Job Description

Sofware System Engineer - NVIDIA CUDA & Kernel Development

Company: - US Based Tech Startup

Location:

Bangalore, India

Experience:

6 + Years

Looking for Early Joiner

Job Summary:

We are looking for an experienced

System Engineer

with deep expertise in

C programming, Linux kernel internals

, and the

NVIDIA CUDA stack

. The ideal candidate will have hands-on experience working with

GPU kernel drivers

,

CUDA runtime

,

memory management

, and

CUDA kernel task execution frameworks

. This is a highly technical role requiring a solid understanding of GPU architecture and low-level system programming.

Key Responsibilities:

Develop and integrate system-level features around

NVIDIA GPU drivers

and the

CUDA stack

. Work closely with Linux kernel modules, device drivers, and user-space libraries. Design and debug GPU memory access, allocation, and

CUDA Unified Memory

implementation. Optimize GPU compute workloads, kernel launches, and task scheduling. Collaborate with cross-functional teams to support system stability and performance tuning.

Required Skills & Experience:

6+ years of hands-on experience in

NVIDIA GPU development

and

CUDA programming

. Strong C programming skills, especially in

system-level or kernel-space development

. Solid understanding of:

CUDA driver/runtime API

GPU kernel drivers

CUDA memory access patterns & optimization

Unified Memory framework

Kernel task execution models

Good debugging skills across user and kernel space (gdb, dmesg, trace tools). Exposure to performance benchmarking and profiling tools (e.g., nvprof, Nsight).

Nice to Have:

Experience with AI/ML workloads, HPC applications, or real-time systems. Contributions to open-source GPU or Linux kernel projects. Familiarity with other GPU platforms like AMD ROCm.
Job Type: Full-time

Pay: ?1,686,349.11 - ?3,095,710.16 per year

Ability to commute/relocate:

Bangalore, Karnataka: Reliably commute or willing to relocate with an employer-provided relocation package (Preferred)
Education:

Bachelor's (Preferred)
Experience:

Systems engineering: 6 years (Preferred) Software development: 6 years (Preferred) NVIDIA CUDA: 2 years (Preferred) C programming: 5 years (Preferred) Linux kernel internals: 3 years (Preferred) GPU architecture : 3 years (Preferred) low-level system programming.: 5 years (Preferred)
Work Location: In person