Unreal engine engineering optimization packaging method suitable for localized platform
The optimization method for Unreal Engine on domestic platforms addresses compatibility and performance issues by adapting shader codes, optimizing Vulkan APIs, and integrating platform-specific toolchains, resulting in enhanced rendering efficiency and stability.
Patent Information
- Application Number
- CN202510383066.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-03-28
- Publication Date
- 2025-07-15
AI Technical Summary
Unreal Engine has compatibility problems on the domestic platform, resulting in reduced performance, inefficient graphics rendering, immature ecosystem, insufficient security and compliance, making it difficult to directly port and deploy.
By performing domestic platform adaptation and compatibility analysis on Unreal Engine, the Vulkan API, MESA library and graphics card driver are optimized, system adjustments and toolchain configuration are performed, Shader code, rendering paths and resource management are optimized, and compatibility and performance are ensured.
It significantly improves the graphics rendering performance and stability on the domestic platform, provides a smooth user experience, and solves the problem of unstable operation of Unreal Engine on the domestic platform.
Smart Images
Figure CN120315757A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of Unreal Engine, and specifically refers to an optimized packaging method for Unreal Engine projects suitable for domestic platforms. Background Art
[0002] With the increasing demand for information security, the domesticization of software and hardware has become a trend. However, many existing development tools such as Unreal Engine encounter compatibility issues when attempting to deploy their projects to domestic platforms, resulting in applications that cannot start or run stably. Especially when dealing with complex graphics rendering tasks, due to differences between different hardware architectures, direct transplantation becomes extremely difficult.
[0003] Unreal Engine (UE) encounters significant compatibility issues when attempting to deploy its projects to domestic platforms. These issues mainly stem from the following aspects:
[0004] 1. Hardware architecture differences
[0005] Modern 3D development tools are usually optimized for the x86 architecture, while domestic hardware platforms often are based on the ARM architecture (such as the Feiteng D2000 / 8 processor). Due to the significant differences in instruction sets, memory management, and parallel processing capabilities between these two architectures, directly transplanting existing projects may lead to performance degradation or incomplete functionality.
[0006] When dealing with complex graphics rendering tasks, the optimization strategies under the x86 architecture may not be directly applicable to the ARM architecture, resulting in low rendering efficiency or impaired picture quality.
[0007] 2. Insufficient support for graphics APIs
[0008] Unreal Engine widely uses Vulkan as one of its graphics APIs. However, on domestic platforms, especially in systems equipped with non-mainstream GPUs (such as the Jingjiawei JM9230), the support for Vulkan may be defective or completely absent. This forces developers to face additional adaptation work and even rewrite some code to ensure compatibility.
[0009] In addition, the MESA library (libVulkan.so), as an open-source implementation of Vulkan, may not provide the same performance and stability as closed-source drivers in some cases, further increasing the transplantation difficulty.
[0010] 3. Immature ecosystem
[0011] The software ecosystem on domesticated platforms is relatively immature, lacking optimization tools and support libraries for specific hardware and operating systems. This means that developers may need to develop or modify a large amount of underlying code themselves during transplantation, increasing the complexity and development cycle of the project.
[0012] At the same time, domestic manufacturers have insufficient experience in promoting the collaborative work of hardware and software, resulting in a lack of effective technical support and best practice guidance in solving cross-platform compatibility problems.
[0013] 4. Security and compliance requirements
[0014] With the improvement of information security awareness, domesticated platforms have put forward higher requirements for the security and compliance of software. Traditional 3D development tools may not fully consider these factors, resulting in potential security risks or compliance obstacles when running on domesticated platforms.
[0015] For example, some open-source components may contain unknown security vulnerabilities or do not comply with domestic relevant laws and regulations, affecting the market access and user trust of the final product.
