At NVIDIA, our NVLink firmware powers the high-speed interconnect fabric at the
heart of the largest multi-GPU systems globally — the backbone of up-to-date AI and
high-performance computing. Our Bangalore firmware development and validation
team is responsible for this firmware end to end. They handle everything from
feature creation to pre-silicon testing, silicon bring-up, and automation that
keeps progress fast. This is firmware engineered at the edge of what the
hardware allows — within tight boot-time and memory-footprint budgets where
every byte counts. We are looking for a top-tier firmware engineer who is
motivated by the full craft — writing the firmware, proving it accurate,
crafting the tooling and infrastructure strategy that makes the whole team
faster, and setting the technical bar for everyone around them. The work spans
next-generation GPU platforms and puts you shoulder to shoulder with
architecture, hardware, and build teams across our Bangalore and American
locations. If you love low-level software, thrive on debugging the hardest
problems,
and want your work to ship in the systems that define the industry,
we’d love to meet you.
What you’ll be doing:
* Own the technical architecture and roadmap for NVLink and high-performance
interconnect firmware across next-generation GPU platforms — staying hands-on
in C, C++, and Python on the highest-leverage problems.
* Take end-to-end ownership of sophisticated interconnect features, driving
seamless execution from design through final verification.
* Architect correctness models, verification plans, and test cases — primarily
in Python — to prove firmware functionality across both pre-silicon and
post-silicon environments.
* Shape silicon bring-up, design-for-debug, and program-level validation
strategy with architecture, hardware, and design leads across Bangalore and
US sites.
* Define the verification methodology, engineering standards, and quality gates
the team builds to — and the automation strat