Platform Emulation Software Engineer - DC GPU

Advanced Micro Devices, Inc•Markham, ON

3d•Hybrid

About The Position

AMD is powering the next generation of supercomputing, high-performance computing, cloud, and AI. Whether you’re designing next-gen processors, enabling AI breakthroughs, or creating go-to-market plans, every role at AMD contributes to something bigger — technology that moves the world forward. The Data Center GPU and Accelerated Processing Team delivers AMD products for Data Center, Machine Learning, and High-Performance Computing. The Platform Emulation team (PEMU) is an integral part of the Data Center GPU and Accelerated Processing group responsible for pre-silicon activities to support product bring-up and launch. AMD is seeking a Platform Emulation Software Engineer to join our Data Center GPU organization. Our products support the rapidly scaling Data Center and High-Performance Compute infrastructure. You will be an integral member of the Platform Emulation team responsible for designing and implementing GPU applications in CUDA/C++ to enhance pre-silicon verification, writing tools to improve the efficiency of debugging hardware on emulation, building and executing GPU benchmarks/applications on the emulator, and developing emulation infrastructure and tools. You will work alongside a team of innovative engineers to support the deployment of AMD’s Instinct ML products targeting Supercomputers and Data Center workloads.

Requirements

Strong knowledge of computer hardware architecture (GPU/CPU, memory hierarchy, interconnects, caches)
Excellent programming skills in: C, C++, Python, Tcl
Shared memory concurrent programming
Relaxed memory models
Cache coherency
Working knowledge of Linux/Unix environments and shell scripting
Knowledge of computer software architecture and boot flow (boot code, BIOS, device drivers, OS)
Excellent oral and written communication skills
Willingness to learn and think outside the box

Nice To Haves

Exposure to Verilog/SystemVerilog and waveform-based debug
Experience with ML workloads and profiling/performance analysis
Familiarity with APIs/platforms such as ROCm, OpenCL, OpenGL, Vulkan

Responsibilities

Design and implement GPU applications in CUDA/HIP to enhance pre-silicon verification
Develop tools and automation to improve the efficiency of debugging hardware on emulation
Run and collect data for analysis on AMD’s high-end emulators and simulation models
Develop scripts/tools to parse and analyze data from emulation runs
Run and collect functional and performance data for AI/ML workloads
Collaborate with senior engineers to support debug of hardware-related failures and performance issues observed on emulation
Attend weekly meetings, provide status communication, and deliver technical presentations