Lead Runtime Engineer

WHAT YOU DO AT AMD CHANGES EVERYTHING At AMD, our mission is to build great products that accelerate next-generation computing experiences—from AI and data centers, to PCs, gaming and embedded systems. Grounded in a culture of innovation and collaboration, we believe real progress comes from bold ideas, human ingenuity and a shared passion to create something extraordinary. When you join AMD, you’ll discover the real differentiator is our culture.

We push the limits of innovation to solve the world’s most important challenges—striving for execution excellence, while being direct, humble, collaborative, and inclusive of diverse perspectives. Join us as we shape the future of AI and beyond. Together, we advance your career.

THE ROLE:

AMD is searching for talented and motivated individuals to support and develop the GPU‑compute language runtime libraries of the AMD ROCm™ platform, with a focus on Windows graphics and compute driver enablement. As part of the Language Runtime group in AMD AI GPU Software, you will build core compute runtime components powering artificial intelligence and high‑performance computing applications across workstations, data centers, and supercomputers.

This role involves deep technical collaboration with customers and partners to understand requirements, resolve driver‑level challenges, and ensure successful deployment of AMD GPU technologies on Windows. THE PERSON:
You are passionate about Windows GPU driver development, kernel‑mode engineering, and building high‑performance runtime systems. You thrive when working close to the metal—solving complex problems in memory management, scheduling, concurrency, and GPU compute execution. You approach challenges with a performance‑first mindset, using profiling, instrumentation, and data analysis to guide improvements. You collaborate effectively across cross‑functional teams and communicate transparently with customers.

You are detail‑oriented, proactive, and motivated to push the boundaries of GPU compute runtime performance on Windows.

KEY RESPONSIBILITIES:

Design, develop, and optimize GPU language runtime components for Windows within the AMD ROCm™ platform Implement and enhance runtime functionalities that interface with Windows kernel‑mode and user‑mode GPU drivers Conduct deep performance analysis using profiling tools, instrumentation, and hardware counters Identify, diagnose, and eliminate performance bottlenecks across driver, runtime, and compute pathways Work with AMD architecture teams to influence future GPU hardware and software features with a focus on performance scalability Participate in new ASIC and hardware bring‑ups, ensuring runtime and driver compatibility, functionality, and peak performance Debug and resolve complex runtime, compute pipeline, and GPU‑driver‑related issues, including customer‑reported problems Interface directly with customers and partners to understand requirements, troubleshoot issues, and optimize performance in real‑world workloads Maintain strong technical relationships with internal teams, external partners, and customers Stay current on Windows graphics subsystems, GPU driver models, runtime technologies, and performance best practices

PREFERRED EXPERIENCE:

Strong systems programming skills with proficiency in C/C++ Experience developing or debugging Windows GPU kernel‑mode drivers (WDDM, KMDF/WDF) Familiarity with Windows Display Driver Model (WDDM), DXGI, Direct3D, compute driver components, or command submission pipelines Hands‑on experience with performance profiling tools (Win Dbg, GPUView, WPA, ETW tracing, PIX, vendor profilers) Demonstrated ability to perform deep‑dive debugging and optimization using hardware counters, scheduling analysis, and memory utilization metrics Background in developing or optimizing runtime systems such as compute runtimes, language runtimes, device runtimes, or driver‑adjacent software

Experience with concurrency, synchronization primitives, and multi‑threaded performance tuning Strong understanding of Windows OS internals, memory models, kernel/user transition costs, and driver framework best practices Proficiency with development tools such…