LLM for software optimization

About

Table of Contents

• Environment Requirement
• Reproduce Results
• Running the pipeline
• Analysis and evaluation
• Code Dependencies

Environment Requirement

This artifact requires a machine with the following capabilities to support RAPL (Running Average Power Limit) and read MSR (Model-Specific Registers):

1. Hardware

• Intel Processor: Machine with Intel processors supporting RAPL (Sandy Bridge or newer).
• MSR Support: Machine must allow access to MSRs.

2. Operating System

• Linux-based OS (e.g., Ubuntu 16.04+).
• Linux Kernel Version 3.13+ required for RAPL support.
• Root Access: MSRs can only be accessed with root/superuser privileges.

3. Software

• msr-tools: Install for reading MSRs:

  sudo apt-get install msr-tools
  

Reproduce Results

To set up the pipeline for energy-efficient code optimization, follow these steps:

1. Clone the repository:

   git clone <repository-link>
   cd <project-directory>

2. Install the required dependencies using the Makefile

   make setup

3. Create .env in root directory This should include:

   API_KEY=your_openai_api_key_here
   USER_PREFIX=/path/to/E2COOL

4. Update RAPL/main.c write path Change the path to language .csv file to match your absolute path

   strcpy(path, “ABSOLUTE_PATH/E2COOL/src/runtime_logs/");

   Then run make in RAPL directory

   make

Running the pipeline

5. Run the main script in the home directory (/E2COOL)

   make run ARGS="--benchmark PIE --llm gpt-4o --self_optimization_step 2 --num_programs 2"

   To use GenAI studio to inference open-source models

   make run ARGS="--benchmark PIE --llm llama3.3:70b --self_optimization_step 2 --num_programs 2 --genai_studio True"