Loading
A
Systems Performance Architect
AppleCupertino, California, United States$150,400 - $277,6003+ years
Apply Imagine what you could do here. At Apple, new ideas have a way of becoming products, services, and customer experiences very quickly. Every single day, people do amazing things at Apple! Do you want to impact billions of users by developing an extraordinary product with a prime focus on accuracy, understandability, and performance of the product? Dynamic, Inspiring people and Innovative technologies are the norm here.
The System Performance Architecture team is responsible for all aspects of performance spanning hardware (CPU, GPU, DRAM, storage), software (OS, drivers, applications), and other system characteristics (thermals, power, acoustics). This spans performance modeling, guiding architecture tradeoffs, performance verification, performance debug, and workload characterization.
Our team is collaborative, creative and passionate about what we do and the value we add in future product designs.
Come join us!
We have an outstanding opportunity for a Performance Architect to contribute to the characterization and optimization of future Apple Silicon platforms. This position is a multi-disciplinary engineering role encompassing computer system design, workload analysis, and performance modeling. The candidate will need the skills to dissect complex workloads (specifically AI and ML), identify architectural bottlenecks, and influence the definition of our next-generation SoCs to advance our products' performance, power, and thermal efficiency.
- Workload Characterization: Analyzing key applications and benchmarks to understand their behavior on current Apple Silicon hardware. Identifying instruction mix, memory access patterns, and compute requirements.
- Performance Modeling: Distilling key system interactions into high-level performance models to project performance for future microarchitectures.
- Trace & Tool Development: Defining and guiding new methodologies for capturing representative workload traces and developing analysis tools to visualize performance bottlenecks.
- Cross-Functional Collaboration: Interacting with OS, Compiler, and ML software teams to understand future software trends and guide hardware trade-offs.
- Validation: Correlating performance models against silicon measurements to continually improve accuracy.
- Debug: Assisting with complex performance debug activities on pre-silicon and post-silicon platforms.
- Bachelors Degree in EE or Computer Engineering and minimum 3 years of experience in Computer Architecture, performance analysis, or hardware modeling
- Understanding of CPU/GPU/NPU microarchitecture concepts (pipelines, caches, coherence, etc.)
- Experience in performance analysis and modeling
- Expert in Python scripting /C++ coding Skills
- Masters Degree in EE/Computer Engineering or equivalent of relevant working experience is preferred
- Experience with workload characterization tools (e.g., hardware performance counters, instruction tracers, profilers)
- Experience with GPU APIs, OpenGL, OpenCL, DirectX, Metal.
- Expertise in one or more focused areas (CPU/GPU, System Architecture, Compiler/SW Stack). Areas could include core/shader micro-architecture, texture pipeline, caching, or memory subsystem design for the SoC. System Areas examples could include IO, Storage, or DRAM Subsystem.
- Exposure to modeling in constrained environments (ie under thermal, power delivery, power management constraints)
- Ability to communicate complex technical concepts to diverse audiences across hardware and software teams