preset number of modules is 5, i.e., 1 main control module 12, 2 GPU acceleration modules, and 2 power supply modules. This design ensures efficient and stable heat dissipation, thereby improving the overall server's operating efficiency and maintenance convenience.
[0037] Furthermore, the cold water distributor 62 divides the cold water entering from the inlet 31 into water channels of the preset number of modules, and delivers them through inlet pipes to the inlets 31 of the main control module 12, the GPU acceleration module group 11, and the power module group 61, respectively.
[0038] In this invention, the immersion water-cooled server is equipped with a coolant distributor 62. The function of the coolant distributor 62 is to distribute the coolant entering through the inlet 31 according to the preset number of modules. That is, the coolant distributor 62 divides the coolant into multiple independent water paths according to the number of modules in the main control module 12, GPU acceleration module group 11, and power module group 61. Each water path delivers the coolant to the inlet 31 of each module through an inlet pipe. This distribution method ensures that each module can receive coolant independently, meeting its heat dissipation requirements, thereby effectively improving the overall heat dissipation efficiency of the server.
[0039] Furthermore, the hot water output from the outlets 32 of the main control module 12, the GPU acceleration module group 11, and the power module group 61 is respectively transported through water cooling pipes to the hot water distributor at the rear of the chassis 14, and then discharged through the outlet 32 on the rear panel 15 of the chassis 14.
[0040] In this invention, the outlets 32 of the main control module 12, GPU acceleration module group 11, and power module group 61 respectively output hot water after the heat has been removed by the coolant. The hot water from each module is transported to the hot water distributor at the rear of the chassis 14 via connected water-cooling pipes. The function of the hot water distributor is to rationally distribute the hot water from each module for effective drainage. In the hot water distributor, the hot water is collected and distributed to appropriate water pipes, and then discharged through the outlet 32 on the rear panel 15 of the chassis 14. This drainage path ensures that the hot water can leave the server smoothly and quickly, preventing heat accumulation inside the chassis 14 and affecting the overall heat dissipation performance of the server, thereby improving the maintainability and long-term stability of the server.
[0041] Furthermore, the main control module 12 and the GPU acceleration module group 11 are inserted from the front panel 13 into the corresponding slots of the chassis 14.
[0042] In this invention, both the main control module 12 and the GPU acceleration module group 11 are equipped with a front panel 13. The front panel 13 has interfaces and connection devices for interfacing with slots within the chassis 14. During installation, the user simply aligns the main control module 12 and the GPU acceleration module group 11 with the slots within the chassis 14 via the front panel 13 and inserts them into the corresponding slots until they are fully seated. This insert-type design of the front panel 13 ensures a secure connection between the modules and the chassis 14, while also facilitating quick disassembly or replacement of modules as needed. This modular design enhances the server's scalability and flexibility, enabling rapid adjustment and configuration of different hardware modules according to actual requirements.
[0043] Furthermore, both the main control module 12 and the GPU acceleration module group 11 are equipped with pull-out aids 22 on their front panels 13, for users to insert or remove the modules from the chassis 14.
[0044] In this invention, both the main control module 12 and the GPU acceleration module group 11 are equipped with pull-out aids 22 on their front panels 13, designed to facilitate easy insertion and removal of modules from the chassis 14 during installation, maintenance, or replacement. The pull-out aid 22 is a convenient mechanical device, typically located at the edge of the front panel 13, designed to provide additional gripping force, helping users operate the modules more easily. When a user needs to insert a module into the chassis 14, the pull-out aid 22 provides additional support, allowing the module to be smoothly aligned and inserted into the slot of the chassis 14. When removing a module, the design of the pull-out aid 22 allows the user to easily remove the module from the slot, avoiding the difficulties or risks of damage that may occur during manual insertion and removal. By providing the pull-out aid 22, the convenience and reliability of hardware replacement are improved, enhancing server maintenance efficiency.
[0045] Furthermore, both the main control module 12 and the GPU acceleration module group 11 are provided with locking screws 21 on their front panels, which can fully lock the modules inserted into the chassis 14 to the chassis 14.
[0046] In this invention, both the main control module 12 and the GPU acceleration module group 11 are equipped with locking screws 21 on their front panels 13. These screws ensure that the modules are securely fixed in their slots after being inserted into the chassis 14, preventing loosening or detachment during operation or transportation. The locking screws 21 are located on the front panel 13. When installing the modules, users simply insert the modules into the corresponding slots of the chassis 14 and tighten the screws to firmly secure the modules within the chassis 14. This enhances the connection stability between the modules and the chassis 14, ensuring that the modules will not loosen due to vibration or external impact during server operation, thus improving the overall safety and reliability of the server.
[0047] Furthermore, the GPU acceleration module group 11 has a PCIe card slot, and the PCIe card slot is equipped with a GPU acceleration card.
[0048] In this invention, the GPU acceleration module group 11 is equipped with multiple PCIe card slots for installing GPU accelerator cards. Each PCIe card slot provides a stable connection interface for the GPU accelerator card, ensuring that the GPU accelerator card can effectively transmit data and interact with the main control module 12 and other components. The GPU accelerator card is a hardware module specifically designed for high-performance computing, capable of significantly improving the processing power of computationally intensive tasks (such as data processing, deep learning, and graphics rendering). Furthermore, the GPU accelerator card installation method is modular, allowing users to flexibly configure different numbers of accelerator cards according to actual needs. When increased computing power is required, users can easily increase computing resources by plugging and unplugging GPU accelerator cards from the PCIe card slots without the need for complex server disassembly or rewiring. Through this design, the GPU acceleration module can not only adapt to different application scenarios but also improve overall computing performance without adding extra complexity.
