Performance Modeling Software Architect, Platform Architecture

Apple•Austin, TX

14h

About The Position

As a member of Apple’s system-on-chip (SoC) performance team within the Hardware Technologies organization, you will help shape the architecture of future devices. The team collaborates with SoC architecture, design, power, and software groups to develop caches and memory systems that enable CPUs, GPUs, machine learning, camera, display, and connectivity components to work together seamlessly. You will work on performance models and infrastructure to produce high-quality, timely data that informs hardware design. The role involves translating design-space studies and hardware execution models into well-written, high-performance C++ software models, and contributing across the performance-modeling lifecycle, from early architectural exploration through post-silicon correlation. You will help ensure the simulator is performant, maintainable, flexible, and supports rapid experimentation by hardware architects. The role also involves developing APIs for non-expert C++ developers; training AI agents and developers in idiomatic code and methods; and structuring the codebase for effective AI-assisted development. You will write clear, self-documenting code with consistent conventions and well-organized modules. Additional work includes learning about SoC IP and chip-level architecture, building data-analysis pipelines and visualization methods, continuous regression testing, running performance simulations, and analyzing and presenting simulation results.

Requirements

Bachelor’s degree.
Experience with hardware architecture and design issues.
Experience coding in C++ and Python, including binding C++ to other languages.
Experience with computer architecture and software engineering principles.
Experience with data structures and algorithms.
Experience using software models to collect behavioral characteristics and transforming that data into compelling performance analysis.

Nice To Haves

MS or PhD in computer science, electrical engineering, or a related field.
20+ years of relevant experience.
Experience with multi-paradigm software development in C++.
Experience with multi-language programming environments and selecting the right tool or language for the job.
Experience with Compiler Explorer and understanding the transformations a compiler performs on code.
Experience treating code as something read by both humans and machines, and establishing conventions, structure, and documentation that make a codebase tractable for AI-assisted reasoning and navigation.
Experience with architectural and micro-architectural performance modeling.

Responsibilities

Develop performance models and infrastructure that provide high-quality, timely data for hardware design.
Translate design-space studies and hardware execution models into well-written, high-performance C++ software models.
Contribute to the full performance-modeling lifecycle, from early architectural exploration to post-silicon correlation.
Improve simulator performance, maintainability, and flexibility so hardware architects can iterate quickly.
Develop APIs usable by non-expert C++ developers.
Train AI agents and non-expert C++ developers on idiomatic code and methods.
Structure the codebase for effective use with AI-assisted development tools.
Write clear, self-documenting code using consistent conventions, meaningful naming, and well-organized modules.
Learn about state-of-the-art SoC IP and chip-level architecture.
Work on data-analysis pipelines, visualization methodologies, and continuous regression testing.
Run performance simulations.
Analyze and present simulation results.