[0016] Therefore, an optimized packaging method for Unreal Engine projects applicable to domesticated platforms has become an urgent problem to be solved. Summary of the Invention
[0017] The technical problem to be solved by the present invention is to provide an optimized packaging method for Unreal Engine projects applicable to domesticated platforms. By comprehensively optimizing multiple links such as Vulkan API, Unreal Engine, MESA library, and graphics card driver, a complete set of solutions is formed, thereby maximizing the graphics rendering performance and stability on the target platform. Through refined system adjustment, it is ensured that the Unreal Engine can run efficiently on different platforms and provide a smooth user experience, especially in the hardware environment of domesticated platforms.
[0018] To solve the above technical problem, the technical solution provided by the present invention is: an optimized packaging method for Unreal Engine projects applicable to domesticated platforms, including the following steps:
[0019] S1. Conduct adaptation and compatibility analysis of the Unreal Engine for domesticated platforms, perform performance optimization, and complete operating system adaptation
[0020] Analyze the GPU architecture and driver characteristics of the domestic platform, modify the Vulkan Shader code used in the Unreal Engine according to the specific hardware configurations of different target platforms; design a new PSO, adjust the parts of the Shader that are not suitable for the platform, optimize the execution path of the shader program, reduce the computational complexity, adjust the precision requirements, and optimize the graphics rendering pipeline to ensure that the Shader program fully utilizes the hardware advantages of the target platform, avoid compatibility issues, and improve the rendering performance;
[0021] S2. Platform-Specific Toolchain Configuration and Integration
[0022] Install and configure the necessary SDKs to ensure the compatibility of the toolchain with the platform and set up the build environment;
[0023] S3. Project Building and Packaging
[0024] Build and package the project into the format required by the target platform to generate an executable file;
[0025] S4. Testing and Debugging
[0026] Test the project on the target platform, debug platform-specific issues, and conduct performance testing and stress testing;
[0027] S5. Deployment and Release
[0028] Release the project to the target platform, and the project is successfully deployed and runs stably on the target platform.
[0029] Furthermore, the domestic platform is a platform equipped with the Kylin v10 operating system, the Feiteng CPU D2000 / 8 processor, and the Jingjiawei JM9230 GPU.
[0030] Furthermore, in step S1, deeply analyze the MESA library in the Linux Arm64 environment, adjust and optimize the key functions related to Vulkan; according to the characteristics of the Linux Arm64 architecture, deeply customize and modify the Vulkan driver layer in the MESA library, enhance the multi-threaded rendering support, improve the memory management mechanism, optimize the GPU instruction scheduling and resource binding; introduce the optimization strategy specific to the Jingjiawei JM9230 GPU, utilize its hardware acceleration function to accelerate texture loading and geometric transformation, and improve the overall rendering efficiency, performance, and stability.
[0031] Furthermore, in step S1, deeply analyze the part of the Unreal Engine that interacts with Vulkan internally, identify the key bottleneck points affecting performance; modify the organization of the Render Pass, optimize the usage logic of the Descriptor Set, make platform-specific modifications to the Vulkan implementation in the engine source code, and adjust the state settings, resource management, and swap chain configuration of the Vulkan rendering pipeline to support the hardware architecture of the target platform.
[0032] Furthermore, in step S1, optimize the graphics card driver based on the technical documentation of Jingjiawei JM9230 GPU and the official recommended practice guide; optimize GPU resource management, command buffer management, and synchronization mechanism, optimize the instruction scheduling algorithm, enhance error detection and recovery capabilities, and fix the performance bottlenecks in the existing driver; utilize the CGroups and NUMA node affinity settings provided by the Linux kernel to refine the resource allocation strategy, reduce GPU computing latency, improve rendering smoothness, and enhance system stability.
[0033] Furthermore, the specific method of modifying the Vulkan Shader code used in the Unreal Engine is as follows: adopt a platform adaptation algorithm, replace the mathematical operation instructions in the Shader code according to the instruction set characteristics of the target platform GPU architecture, and split the unsupported complex instructions into multiple simple instructions; adjust the texture sampling method in the Shader code according to the driver characteristics to adapt to the texture storage format of the domesticated platform.
