Overview As a Computer Architect at NVIDIA, you will shape the memory subsystem and energy-efficient architectures for future high-performance processors. You will work across the team to push perf/W and perf/mm, applying PPA analysis to drive impactful design choices. You'll develop proto models for power/noise, and evaluate performance and power through simulations and benchmarks. This role blends deep architectural work with collaboration across partners to scale cutting-edge HPC GPUs. You will contribute to a mission-driven, fast-moving environment that values innovation and impact.
Compensation / Benefits Responsibilities- Design high-performance processor and system architectures with a focus on memory systems and energy efficiency
- Advance architecture and micro-architecture features to improve GPU memory performance metrics (perf/W, perf/mm, perf/$)
- Develop and refine architecture prototype models for power and noise analysis
- Perform performance and power simulations to optimize energy per byte
- Analyze benchmarks and workloads to identify architectural optimizations
- Debug power, performance, and functional issues across models, RTL, silicon, and systems
- Collaborate with external partners on system infrastructure
Key requirements- 10+ years in CPU/GPU architecture with memory systems design and energy efficiency focus
- Bachelors in Electrical Engineering, Computer Science, or related field
- Expertise in power analysis and modeling in pipeline-based architectures
- Familiarity with power analysis tools such as Ansys PowerArtist and Synopsys PT PX
- Experience with C, C++, and scripting languages
- excellent communication
- team collaboration
- problem solving
- memory systems architecture
- power analysis and modeling
- pipeline-based architectures
