diff --git a/docs/zh/server/_toc.yaml b/docs/zh/server/_toc.yaml
index af3220972c8ed0b3fb8e7901aed8ef0f69bdea94..03925b5f2456314959e537bcaf42464e412ed824 100644
--- a/docs/zh/server/_toc.yaml
+++ b/docs/zh/server/_toc.yaml
@@ -78,6 +78,7 @@ sections:
           upstream: https://gitee.com/openeuler/compiler-docs/blob/openEuler-24.03-LTS-SP2/docs/zh/bisheng_autotuner/_toc.yaml
       - href: ./development/ai4c/_toc.yaml
       - href: ./development/fangtian/_toc.yaml
+      - href: ./development/annc/_toc.yaml
   - label: HA高可用
     sections:
       - href: ./high_availability/ha/_toc.yaml
diff --git a/docs/zh/server/development/annc/_toc.yaml b/docs/zh/server/development/annc/_toc.yaml
new file mode 100644
index 0000000000000000000000000000000000000000..ab3b9e9edca68e0b23dd4cd6d4334b0fdbef1b1d
--- /dev/null
+++ b/docs/zh/server/development/annc/_toc.yaml
@@ -0,0 +1,6 @@
+label: ANNC使用手册
+isManual: true
+description: ANNC（Accelerated Neural Network Compiler）是专注于加速神经网络计算的编译器，聚焦于通过计算图优化，高性能融合算子生成和对接技术以及高效代码生成和优化能力，加速推荐和大模型的推理性能，支持主流开源推理框架接入。
+sections:
+  - label: ANNC使用手册
+    href: ./annc_user_manual.md
diff --git a/docs/zh/server/development/annc/annc_user_manual.md b/docs/zh/server/development/annc/annc_user_manual.md
new file mode 100644
index 0000000000000000000000000000000000000000..f8d229c2d11b838a5f7ee2e1ca7064993a44caaa
--- /dev/null
+++ b/docs/zh/server/development/annc/annc_user_manual.md
@@ -0,0 +1,248 @@
+# ANNC 使用手册
+
+## 1 ANNC 介绍
+
+ANNC（Accelerated Neural Network Compiler）是专注于加速神经网络计算的编译器，聚焦于通过计算图优化，高性能融合算子生成和对接技术以及高效代码生成和优化能力，加速推荐和大模型的推理性能，支持主流开源推理框架接入。
+
+## 2 ANNC 的安装构建
+
+### 2.1 直接安装ANNC
+
+若用户使用最新的openEuler系统（24.03-LTS-SP2），可以直接安装`ANNC`包。
+
+```shell
+yum install -y ANNC
+```
+
+若用户使用其他版本的`ANNC`特性或在其他OS版本中安装`ANNC`，需重新构建`ANNC`，可以参考以下步骤。
+
+### 2.2 RPM包构建安装流程（推荐）
+
+1. 使用 root 权限，安装 rpmbuild、rpmdevtools，具体命令如下：
+
+   ```bash
+   # 安装 rpmbuild
+   yum install dnf-plugins-core rpm-build
+   # 安装 rpmdevtools
+   yum install rpmdevtools
+   ```
+
+2. 在主目录`/root`下生成 rpmbuild 文件夹：
+
+   ```bash
+   rpmdev-setuptree
+   # 检查自动生成的目录结构
+   ls ~/rpmbuild/
+   BUILD  BUILDROOT  RPMS  SOURCES  SPECS  SRPMS
+   ```
+
+3. 使用`git clone -b openEuler-24.03-LTS-SP2 https://gitee.com/src-openeuler/ANNC.git`，从目标仓库的 `openEuler-24.03-LTS-SP2` 分支拉取代码，并把目标文件放入 rpmbuild 的相应文件夹下：
+
+   ``` shell
+   cp ANNC/*.tar.gz* ~/rpmbuild/SOURCES
+   cp ANNC/*.patch ~/rpmbuild/SOURCES/
+   cp ANNC/ANNC.spec ~/rpmbuild/SPECS/
