Characterize and tune controller behaviour across workloads, operating conditions, silicon variation, and product power and thermal limits. Maximize the perf and Perf/Watt with-in architectural bottlenecks & limitations, Power/Thermal system limits to provide best customer experience for the products
Correlate silicon behaviour with pre-silicon expectations, debug & resolve complex problems during bring up and final closed loop performance stage involving silicon HW, firmware, software, power delivery network, system thermals etc.
Apply AI/ML and data-driven methods to analyze silicon characterization and HW telemetry data, identify anomalies and trends, accelerate controller tuning and improve debug as well as silicon validation execution efficiency.
What we need to See
B. Tech or M. Tech in Electronics Engineering stream.
At least 5+ years of proven experience in silicon validation, system level design, and power/performance optimization.
Strong understanding of aspects related to silicon power and performance, technology node impacts, Hardware and Software interactions at system level.
Hands-on experience with silicon bring up, validation and productization
Good HW architectural understanding and orientation
Very good problem solving and hardware debugging skills.
Strong good data analysis and logical reasoning skills.
Strong familiarity with HW lab environment.
Must be a great teammate and ready to collaborate with global teams from diverse cultural backgrounds in a high energy environment.
Ways to stand out from the crowd
Extensive background in DVFS, voltage/frequency management, HW telemetries, Performance throttling, silicon protection mechanisms on advanced ASIC/SoC products.
Contributions that improved controller behaviour, silicon performance, power efficiency, thermal management, bring-up/productization methodologies.
Experience connecting controller behaviour across simulation, silicon, firmware, and system-level measurements and converting findings into production tuning.
Experience applying AI/ML to engineering workflows, such as controller tuning, characterization analytics, anomaly detection, or debug assistants deployed beyond personal use.
With every new architecture and process node, balancing performance against power and thermal limits becomes tougher. Our controllers are at the forefront of this issue: drawing out the performance the silicon was crafted to achieve while shielding it when workloads push its limits. If you like understanding dynamic silicon behaviour, addressing hardware-software challenges, and influencing NVIDIA products at scale, we'd love to connect!
#LI-Hybrid