GPU Performance Modeling Engineer

Summary

Do you love creating elegant solutions to highly complex challenges? Do you intrinsically see the importance in every detail? As part of our Silicon Technologies group, you’ll help design and manufacture our next-generation, high-performance, power-efficient GPU! You’ll ensure Apple products and services can seamlessly and efficiently handle the tasks that make them beloved by millions. Do you love working on challenges that no one has solved yet? Join us to help deliver the next groundbreaking product containing an Apple designed GPU. In this team, we are passionate about writing code for performance modeling for GPUs and the performance model infrastructure. This includes working with architects and designers to understand the units being modeled.

Minimum Qualifications

We are looking for candidates with a BS and a minimum of 3 years relevant industry experience. Background in GPU, CPU, or SIMD architectures. Excellent C/C++ programming skills. Experience in hardware performance modeling, analysis, and debug. Ability to work well in a team in order to be productive under aggressive product schedules. We believe that there is power in teams working together so you must have strong communication, influence, and negotiation skills.

Preferred Qualifications

Experience writing performance models for GPU Schedulers, Execution Units, Texture/Pixel Units, or Caches. Experience with GPU APIs, OpenGL, OpenCL, DirectX, Metal. Expertise in one or more focused CPU/GPU areas. Areas could include core/shader micro-architecture, texture pipeline, caching, or memory subsystem design. Experience with scripting, Verilog/VHDL, design verification methodology and tools a plus.

Description

As a GPU performance modeling engineer, you will be responsible for developing cycle-approximate perf C/C++ models in close collaboration with architects and designers. Additionally, you will analyze performance results and help propose architectural and micro-architectural changes. You will be critical in supporting the hardware team during model and design performance bring-up.