[0034] Furthermore, the specific method of adjusting and optimizing the key functions related to Vulkan is as follows: through function overloading technology, re-implement the key functions related to multi-threaded rendering in the MESA library to optimize the thread synchronization mechanism; utilize memory mapping technology to improve the memory management mechanism, and write specific function interfaces for the hardware acceleration function of Jingjiawei JM9230 GPU, and call GPU hardware acceleration instructions in geometric transformation operations.
[0035] Furthermore, the specific method of making platform-specific modifications to the Vulkan implementation in the engine source code is as follows: use a rendering path analysis tool to identify redundant operations in the Render Pass organization method, reduce the number of renderings by re-planning the RenderPass order; utilize data structure optimization algorithms to optimize the usage logic of the Descriptor Set and reduce resource lookup time; in the engine source code, modify the Vulkan rendering pipeline state setting function for the target platform hardware architecture to make it conform to the best state configuration of the hardware.
[0036] Further, the specific method for optimizing the graphics card driver is as follows: Based on the hardware architecture of the Jingjiawei JM9230 GPU, a new instruction scheduling algorithm is written to send high-priority rendering instructions to the GPU for execution first; an error detection module is added to the driver to monitor errors during the GPU calculation process in real time. Once an error is detected, a recovery mechanism is started to reallocate resources; the CGroups mechanism of the Linux kernel is used to allocate independent resource groups for different rendering tasks, and through NUMA node affinity settings, the GPU resources are bound to the corresponding CPU nodes to improve resource access efficiency.
[0037] The advantages of the present invention compared with the prior art are as follows: By performing domestic platform adaptation, compatibility analysis, and performance optimization on the Unreal Engine, the potential of the domestic platform hardware is fully exploited, and the rendering performance is significantly improved. Compared with before optimization, the rendering efficiency is greatly increased. For example, the frame rate is significantly improved during complex scene rendering, and a smoother visual experience can be provided for users.
[0038] Through a complete optimization and packaging process, from platform adaptation to project deployment and release, the present invention ensures that projects developed based on the Unreal Engine can run stably on domestic platforms, solving the long-term problem of poor running effects of Unreal Engine projects on domestic platforms. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1 It is a flowchart of a method for optimizing and packaging an Unreal Engine project applicable to domestic platforms according to the present invention.
[0040] Figure 2 It is a flowchart of key technical improvement points of a method for optimizing and packaging an Unreal Engine project applicable to domestic platforms according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0041] Various exemplary embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be noted that: Unless otherwise specifically stated, the relative arrangements of components and steps, numerical expressions, and numerical values set forth in these embodiments do not limit the scope of the present invention.
[0042] The description of at least one exemplary embodiment below is actually only illustrative and in no way limits the present invention and its application or use.
[0043] Technologies, methods, and devices known to those of ordinary skill in the relevant art may not be discussed in detail, but where appropriate, the said technologies, methods, and devices should be regarded as part of the specification.
[0044] In all the examples shown and discussed here, any specific values should be construed as merely exemplary and not as a limitation. Thus, other examples of the exemplary embodiments may have different values.
[0045] The following further elaborates in detail on an optimized packaging method for Unreal Engine projects applicable to domestic platforms in combination with the accompanying drawings.
[0046] Combined with the attached Figure 1-2 , the present invention is introduced in detail.
[0047] Step S1: Conduct adaptation and compatibility analysis for the domestic platform with respect to the Unreal Engine, perform performance optimization, and complete the operating system adaptation
[0048] Deeply analyze the GPU architecture and driver characteristics of the domestic platform. Taking the platform with Kylin v10 operating system, FT-2000 / 8 processor, and JM9230 GPU as an example, study the hardware characteristics of this platform. For the specific hardware configurations of different target platforms, modify the Vulkan Shader code used in the Unreal Engine. Specifically, adopt a platform adaptation algorithm to replace the mathematical operation instructions in the Shader code according to the instruction set characteristics of the target platform's GPU architecture, and split the unsupported complex instructions into multiple simple instructions; according to the driver characteristics, adjust the texture sampling method in the Shader code to adapt to the texture storage format of the domestic platform.
