GPU Memory Architect — Advanced Memory Systems & Equity

NVIDIA Corporation; Santa Clara (CA); On-site; USD 184,000 - 288,000; Full time

Job summary

NVIDIA Corporation in Santa Clara, CA, is seeking a motivated architect to advance GPU memory architecture, optimizing performance, power, area, and complexity across leading-edge silicon processes. You will develop architecture and micro-architecture, run performance simulations, analyze benchmarks, define experiments, and create architectural specifications.

Collaboration with industry partners to define software interfaces to the GPU is also required.

Qualifications

• Master degree or equivalent in Electrical Engineering, Computer Science, Computer Engineering or related field.
• PhD in computer architecture is a plus.
• 3+ years of meaningful work experience in GPU or CPU Architecture and development, especially interconnects and QoS.
• Memory hierarchy, Memory model & ordering, multifunction HW accelerator, software, virtualization, and security.
• Strong communication and interpersonal skills; ability to work in a dynamic, product-oriented, distributed team.
• Experience mentoring junior engineers and interns is a plus.
• Experience with hardware memory management unit, prefetching, accelerator IPs, memory subsystem hierarchy.
• Experience with PCIe/CXL, encryption, confidential compute, virtualization & security.

Responsibilities

• Developing architecture and micro-architecture to improve the state-of-the-art in GPU memory system, Memory management, Virtualization & security optimizing along the axes of performance, power efficiency, complexity, area, effort, and schedule.
• Participating in performance simulation of features to improve translation efficiency to access memory and io using protocol like PCIe.
• Implementing and maintaining high-level functional and performance models.
• Analyzing benchmarks, application workloads and performance simulation results to identify areas for microarchitecture optimizations.
• Defining and performing experiments to study the machine in action, presenting experiment results to the larger group and proposing mechanisms for improvement.
• Creating architectural specifications and customer-facing documents.
• Working with partners in the industry to generate specifications which consider software interfaces to the GPU.
• Debugging performance and functional issues with high-level models, RTL simulation, silicon & systems.

Skills

GPU Architecture

Interconnects

QoS

Memory hierarchy

Memory modeling

Virtualization

Security

Education

Master's degree or equivalent (EE/CS/CE)

PhD in computer architecture (preferred)

Tools

PCIe/CXL

Job description

NVIDIA Corporation in Santa Clara, CA, is seeking a motivated architect to advance GPU memory architecture, optimizing performance, power, area, and complexity across leading-edge silicon processes. You will develop architecture and micro-architecture, run performance simulations, analyze benchmarks, define experiments, and create architectural specifications.

Collaboration with industry partners to define software interfaces to the GPU is also required.