Fellow GPU Architect - Hardware​/Software Co-Design

The Role

We are seeking two Fellow GPU Architects - Hardware/ Software Co-Design to join our team to develop product roadmap for instinct GPU's, In this role, you'll drive roadmap definition and HW/SW co-design activities to deliver new capabilities of data center GPU's. You will work with customers to deeply understand AI/ML models and trends to customize instinct GPU architecture. Your work will straddle AI model research and hardware design to deliver state of the art GPU's that excel at AI model performance for inference and training.

You bring excellent communication, management, and presentation skills, demonstrated through technical publications, presentations, trainings, executive briefings, etc. You are highly adept at collaboration among top-thinkers and engineers alike, ready to mentor and guide, and help to elevate the knowledge and skills of the team around you. Adept at collaboration among top-thinkers and senior architects with strong interpersonal skills to work across teams in different geographies.

• Utilize extensive understanding of GPUs/ AI accelerators to identify key bottlenecks in AI workloads
• Maintain a continuous learning of latest trends in AI algorithms as well as inference and training frameworks
• Work closely with product management to study perf/tco, customer requirements to tailor GPU product architecture
• Define product features and capabilities, close architecture, and micro-architecture requirements, drive technical specifications for SoC and IP blocks to meet those requirements, and provide technical direction to execution teams for AMD's next generation Instinct GPU's
• Collaborate with architects, hardware engineers, and firmware engineers to understand the new features to be developed
• Work cross functionally with IP/Domain architects to identify and assess complex technical issues/risks and develop architectural solutions to achieve product requirements

• Experience with HIP, Triton, Cuda and mapping key Algorithms used in AI to hardware units like tensor cores
• Deep understanding of transformer models , deep intuition for model break down and parallelization techniques and trade off’s involved
• Experience analyzing CPU, GPU or System-level Micro-Architectural features to identify performance bottlenecks within different workloads
• Outstanding foundation in SoC architecture, with expertise in one or more of the following: CPU or GPU, Memory sub-system, Fabrics, CPU/GPU coherency, Multimedia, I/O subsystems, Clocks, Resets, Virtualization and Security
• Demonstrated expertise in power management microarchitecture, low power design and power optimization, along with power impact at architecture, logic design, and circuit levels

• Undergrad degree required. Bachelors or Masters degree in computer engineering/Electrical Engineering preferred.

This role is not eligible for visa sponsorship.

Benefits offered are described: AMD benefits at a glance.

AMD does not accept unsolicited resumes from headhunters, recruitment agencies, or fee-based recruitment services. AMD and its subsidiaries are equal opportunity, inclusive employers and will consider all applicants without regard to age, ancestry, color, marital status, medical condition, mental or physical disability, national origin, race, religion, political and/or third-party affiliation, sex, pregnancy, sexual orientation, gender identity, military or veteran status, or any other characteristic protected by law.

We encourage applications from all qualified candidates and will accommodate applicants’ needs under the respective laws throughout all the stages of the recruitment and selection process.

AMD may use Artificial Intelligence to help screen, assess or select applicants for this position. AMD’s “Responsible AI Policy” is available here.

This posting is for an existing vacancy.