[0049] Furthermore, the front panel 13 of the GPU acceleration module group 11 is provided with a slot indicator light 23 to indicate the position status of the internal PCIe card slot.
[0050] In this invention, the front panel 13 of the GPU acceleration module group 11 is equipped with slot indicator lights 23 to display the real-time position status of the internal PCIe card slots. Each indicator light corresponds one-to-one with a corresponding PCIe card slot, clearly reflecting the installation status of the GPU accelerator card in each slot. When the GPU accelerator card is correctly inserted and connected to the PCIe card slot, the corresponding slot indicator light 23 will light up, indicating that the slot has been successfully installed and is in normal working condition. If the GPU accelerator card is not correctly inserted, or if no GPU accelerator card is installed in the slot, the slot indicator light 23 will not light up, or will display a warning color, prompting the user that there is a potential problem or maintenance is required. This design not only provides users with intuitive feedback to ensure correct hardware installation, but also facilitates the rapid detection of any possible hardware failures or connection problems during operation.
[0051] In summary, the chassis 14 has a water inlet 31 and a water outlet 32 arranged on its rear panel 15, serving as a unified water inlet and outlet channel. The main control module 12 includes a main control cold water chamber 41, a main control heat exchange chamber 42, and a main control hot water chamber 43. These chambers are separated by a perforated partition 44. Cold water entering through the water inlet 31 flows sequentially through these chambers. 3. After carrying away heat from the components, the water is discharged through the outlet 32; GPU acceleration module group 11, the GPU acceleration module group 11 includes an accelerated cold water chamber 51, an accelerated heat exchange chamber 52, and an accelerated hot water chamber 53, wherein the accelerated cold water chamber 51, the accelerated heat exchange chamber 52, and the accelerated hot water chamber 53 are separated by a perforated partition 44. Cold water entering through the inlet 31 flows sequentially through the accelerated cold water chamber 51, the accelerated heat exchange chamber 52, and the accelerated hot water chamber 53, carrying away heat from the components and then being discharged through the outlet 32. This achieves the technical effect of enabling quick plug-and-play maintenance through modular decoupling design, improving configuration flexibility and maintenance efficiency.
[0052] Although the present invention has been described in conjunction with specific features and embodiments, it is apparent that various modifications and combinations can be made thereto without departing from the spirit and scope of the present invention. Accordingly, this specification and accompanying drawings are merely illustrative examples of the present invention as defined herein, and are to be considered as covering any and all modifications, variations, combinations, or equivalents within the scope of the present invention. Clearly, those skilled in the art can make various alterations and modifications to the present invention without departing from its scope. Thus, if such modifications and modifications of the present invention fall within the scope of the present invention and its equivalents, the present invention intends to include such modifications and modifications.
Claims
1. An immersion water-cooled server characterized by, The server includes: The chassis has an inlet and an outlet on its rear panel, which serve as a unified water inlet and outlet channel for the whole system. The main control module includes a main control cold water chamber, a main control heat exchange chamber, and a main control hot water chamber. The main control cold water chamber, the main control heat exchange chamber, and the main control hot water chamber are separated by a perforated partition. Cold water entering through the inlet flows sequentially through the main control cold water chamber, the main control heat exchange chamber, and the main control hot water chamber, carrying away the heat from the components before being discharged through the outlet. The GPU acceleration module group includes an accelerated cold water chamber, an accelerated heat exchange chamber, and an accelerated hot water chamber. The accelerated cold water chamber, the accelerated heat exchange chamber, and the accelerated hot water chamber are separated by a perforated partition. Cold water entering through the inlet flows sequentially through the accelerated cold water chamber, the accelerated heat exchange chamber, and the accelerated hot water chamber, carrying away the heat of the components before being discharged through the outlet.
2. The submerged water-cooled server of claim 1, wherein, The server also includes a power module assembly that is inserted from the rear panel into a corresponding slot in the chassis.
3. The submerged water-cooled server of claim 2, wherein, The rear of the chassis is equipped with a cold water separator, which is used to divide the cold water coming in from the inlet into water channels with a preset number of modules, wherein the preset number of modules is the sum of the number of modules in the main control module, the GPU acceleration module group and the power module group.
4. The submerged water-cooled server of claim 3, wherein, The cold water separator divides the cold water coming in from the inlet into water channels of the preset number of modules, and delivers them to the inlets of the main control module, the GPU acceleration module group, and the power module group respectively through the water inlet pipe.
5. The submerged water-cooled server of claim 4, wherein, The hot water output from the outlets of the main control module, the GPU acceleration module group, and the power module group is transported through water cooling pipes to the hot water distributor at the rear of the chassis, and then discharged through the outlet on the rear panel of the chassis.
6. The submerged water-cooled server of claim 1, wherein, The main control module and the GPU acceleration module are inserted into the corresponding slots of the chassis from the front panel.
7. The submerged water-cooled server of claim 1, wherein, Both the main control module and the GPU acceleration module group have pull-out aids on their front panels for users to insert or remove the modules from the chassis.
8. The submerged water-cooled server of claim 7, wherein, Both the main control module and the GPU acceleration module group have locking screws on their front panels, which can fully lock the modules inserted into the chassis to the chassis.
9. The submerged water-cooled server of claim 1, wherein, The GPU acceleration module group has a PCIe card slot, and the PCIe card slot is equipped with a GPU acceleration card.
10. The submerged water-cooled server of claim 9, wherein, The front panel of the GPU acceleration module group is equipped with slot indicator lights to indicate the presence status of the internal PCIe card slots.