ncnn
ncnn is a high-performance neural network inference framework optimized for the mobile platform
About ncnn
ncnn is a high-performance neural network inference framework optimized for the mobile platform
What's New in v20260113
编译版本,默认配置,android-ndk-r29,xcode 16.4,ubuntu-22.04,ubuntu-24.04,vs2015,vs2017,vs2019,vs2022,emscripten-3.1.28
| file | content | arch |
|---|---|---|
| ncnn-full-source.zip | 包含全部 submodule 代码的完整源码 | |
| ncnn-android.zip | android 静态库/动态库 | armeabi-v7a + arm64-v8a + x86 + x86_64 + riscv64 |
| ncnn-android-vulkan.zip | android 静态库/动态库,支持 GPU | armeabi-v7a + arm64-v8a + x86 + x86_64 + riscv64 |
| ncnn-apple.zip | apple xcframework,ios + ios-simulator + macos + mac-catalyst + watchos + watchos-simulator + tvos + tvos-simulator + visionos + visionos-simulator | arm64 + arm64e + x86_64 |
| ncnn-apple-vulkan.zip | apple xcframework,ios + ios-simulator + macos + mac-catalyst + watchos + watchos-simulator + tvos + tvos-simulator + visionos + visionos-simulator,支持 GPU | arm64 + arm64e + x86_64 |
| ncnn-ios.zip | ios 静态库 | arm64 |
| ncnn-ios-vulkan.zip | ios 静态库,支持 GPU | arm64 |
| ncnn-ios-simulator.zip | ios simulator 静态库 | x86_64 + arm64 |
| ncnn-ios-simulator-vulkan.zip | ios simulator 静态库,支持 GPU | x86_64 + arm64 |
| ncnn-macos.zip | macos 静态库 | x86_64 + arm64 |
| ncnn-macos-vulkan.zip | macos 静态库,支持 GPU | x86_64 + arm64 |
| ncnn-mac-catalyst.zip | mac catalyst 静态库 | x86_64 + arm64 |
| ncnn-mac-catalyst-vulkan.zip | mac catalyst 静态库,支持 GPU | x86_64 + arm64 |
| ncnn-watchos.zip | watchos 静态库 | armv7k + arm64_32 |
| ncnn-watchos-simulator.zip | watchos simulator 静态库 | x86_64 + arm64 |
| ncnn-tvos.zip | tvos 静态库 | x86_64 + arm64 |
| ncnn-tvos-vulkan.zip | tvos 静态库,支持 GPU | x86_64 + arm64 |
| ncnn-tvos-simulator.zip | tvos simulator 静态库 | x86_64 + arm64 |
| ncnn-tvos-simulator-vulkan.zip | tvos simulator 静态库,支持 GPU | x86_64 + arm64 |
| ncnn-visionos.zip | visionos 静态库 | arm64 |
| ncnn-visionos-vulkan.zip | visionos 静态库,支持 GPU | arm64 |
| ncnn-visionos-simulator.zip | visionos simulator 静态库 | x86_64 + arm64 |
| ncnn-visionos-simulator-vulkan.zip | visionos simulator 静态库,支持 GPU | x86_64 + arm64 |
| ncnn-ubuntu.zip | ubuntu linux 静态库/动态库,支持 GPU,模型转换工具 | x86_64 |
| ncnn-windows.zip | windows 静态库/动态库,支持 GPU,模型转换工具 | x86 + x64 + arm + arm64 |
| ncnn-webassembly.zip | webassembly 静态库 | wasm32 + simd + threads + simd-threads |
新增sdpa layer和pnnx torch.scaled_dot_product_attention的转换,支持gqa合并
新增rotaryembed layer
sdpa支持kvcache
multiheadattention支持kvcache
layer可选实现support_vulkan_packing
layer可选实现support_vulkan_any_packing
vulkan支持bf16开关,支持旧显卡模拟转换bf16
rmsnorm vulkan优化(@futz12)
selu vulkan优化(@futz12)
vulkan eltwise统一elempack shader
简化vulkan cast
改善M较小时在N上切块的多线程调度
gemm x86 avx512采用N维度16切块优化
sdpa x86使用gemm和softmax优化(@futz12)
arm neon数学函数优先使用fma指令优化(@Abandon-ht)
unaryop tan rvv优化(@ihb2032 @lyd1992)
新增cmake NCNN_WINXP开关,不再主动定义_WIN32_WINNT宏
c-api新增ncnn_version_number()接口返回数值
c-api新增更多option setter getter接口
net加载模型接口新增wchar_t参数类型
新增float8和bfloat8转换函数(@chloeee99)
格式化glsl文件
删除shader注释和额外的空格
不再编译onnx2ncnn
benchncnn内置模型param,运行时不再需要param文件
修复modelwriter访问空bias数据崩溃问题(@csukuangfj)
修复param解析时尝试对已读取数据再次读取的逻辑错误(@futz12)
修复softmax多线程尾部余数的错误,优化倒数计算(@futz12)
修复msvc编译器x86 lstm int8开启vnni指令集时计算错误
修复x86 lstm int8越界读写问题
修复加载模型param出错时的退出逻辑(@Cat-myq)
修复vulkan驱动返回无效subgroup size导致加载卡死的问题(@Cat-myq)
修复加载模型时CRLF行尾解析逻辑错误(@chennevwin)
修复sdpa单通道attnmask的处理逻辑
修复ncnn2int8对仅有反量化的输出int8 scales保存崩溃问题
modelwriter支持tile层
ncnn2mem支持新的数组和字符串类型
x86上引用传参simd寄存器类型,函数无法接受对齐类型传值
