Matching Items (1)
Description
This project's goal was to design a Central Processing Unit (CPU) incorporating a fairly large instruction set and a multistage pipeline design with the potential to be used in a multi-core system. The CPU was coded and synthesized with Verilog. This was accomplished by building on the CPU design from fundamentals learned in CSE320 and increasing the instruction set to resemble a proper Reduced Instruction Set Computing (RISC) CPU system. A multistage pipeline was incorporated to the CPU to increase instruction throughput, or instructions per second. A major area of focus was on creating a multi-core design. The design used is master-slave in nature. The master core instructs the sub-cores where they should begin execution, the idea being that the operating system or kernel will be executing on the master core and the "user space" programs will be run on the sub-cores. The rationale behind this is that the system would specialize in running several small functions on all of its many supported cores. The system supports around 45 instructions, which include several types of jumps and branches (for changing the program counter based on conditions), arithmetic operations (addition, subtraction, or, and, etc.), and system calls (for controlling the core execution). The system has a very low Clocks per Instruction ratio (CPI), but to achieve this the second stage contains several modules and would most likely be a bottleneck for performance if implemented. The CPU is not perfect and contains a few errors and oversights, but the system as a whole functions as intended.
ContributorsKolden, Brian Andrew (Author) / Burger, Kevin (Thesis director) / Meuth, Ryan (Committee member) / Computer Science and Engineering Program (Contributor) / Barrett, The Honors College (Contributor)
Created2016-05