[0049] Design a new PSO (Pipeline State Object), and adjust the parts of the Shader that are not suitable for the platform. By optimizing the execution path of the shader program, reducing the computational complexity, adjusting the precision requirements, and optimizing the graphics rendering pipeline, ensure that the Shader program fully utilizes the hardware advantages of the target platform, avoid compatibility issues, and improve the rendering performance.
[0050] Deeply analyze the MESA library in the Linux Arm64 environment, and adjust and optimize the key functions related to Vulkan. Through function overloading technology, re-implement the key functions related to multi-threaded rendering in the MESA library, and optimize the thread synchronization mechanism; use memory mapping technology to improve the memory management mechanism, and write specific function interfaces for the hardware acceleration function of the JM9230 GPU, and call the GPU hardware acceleration instructions in the geometric transformation operation.
[0051] Deeply customize and modify the Vulkan driver layer in the MESA library according to the characteristics of the Linux Arm64 architecture. Enhance multi-threaded rendering support, improve the memory management mechanism, and optimize GPU instruction scheduling and resource binding; introduce optimization strategies specific to the Jingjiawei JM9230 GPU, and utilize its hardware acceleration function to accelerate texture loading and geometric transformation, improving the overall rendering efficiency, performance, and stability.
[0052] Deeply analyze the part of the Unreal Engine that interacts with Vulkan to identify the key bottlenecks affecting performance. Use a rendering path analysis tool to identify redundant operations in the Render Pass organization method, and reduce the number of renderings by re-planning the Render Pass order; utilize data structure optimization algorithms to optimize the usage logic of Descriptor Sets and reduce resource lookup time; in the engine source code, modify the Vulkan rendering pipeline state setting function for the target platform's hardware architecture to make it conform to the best state configuration of the hardware.
[0053] Optimize the graphics card driver based on the technical documentation of the Jingjiawei JM9230 GPU and the official recommended practice guide. Based on the hardware architecture of the Jingjiawei JM9230 GPU, write a new instruction scheduling algorithm to send high-priority rendering instructions to the GPU for execution first; add an error detection module to the driver to monitor errors in the GPU calculation process in real time. Once an error is detected, start a recovery mechanism to reallocate resources; utilize the CGroups mechanism of the Linux kernel to allocate independent resource groups for different rendering tasks, and bind the GPU resources to the corresponding CPU nodes through NUMA node affinity settings to improve resource access efficiency.
[0054] Step S2: Platform-Specific Toolchain Configuration and Integration
[0055] Install and configure the necessary SDK (Software Development Kit) to ensure that the toolchain is compatible with the platform, set up the build environment, and prepare for the build and packaging of the project.
[0056] Step S3: Project Build and Packaging
[0057] Build and package the project into the format required by the target platform to generate an executable file, enabling the project to run on the target domestic platform.
[0058] Step S4: Testing and Debugging
[0059] Test the project on the target platform, debug platform-specific issues, conduct performance testing and stress testing to ensure the stability and performance of the project in different scenarios.
[0060] Step S5: Deployment and Release
[0061] Release the project to the target platform. The project is successfully deployed and runs stably on the target platform, completing the entire optimized packaging process.
[0062] The specific implementation process of an optimized packaging method for Unreal Engine projects applicable to domestic platforms in the present invention is as follows:
[0063] Step S1, Vulkan Shader code modification: Scan the Vulkan Shader code in the Unreal Engine according to the instruction set of the Jingjiawei JM9230 GPU. For example, if complex matrix operation instructions not supported by the platform are found in the code, use a platform adaptation algorithm to replace them with a combination of multiple simple vector operation instructions to adapt to the hardware architecture of the GPU. At the same time, according to the texture storage format characteristics of the Jingjiawei JM9230 GPU, adjust the parameters and logic of the texture sampling function in the Shader code to ensure that data can be correctly and efficiently obtained from the texture memory.