+   ```
+
+4. 用户可通过以下步骤生成 `ANNC` 的 RPM 包：
+
+   ```bash
+   # 安装 ANNC 所需依赖
+   yum-builddep ~/rpmbuild/SPECS/ANNC.spec
+   # 构建 ANNC 依赖包
+   # 若出现 check-rpaths 相关报错，则需要在 rpmbuild 前添加 QA_RPATHS=0x0002，例如
+   # QA_RPATHS=0x0002 rpmbuild -ba ~/rpmbuild/SPECS/ANNC.spec
+   rpmbuild -ba ~/rpmbuild/SPECS/ANNC.spec
+   # 安装 RPM 包
+   cd ~/rpmbuild/RPMS/<arch>
+   rpm -ivh ANNC-<version>-<release>.<arch>.rpm
+   ```
+
+   注意事项：若系统因存有旧版本的 RPM 安装包而导致文件冲突，可以通过以下方式解决：
+
+   ```bash
+   # 解决方案一：强制安装新版本
+   rpm -ivh ANNC-<version>-<release>.<arch>.rpm --force
+   # 解决方案二：更新安装包
+   rpm -Uvh ANNC-<version>-<release>.<arch>.rpm
+   ```
+
+### 2.3 源码构建安装流程
+
+ANNC 的源码地址：<https://gitee.com/openeuler/ANNC>
+
+保证以下依赖包已安装：
+
+```shell
+yum install -y gcc gcc-c++ bzip2 python3-devel python3-numpy python3-setuptools python3-wheel libstdc++-static java-11-openjdk java-11-openjdk-devel make
+```
+
+安装bazel, 从该地址 <https://releases.bazel.build/6.5.0/release/bazel-6.5.0-dist.zip> 获取bazel-6.5.0包
+
+```bash
+unzip bazel-6.5.0-dist.zip -d bazel-6.5.0
+cd bazel-6.5.0
+env EXTRA_BAZEL_ARGS="--tool_java_runtime_version=local_jdk" bash ./compile.sh
+
+export PATH=/path/to/bazel-6.5.0/output:$PATH
+bazel --version
+```
+
+安装ANNC, 从源码地址下载ANNC源码包
+
+```bash
+git clone https://gitee.com/openeuler/ANNC.git
+
+cd ANNC
+
+bazel --output_user_root=./output build -c opt \
+    --copt="-DANNC_ENABLE_GRAPH_OPT" \
+    --copt="-DANNC_ENABLE_OPENBLAS" \
+    annc/service/cpu:libannc.so
+
+cp bazel-bin/annc/service/cpu/libannc.so /usr/lib64
+mkdir -p /usr/include/annc
+cp annc/service/cpu/kdnn_rewriter.h /usr/include/annc
+cd python
+python3 setup.py bdist_wheel
+python3 -m pip install dist/*.whl
+```
+
+## 3 使用流程
+
+**注意事项：**
+
+* ANNC使用者需提前部署好tf-serving，通过编译选项和代码补丁的方式接入ANNC编译优化扩展套件
+
+### 3.1 图融合结合手动大算子
+
+下载基线模型
+
+```bash
+git clone https://gitee.com/openeuler/sra_benchmark.git
+```
+
+从基线模型库中获取以下目标推荐模型 **DeepFM、DFFM、DLRM、W&D**
+
+命令行实现图融合
+
+```bash
+# 安装依赖库
+
+python3 -m pip install tensorflow==2.15.1
+
+# 运行模型转换，以DeepFM模型为例
+
+annc-opt -I /path/to/model_DeepFM/1730800001/1 -O deepfm_new/1 dnn_sparse linear_sparse
+cp -r /path/to/model_DeepFM/1730800001/1/variables deepfm_new/1
+```
+
+输出目录deepfm_new/1下应生成新的模型文件saved_model.pbtxt，搜索KPFusedSparseEmbedding, 确认图融合算子正确生成
+   
+然后将ANNC提供的开源算子库注册到tf-serving
+
+```bash
+# 进入tf-serving目录，创建自定义算子文件夹
+
+cd /path/to/serving
+mkdir tensorflow_serving/custom_ops
+
+# 将ANNC算子拷贝到该目录下
+
