NVIDIA is seeking computer architects to help design processor and system architectures that will enable compelling HPC performance, architecture and efficiency improvements. This role offers the opportunity to be part of technology development that impacts the future hardware roadmap for HPC in NVIDIA. If you are keen on analyzing, predicting, and improving scientific and compute oriented performance to its limits with tomorrow’s hardware and software, this is the place to be.

What you’ll be doing:

• Creating new analysis methodologies for automating performance analysis on NVIDIA systems.

• Develop tools to semi-automate bottleneck identifications for HPC applications.

• Learn about how applications are accelerated by GPUs and impact of various system design choices on their performance.

• Develop methods to quantify sensitivity of applications to architectural parameters.

What we need to see:

• You are pursuing a PhD or MS in CS, EE or CSEE.

• Background in computer architecture or parallel programming.

• Programming skills in C++/Java/Python.

• Familiarity with GPU computing (CUDA) is a plus.

• Past experience in scalable tool development is a plus.

NVIDIA has some of the most forward-thinking and hardworking people in the world working for us. Are you a creative and autonomous engineer with a real passion for computer architecture? If so, we want to hear from you!