[0064] PSO design and Shader adjustment: Use a graphics rendering analysis tool to analyze the execution path of the existing Shader program and find parts with high computational complexity. Redesign the execution path through an optimization algorithm to reduce unnecessary calculation steps. For example, for some constants that can be calculated in advance in the vertex shader stage, advance the calculation steps to avoid repeated calculations in the fragment shader, thereby reducing the computational complexity. At the same time, reasonably adjust the precision requirements in the Shader according to the precision support range of the platform hardware to improve the calculation efficiency while ensuring the graphics quality.
[0065] Optimization related to the MESA library: In the Linux Arm64 environment, use function overloading technology to re-implement key functions related to multi-threaded rendering in the MESA library, such as thread synchronization functions. In a multi-threaded rendering scenario, the optimized thread synchronization mechanism can effectively reduce thread waiting time and improve multi-threaded rendering efficiency. Use memory mapping technology to improve the memory management mechanism in the MESA library. For example, map frequently accessed texture data to a specific memory area to reduce memory access latency. Write a dedicated function interface for the hardware acceleration function of the Jingjiawei JM9230 GPU. When performing geometric transformation operations, trigger GPU hardware acceleration instructions by calling this interface to accelerate the geometric transformation process.
[0066] Customization and Modification of Vulkan Driver Layer: In view of the characteristics of the Linux Arm64 architecture, the Vulkan driver layer in the MESA library is deeply modified. In terms of multi-threaded rendering support, the thread scheduling algorithm is optimized to ensure that the rendering tasks of different threads can be reasonably allocated to CPU cores, making full use of the multi-core advantages. The memory management mechanism is improved, and memory resources are reasonably allocated according to the priority of rendering tasks and data access frequency to reduce memory fragmentation. The GPU instruction scheduling is optimized, and high-priority rendering instructions are preferentially sent to the GPU for execution. At the same time, the resource binding process is optimized to reduce the resource binding time and improve the rendering efficiency.
[0067] Optimization of the Internal Vulkan Interaction in Unreal Engine: Use a rendering path analysis tool to analyze the organization method of Render Pass in the internal Vulkan interaction of the Unreal Engine, and identify redundant Render Pass. For example, for some front and back processing Render Pass that can be merged, they are merged into one Render Pass by re-planning the order to reduce the number of renderings. The usage logic of Descriptor Set is optimized using data structure optimization algorithms. For example, the data structure of DescriptorSet is changed from linear search to hash search to reduce the resource search time. In the engine source code, for the hardware architecture of Jingjiawei JM9230 GPU, the Vulkan rendering pipeline state setting function is modified to ensure that the state setting of the rendering pipeline conforms to the best performance configuration of the hardware.
[0068] Optimization of Graphics Card Driver: Based on the hardware architecture of Jingjiawei JM9230 GPU, a new instruction scheduling algorithm is written. During the execution of rendering tasks, high-priority rendering instructions such as scene drawing and special effect rendering are preferentially sent to the GPU for execution to ensure that key rendering tasks can be completed in a timely manner. An error detection module is added to the driver to monitor errors in the GPU calculation process in real time, such as detecting GPU memory access errors and instruction execution errors. Once an error is detected, the recovery mechanism is immediately started to reallocate resources to ensure that the rendering tasks can continue. Using the CGroups mechanism of the Linux kernel, independent resource groups are allocated for different rendering tasks, such as scene rendering and UI rendering, to ensure resource isolation between tasks. Through NUMA node affinity settings, the GPU resources are bound to the corresponding CPU nodes to reduce the cross-node data transfer latency and improve the resource access efficiency.