检查gpu显存分配失败错误,返回错误码(@Upliner)
simplevk支持查找高通windows vulkan驱动文件(@strongtz)
simplevk支持apple平台动态加载vulkan驱动
simplevk支持VK_DRIVER_FILES环境变量加载vulkan驱动
禁用windows amd rdna2驱动的cooperative matrix软件模拟功能以提升性能
glslang更新到20260109
适配新windows-sdk的更多arm处理器特性检查功能
更新pybind 3.0.1,修复python-3.14使用pyncnn崩溃问题
更新pnnx到torch-2.9,支持onnx external data,支持dynamo-exported onnx
pnnx支持转换torch.shrink Tensor.unflatten torch.flatten
pnnx转换torch.flatten到ncnn支持多动态维度
pnnx支持转换F.interpolate nearest-exact
pnnx修复转换Tensor.expand到ncnn缺失的repeats
pnnx支持转换onnx gelu groupnorm rmsnorm gridsample
pnnx支持合并更多transformer attention变种
pnnx支持合并更多sdpa attention变种
pnnx支持合并更多rmsnorm变种
pnnx合并连续permute,删除无用的permute
pnnx添加deepseek_v3和qwen2 attention转换测试
pnnx合并非interleaved和更多的interleaved rope模块
pnnx合并t5风格的无gamma layernorm
pnnx总是删除contiguous,view统一转为reshape
pnnx转换onnx reshape丢弃allowzero参数
pnnx修复onnx旧版opset模型的部分shape折叠
pnnx修复折叠的常量输入丢弃逻辑
pnnx修复onnx padding非常量数值的转换
pnnx修复转换torch.stack负数axis越界崩溃问题
pnnx支持转换onnx动态resize
pnnx合并相同常量为一个
pnnx改善paddle风格的tensor.size模式
pnnx改善合并whisper风格的attention
pnnx自动从onnx模型中获取输入shape
pnnx生成的推理代码在自动shape时生成有效shape
pnnx改善pnnx.py中浮点数的表示方式
pnnx转换onnx模型不再输出无用的open failed警告
pnnx在pnnx.py中生成export_pnnx和export_ncnn工具函数
pnnx检查import xxx_pnnx路径,跳过目录检查(@glenn-jocher)
修复pnnx windows编译
ppocrv5分割英文文本时保留空格(@sxj731533730)
修复whisper例子中ffmpeg命令错误(@quink-black)
whisper截断音频时长到30秒
新增arcface示例(@heabeounMKTO)
gpu单元测试丢弃shape hint测试的pipeline缓存减少gpu显存占用
删除无用的testutil layer hook功能
新增gemm oom单元测试
ci比较二进制任务改用pull_request触发
ci修复windows-xp编译,统一workflow文件
ci更新mingw工具链下载地址
ci asan任务优化存储占用
ci新增aarch64 asan任务
ci更新macos-13到macos-15-intel(@Willaaaaaaa)
ci更新windows-sdk和swiftshader
删除已停用的tencent ci(@mpj1234)
更新onnx模型转换文档
readme添加8bit量化文档链接(@mlbo)
编译步骤增加make install(@roachsinai)
添加打印VkMat内容的文档
新增Arduino UNO Q性能数据(@SimoSbara)
发布linux riscv64的python wheel
发布macos arm64的python pypy wheel
New Contributors
- @sxj731533730 made their first contribution in https://github.com/Tencent/ncnn/pull/6350
- @mpj1234 made their first contribution in https://github.com/Tencent/ncnn/pull/6355
- @glenn-jocher made their first contribution in https://github.com/Tencent/ncnn/pull/6379
- @Abandon-ht made their first contribution in https://github.com/Tencent/ncnn/pull/6393
- @Cat-myq made their first contribution in https://github.com/Tencent/ncnn/pull/6383
- @heabeounMKTO made their first contribution in https://github.com/Tencent/ncnn/pull/6386
- @SimoSbara made their first contribution in https://github.com/Tencent/ncnn/pull/6454
- @ihb2032 made their first contribution in https://github.com/Tencent/ncnn/pull/6460
- @chennevwin made their first contribution in https://github.com/Tencent/ncnn/pull/6472
- @0130w made their first contribution in https://github.com/Tencent/ncnn/pull/6286
- @chloeee99 made their first contribution in https://github.com/Tencent/ncnn/pull/6495
Full Changelog: https://github.com/Tencent/ncnn/compare/20250916...20260113
Legal Notice
This software is provided by its original developers. We only index and provide links to official download sources. All downloads are from the original publishers and are completely legal and safe.
Download Information
Related Apps
scrcpy
GitHubDisplay and control your Android device
react-native
GitHubA framework for building native applications using React
rustdesk
GitHubAn open-source remote desktop application designed for self-hosting, as an alternative to TeamViewer.
rustdesk
GitHubAn open-source remote desktop application designed for self-hosting, as an alternative to TeamViewer.