+cp /usr/include/annc/fused*.cc tensorflow_serving/custom_ops/
+```
+
+创建算子编译文件tensorflow_serving/custom_ops/BUILD, 并在该文件中写入以下内容：
+
+```ini
+package(
+   default_visibility = [
+        "//visibility:public",
+       ],
+       licenses = ["notice"],
+)
+
+cc_library(
+    name = 'recom_embedding_ops',
+    srcs = [
+      "fused_sparse_embedding.cc",
+      "fused_linear_embedding_with_hash_bucket.cc",
+      "fused_dnn_embedding_with_hash_bucket.cc"
+     ],
+     alwayslink = 1,
+     deps = [
+       "@org_tensorflow//tensorflow/core:framework",
+     ]
+)
+```
+
+```bash
+# 打开 tensorflow_serving/model_servers/BUILD, 搜索SUPPORTED_TENSORFLOW_OPS, 添加以下内容注册我们的算子:
+
+"//tensorflow_serving/custom_ops:recom_embedding_ops"
+```
+
+完成算子注册后，使用以下命令重新编译tf-serving，编译成功即表示算子成功注册：
+
+```bash
+ bazel --output_user_root=./output build -c opt --distdir=./proxy \
+   tensorflow_serving/model_servers:tensorflow_model_server
+```
+
+### 3.2 自动图融合和算子融合
+
+进入tf-serving目录，使用如下选项重新编译tf-serving使能ANNC：
+
+```bash
+bazel --output_user_root=./output build -c opt --distdir=./proxy \
+   --copt=-DANNC_ENABLED_KDNN \
+   tensorflow_serving/model_servers:tensorflow_model_server
+```
+
+启动tf-serving，制定目标模型，确认服务正常启动即可
+
+### 3.3 算子内存和布局优化
+
+进入tf-serving目录，需打上相关patch并重新编译
+
+```bash
+# 打上相关patch， patch位于ANNC目录下
+
+export ANNC_PATH=/path/to/ANNC
+cp $ANNC_PATH/tfserver/llvm/llvm.sh /path/to/serving/output/{id}/external/llvm-raw
+cp ANNC_PATH/tfserver/xla/xla.sh /path/to/serving/output/{id}/external/org_tensorflow/third_party/xla
+
+cd path/to/serving/output/{id}/external/llvm-raw
+bash ./llvm.sh
+cd /path/to/serving/output/{id}/external/org_tensorflow/third_party/xla
+bash ./xla.sh
+
+# 重新编译
+
+bazel --output_user_root=./output build -c opt --distdir=./proxy \
+   --copt=-DANNC_ENABLED_KDNN \
+   tensorflow_serving/model_servers:tensorflow_model_server
+```
+
+设置环境变量，开启优化特性
+
+```bash
+export XLA_FLAGS="--xla_cpu_use_xla_runtime=true --xla_cpu_enable_concat_optimization=true --xla_cpu_enable_output_tensor_reuse=true --xla_cpu_enable_mlir_tiling_and_fusion=true"
+```
+
+### 3.4  算子选择和算子库对接
+
+进入tf-serving目录，重新编译
+
+```bash
+cd ./output/{id}/external/org_tensorflow
+patch -p1 < /usr/include/tensorflow.patch
+
+# 开启MatMul算子下发和算子库对接编译选项
+
+cd /path/to/serving
+bazel --output_user_root=./output build -c opt --distdir=./proxy \
+   --copt=-DANNC_ENABLED_KDNN \
+   --copt=-DDISABLE_TF_MATMUL_FUSION \
+   tensorflow_serving/model_servers:tensorflow_model_server
+```