[0069] Step S2: For the platform equipped with Kylin v10 operating system, Feiteng CPU D2000 / 8 processor, and Jingjiawei JM9230 GPU, download and install the SDK applicable to this platform from the official channels, including the graphics library SDK, compiler SDK, etc. During the installation process, strictly configure according to the installation guide provided by the SDK to ensure the compatibility between the SDK and the platform. Set the build environment variables and add the relevant paths of the SDK to the system environment variables so that the tools and libraries in the SDK can be correctly called during the project build process.
[0070] Step S3: After completing the configuration and integration of the platform-specific toolchain, use the project build tool provided by the Unreal Engine and select the target platform as the one equipped with Kylin v10 operating system, Feiteng CPU D2000 / 8 processor, and Jingjiawei JM9230 GPU. According to the requirements of this platform, set the packaging format of the project, such as generating the executable file format applicable to this platform and the packaging method of relevant resource files. Start the build and packaging process and wait for the build tool to complete operations such as project compilation and linking to generate the final executable file and relevant resource packages.
[0071] Step S4: Deploy the generated project executable file and resource packages to the target platform. Start the project on the target platform for functional testing to check whether each functional module of the project can run normally. Conduct performance testing, use performance monitoring tools such as frame rate monitoring tools and memory occupancy monitoring tools to monitor performance metrics such as frame rate and memory occupancy during the running of the project. Conduct stress testing, simulate high-load scenarios such as loading a large number of models and special effects simultaneously, and test the stability and performance of the project under stress conditions. For platform-specific problems found during the testing process, such as abnormal rendering and performance bottlenecks, use debugging tools to debug, locate the problems and fix them.
[0072] Step S5: After completing the testing and debugging to ensure that the project can run stably and with high performance on the target platform, release the project to the application market of the target platform or the designated release channels. During the release process, provide detailed project descriptions and installation guides to facilitate users to download and install the project. At the same time, continuously pay attention to the feedback during the actual use of the project, promptly handle the problems encountered by users, and conduct subsequent optimization and updates of the project.
[0073] The above describes the present invention and its implementation manners. Such description is not restrictive. What is shown in the drawings is only one of the implementation manners of the present invention, and the actual structure is not limited thereto. Generally speaking, if those of ordinary skill in the art are inspired by it and without departing from the purpose of the present invention creation, they design similar structural manners and embodiments to this technical solution without creative efforts, and all should fall within the protection scope of the present invention.
Claims
1. An optimization and packaging method for Unreal Engine projects applicable to domestic platforms, characterized in that: It includes the following steps: S1. Conduct domesticated platform adaptation and compatibility analysis for the Unreal Engine, perform performance optimization, and complete the operating system adaptation. Analyze the GPU architecture and driver characteristics of the domesticated platform. For the specific hardware configurations of different target platforms, modify the Vulkan Shader code used in the Unreal Engine; design a new PSO, adjust the parts of the Shader that are not suitable for the platform, optimize the execution path of the shader program, reduce the computational complexity, adjust the precision requirements, and optimize the graphics rendering pipeline to ensure that the Shader program fully utilizes the hardware advantages of the target platform, avoid compatibility issues, and improve the rendering performance. S2. Configure and integrate the platform-specific toolchain. Install and configure the necessary SDKs to ensure that the toolchain is compatible with the platform and set up the build environment. S3. Build and package the project. Build and package the project into the format required by the target platform to generate an executable file. S4. Test and debug. Test the project on the target platform, debug platform-specific issues, and perform performance testing and stress testing. S5. Deploy and release. Release the project to the target platform, and the project is successfully deployed and runs stably on the target platform.
2. The optimized packaging method for Unreal Engine projects applicable to domestic platforms according to claim 1, wherein: The domesticated platform is a platform equipped with the Kylin v10 operating system, the Feiteng CPU D2000 / 8 processor, and the Jingjiawei JM9230 GPU.
