System Software Engineer, Performance - CUDA Driver - CA, Santa Clara, United States
$124,000 - $195,500 USD yearly
Originally posted 1 September 2026 by the employer.
Work on the CUDA driver and runtime.
About the role
This role involves designing and shipping production C/C++ features and optimizations in the CUDA driver and runtime, tracing important workloads, bringing up new platforms, and informing future software and hardware direction.
What you'll do
- Design, implement, validate, and ship performance-centric features and programming-model capabilities in the CUDA driver and runtime.
- Optimize critical execution paths, including kernel launch, synchronization, memory management & movement, CPU–GPU coordination, and system interconnect use.
- Own complex performance problems end-to-end, understanding workloads, forming hypotheses, creating measurements and models, isolating root causes, and implementing production solutions.
- Establish performance expectations for current and future platforms, characterize new silicon, close software and hardware gaps, and drive performance readiness.
- Translate workload and platform evidence into CUDA API and programming-model improvements, systems-software direction, and recommendations for future hardware architecture.
- Partner with application, library, framework, operating-system, driver, runtime, firmware, GPU architecture, silicon, product, and customer-facing teams.
What you'll need
- BS, MS, or PhD in Computer Science, Computer Engineering, Electrical Engineering, or a related field—or equivalent practical experience — with at least 2 years of relevant systems-software development experience.
- Strong production C/C++ systems-programming experience, including delivery of substantial features, optimizations, or production fixes in a complex codebase.
- Strong operating systems and concurrency foundations, including threads, synchronization, processes, virtual memory, and user/kernel interactions.
- Strong computer-architecture foundations, including processors, memory hierarchy, caching and coherence, data movement, and system interconnects.
- Demonstrated success improving real software performance: measuring behavior, identifying the limiting mechanism, implementing an effective solution, and validating the result quantitatively.
- Sound technical judgment, ownership of ambiguous problems, and clear communication across organizational and disciplinary boundaries.
Skills: CUDA driver, runtime, C/C++, GPU, performance optimization, systems software
This role has been open 43 days — well below the 56-day median for Infrastructure/Platform Software roles.
Infrastructure/Platform Software · GPU Software Stack
|
Open roles in category
231
|
Median days open
56 d
|
Median salary
$236k
|
See the full market breakdown ▾Category comparison, skills in demand, and who else is hiring
| Metric | NVIDIA | All employers we track in this specialty (231 roles · 20 employers) |
|---|---|---|
| Open roles in this specialty | 114 | 231 |
| Open roles in the wider Software, Firmware & Systems family | 993 | 6140 · 145 employers |
| Median days open | 55 d | 56 d (−1 d vs this employer) |
| Median salary (USD postings) | — | $236k |
Skills observed across this category: CUDA driver, runtime, C/C++, GPU, performance optimization, systems software
Who's hiring in this category
- NVIDIA (this employer) · 117 open roles · median 55 d
- AMD · 53 open roles · median 65 d
- Qualcomm · 19 open roles · median 106 d
- Intel Corporation · 17 open roles · median 6 d
- Arm Holdings · 5 open roles · median 112 d
- d-Matrix · 2 open roles · median 120 d
How we counted: 231 open Infrastructure/Platform Software (GPU Software Stack) roles from 20 employers tracked in the SemiconductorJobs index, counted 15 Oct 2026. Specialty figures count only roles carrying this exact specialty label, so an employer's related work in neighbouring specialties is not included there — it is counted in the wider Software, Firmware & Systems family row. Figures refresh nightly.