diff --git a/docs/quickstart.md b/docs/quickstart.md
new file mode 100644
index 0000000000000000000000000000000000000000..0ad6f1cf1ff8f819c0cc51fe9eea7ec80bc9289f
--- /dev/null
+++ b/docs/quickstart.md
@@ -0,0 +1,63 @@
+# 快速开始
+
+## 前提条件
+
+- OS: Ubuntu 20.04
+- Python: 3.x
+- Docker
+- make
+- GCC && Linux kernel
+
+|Linux kernel版本|GCC版本|
+|:---:|:---:|
+|3.x|GCC 4|
+|4.0 ~ 4.5|GCC 5|
+|4.6 ~ 5.18|GCC 7/9|
+|5.19及以上|GCC 12|
+- 天数智算软件栈必要的头文件，详情请联系您的应用工程师
+- 确认已安装天数卡：
+```bash
+[root@localhost ~]# lspci -vv | grep 1e3e
+# 0003表示天垓150加速卡，0002表示智铠加速卡，0001表示天垓100加速卡。
+```
+
+## 软件栈安装-Docker方式
+我们以天数智算软件栈版本4.3.0为例进行阐述安装过程
+1. 登录[天数智芯官网](https://www.iluvatar.com/)，进入`客户支持`>`资源中心`页面进行选择下载。如下图所示，我们选择下载智铠100的大模型软件栈安装包。
+![get-docker-installer](../resources/get-docker-installer.png)
+2. 点击右侧按钮，生成wget下载链接，然后复制链接到`机器上`执行下载
+![wget-docker-link](../resources/wget-docker-link.png)
+![download-docker-installer](../resources/download-docker-installer.png)
+3. 解压并执行安装
+```bash
+[root@localhost ~]# unzip '软件栈V4.3.0(x86)容器化部署安装包_适用大模型.zip'
+[root@localhost ~]# cd '软件栈V4.3.0(x86)容器化部署安装包_适用大模型'/
+[root@localhost 软件栈V4.3.0(x86)容器化部署安装包_适用大模型]# bash corex-docker-installer-4.3.0-10.2-ubuntu20.04-llm-py3.10-x86_64.run
+```
+4. 如您同意协议的条件条款，请输入`accept`并按回车继续；如您不同意，请输入`decline`，将退出安装。  
+![accept-docker](../resources/accept-docker.png)
+5. 勾选`Install driver`，选择`Install`，按回车键进行安装            
+![install-docker](../resources/install-docker.png)
+6. 等待安装完成                 
+![complete-install](../resources/complete-install.png)
+
+## 示例举例
+**注：如果是小模型，请下载安装适用小模型的软件栈容器。**
+### 推理示例
+1. 启动进入容器
+```bash
+docker run -it -v /usr/src:/usr/src -v /lib/modules:/lib/modules -v /dev:/dev --privileged --cap-add=ALL --pid=host corex:4.3.0
+```
+2. 克隆[DeepSparkInference](https://gitee.com/deep-spark/deepsparkinference)仓库
+3. 按照[qwen2-7b](https://gitee.com/deep-spark/deepsparkinference/tree/master/models/nlp/llm/qwen2-7b/vllm)说明，执行推理
+4. 当然您也可以执行[这里](https://gitee.com/deep-spark/deepsparkinference#%E6%A8%A1%E5%9E%8B%E5%BA%93)的任何一个`IXUCA SDK 4.3.0`的模型进行推理
+
+
+### 训练示例
+1. 启动进入容器
+```bash
+docker run -it -v /usr/src:/usr/src -v /lib/modules:/lib/modules -v /dev:/dev --privileged --cap-add=ALL --pid=host corex:4.3.0
+```
+2. 克隆[DeepSparkHub](https://gitee.com/deep-spark/deepsparkhub)仓库
+3. 按照[llama3_8b](https://gitee.com/deep-spark/deepsparkhub/tree/master/models/nlp/llm/llama3_8b/llamafactory)说明，执行训练
+4. 当然您也可以执行[这里](https://gitee.com/deep-spark/deepsparkhub#%E6%A8%A1%E5%9E%8B%E5%BA%93)的任何一个`IXUCA SDK 4.3.0`的模型进行训练
\ No newline at end of file
diff --git a/resources/accept-docker.png b/resources/accept-docker.png
new file mode 100644
index 0000000000000000000000000000000000000000..ea321de272f161fca036327687a181f093298ce0
Binary files /dev/null and b/resources/accept-docker.png differ
diff --git a/resources/complete-install.png b/resources/complete-install.png
new file mode 100644
index 0000000000000000000000000000000000000000..e2fda24fe1b4ee4503413c3004d19f0d6076ac18
Binary files /dev/null and b/resources/complete-install.png differ
diff --git a/resources/download-docker-installer.png b/resources/download-docker-installer.png
new file mode 100644
index 0000000000000000000000000000000000000000..fda1d769ff00821ed2303ba971cdc008cf908553
Binary files /dev/null and b/resources/download-docker-installer.png differ
diff --git a/resources/get-docker-installer.png b/resources/get-docker-installer.png
new file mode 100644
index 0000000000000000000000000000000000000000..bd955a2e1a2ae90b98765ac746fbabb011ac1961
Binary files /dev/null and b/resources/get-docker-installer.png differ
diff --git a/resources/install-docker.png b/resources/install-docker.png
new file mode 100644
index 0000000000000000000000000000000000000000..920260a932de93bef58f254e4fe268fd3901a824
Binary files /dev/null and b/resources/install-docker.png differ
diff --git a/resources/wget-docker-link.png b/resources/wget-docker-link.png
new file mode 100644
index 0000000000000000000000000000000000000000..591ab73fcc5b58c7c4c637ecd0ea2df1c04766f9
Binary files /dev/null and b/resources/wget-docker-link.png differ