3. The optimized packaging method for Unreal Engine projects applicable to domestic platforms according to claim 2, characterized in that: In step S1, deeply analyze the MESA library in the Linux Arm64 environment, and adjust and optimize the key functions related to Vulkan; for the characteristics of the Linux Arm64 architecture, deeply customize and modify the Vulkan driver layer in the MESA library, enhance the multi-threaded rendering support, improve the memory management mechanism, optimize the GPU instruction scheduling and resource binding; introduce the optimization strategy specific to the Jingjiawei JM9230 GPU, and utilize its hardware acceleration function to accelerate texture loading and geometric transformation to improve the overall rendering efficiency, performance, and stability.
4. The optimized packaging method for Unreal Engine projects applicable to domestic platforms according to claim 3, characterized in that: In step S1, deeply analyze the part of the Unreal Engine that interacts with Vulkan internally, identify the key bottleneck points affecting performance; modify the organization method of the Render Pass, optimize the usage logic of the Descriptor Set, make platform-specific modifications to the Vulkan implementation in the engine source code, and adjust the state settings, resource management, and swap chain configuration of the Vulkan rendering pipeline to make it support the hardware architecture of the target platform.
5. The optimized packaging method for Unreal Engine projects applicable to domestic platforms according to claim 4, characterized in that: In step S1, optimize the graphics card driver based on the technical documentation of the Jingjiawei JM9230 GPU and the official recommended practice guide. Optimize the GPU resource management, command buffer management, and synchronization mechanism, optimize the instruction scheduling algorithm, enhance the error detection and recovery ability, and fix the performance bottlenecks in the existing driver; utilize the CGroups and NUMA node affinity settings provided by the Linux kernel to refine the resource allocation strategy, reduce the GPU computing latency, improve the rendering fluency, and enhance the system stability.
6. The method for optimizing and packaging an Unreal Engine project applicable to a domestic platform according to claim 5, wherein: The specific way to modify the Vulkan Shader code used in the Unreal Engine is as follows: adopt a platform adaptation algorithm, according to the instruction set characteristics of the target platform's GPU architecture, replace the mathematical operation instructions in the Shader code, and split the unsupported complex instructions into multiple simple instructions; adjust the texture sampling method in the Shader code according to the driver characteristics to adapt to the texture storage format of the domestic platform.
7. A method for optimizing and packaging Unreal Engine projects suitable for domestic platforms according to claim 6, characterized in that: The specific way to adjust and optimize the key functions related to Vulkan is as follows: through function overloading technology, re-implement the key functions related to multi-threaded rendering in the MESA library, and optimize the thread synchronization mechanism; use memory mapping technology to improve the memory management mechanism, and write specific function interfaces for the hardware acceleration function of the Jingjiawei JM9230 GPU, and call GPU hardware acceleration instructions in geometric transformation operations.
8. A method for optimizing and packaging Unreal Engine projects applicable to domestic platforms according to claim 7, characterized in that: The specific way to make platform-specific modifications to the Vulkan implementation in the engine source code is as follows: adopt a render path analysis tool to identify redundant operations in the Render Pass organization method, and reduce the number of renders by re-planning the Render Pass order; use a data structure optimization algorithm to optimize the usage logic of the Descriptor Set and reduce the resource lookup time; in the engine source code, modify the Vulkan render pipeline state setting function for the target platform's hardware architecture to make it conform to the best state configuration of the hardware.
9. The optimized packaging method for Unreal Engine projects applicable to domestic platforms according to claim 8, characterized in that: The specific way to optimize the graphics card driver is as follows: based on the hardware architecture of the Jingjiawei JM9230 GPU, write a new instruction scheduling algorithm to give priority to sending high-priority rendering instructions to the GPU for execution; add an error detection module to the driver to monitor errors in the GPU calculation process in real time. Once an error is detected, start the recovery mechanism and re-allocate resources; use the CGroups mechanism of the Linux kernel to allocate independent resource groups for different rendering tasks, and bind the GPU resources to the corresponding CPU nodes through NUMA node affinity settings to improve resource access efficiency.