为突破传统计算机专业核心课程体系存在知识前置与教学时序的结构性矛盾问题，提出“极小化实现”教学范式，具体阐述如何设计图灵完备的12指令原型机VSPM，通过硬件实现与软件模拟双路径，构建冯·诺依曼体系的可视化载体；开发轻量化编译工具链miniCC，支持包括整型变量与控制结构及简单I/O操作的C语言子集编译，以28个渐进式实验关卡实现从“源码→四元式→汇编指令”的透明化转换，通过教学实践说明该方案在教育部“101计划”平台及头歌实训环境中成功建立跨抽象层级的可观测链路，使低年级学生在无先修知识条件下贯通理解指令集、体系结构、内存寻址等核心机制，能够有效弥合基础认知与专业深化间的断层。

In the training of computer system capabilities, it is crucial to develop an early understanding of the entire process from high-level programming languages to hardware execution. However, the traditional curriculum often faces a fundamental conflict between the knowledge required beforehand and the order in which it is taught. To address this issue, the paper introduces a teaching approach called "minimal implementation." It starts by creating a Turing-complete prototype machine, VSPM, with 12 instructions, which serves as a visual representation of the von Neumann architecture through both hardware and software simulations. Additionally, a streamlined compilation tool chain, miniCC, is developed to support a subset of the C language, including basic I/O operations, control structures, and integer variables. This is achieved through 28 progressive experiments that transparently transform source code into quadruples and then into assembly instructions. This approach has been successfully implemented in the Project 101 platform and the TouGe training environment, allowing students to grasp essential concepts like instruction sets, system architecture, and memory addressing without prior knowledge, effectively bridging the gap between foundational understanding and specialized learning.