AI Software Technical Intern at Intel in Santa Clara, CA

Job description

Job Details: Job Description: Role Overview We are seeking a driven and curious Technical Intern to join our AI Software and Hardware Architecture team. In this role, you will work at the intersection of next-generation GPU architecture, performance simulation, and modern AI engineering. Our team develops architectural models and virtual platforms (VPs) to project, evaluate, and optimize the performance of future high-performance computing (HPC) and AI GPU architectures before silicon is taped out. As part of this mission, you will actively integrate the latest generative AI capabilities-including GitHub Copilot, custom Large Language Model (LLM) agents, and developer tooling integrations-directly into our architectural modeling, automated testing, and performance analysis pipelines. This internship offers an opportunity to gain hands-on experience with production-scale architectural simulators while pioneering modern, AI-augmented hardware engineering practices. Key Responsibilities Virtual Platform Development and Architecture Simulation Collaborate with senior architects to develop, calibrate, and validate cycle-approximate or cycle-accurate Virtual Platform (VP) models for future GPU architectures (compute engines, memory fabric, cache hierarchies, and interconnects). Implement functional and performance simulation components in modern C++ and SystemC/TLM frameworks. Assist in porting, configuring, and executing graphics and compute workloads (e.g., PyTorch, SYCL, OpenCL, Vulkan/DirectX) across simulated environments. AI-Augmented Workflow Integration and Automation Accelerate development and debugging efficiency by leveraging GitHub Copilot and state-of-the-art LLM capabilities across team repositories. Prototype and deploy agentic AI workflows and tools (e.g., Model Context Protocol [MCP] integrations, automated log and trace summarization, error diagnosis). Build AI-assisted test generation tools and validation scripts using Python and unit testing frameworks (e.g., GoogleTest). Performance Analysis and Telemetry Collect, profile, and analyze trace-driven simulation data to pinpoint memory subsystem bottlenecks, cache miss penalties, and pipeline stalls. Create automated visualization dashboards and reporting scripts to present performance trade-offs to senior architecture and design teams. Investigate anomaly detection in large-scale simulation telemetry using machine learning and statistical analysis techniques. Qualifications: Minimum Qualifications Currently enrolled in an accredited Master's or Ph.D. program in Computer Science, Computer Engineering, Electrical Engineering, or a closely related discipline. Strong proficiency in C++ (modern C++17/20 preferred) and Python. Solid foundational knowledge of computer architecture principles (cache hierarchies, memory coherence, pipelining, and vector/SIMD processing). Experience with Linux/Unix environments, version control systems (Git), and modern build pipelines (CMake, Ninja). Familiarity with utilizing AI developer tooling (e.g., GitHub Copilot, Anthropic APIs, or local LLM runtimes) for software engineering. Preferred Qualifications Prior coursework or project experience in: GPU architecture, parallel programming, or accelerated computing (CUDA, SYCL, OpenCL, or GPU compute APIs). Virtual Platforms, architectural simulators (e.g., gem5, SystemC/TLM, or proprietary simulator platforms). Software testing and test-driven development (TDD) using frameworks like GoogleTest (gtest). Experience building LLM-assisted developer agents, prompt engineering pipelines, or custom developer automation tooling. Familiarity with hardware/software co-design concepts and performance modeling methodologies. What we offer Mentorship from industry-leading Principal Engineers and Architects in GPU architecture and AI system design. Exposure to pre-silicon hardware validation methodologies used across world-class compute products. An innovative environment where you are encouraged to expe