Deep Learning Software Engineer, TensorRT Performance - New College Grad 2027

at Nvidia

USD 124,000-241,500 per year

JUNIOR

✅ On-site

Tech Stack
Tag name is followed by "@" symbol and proficiency level value. About proficiency levels:

1-2 — basic awareness. Minimal hands-on experience, and a rudimentary understanding of the technology's purpose;

3-6 — daily use. Comfortable and regular usage, capable of handling common tasks and challenges related to the technology;

7-9 — you are an expert, you can teach others, you know all the pitfalls and tricks;

10 — exceptional knowledge, comprehensive understanding, and adeptness in all aspects of the technology, including advanced problem-solving. Think twice before claiming or demanding such level.

AI @ 3 Algorithms CUDA @ 5 Deep Learning @ 3 GPU GenAI Generative AI @ 3 JAX @ 3 LLM @ 3 Performance Analysis @ 3 PyTorch @ 3 Python @ 6 Robotics @ 3 SGLang @ 3 Software Development @ 3 TensorFlow @ 3 TensorRT @ 3 vLLM @ 3

Details

NVIDIA is seeking a Deep Learning Software Engineer to analyze and improve the performance of its inference ecosystem. The role focuses on developing GPU-accelerated deep learning inference software, including TensorRT, deep learning benchmarking software, and solutions for deploying and serving deep learning models.

The engineer will collaborate with the deep learning community to integrate TensorRT into open-source frameworks such as TensorRT-EdgeLLM and PyTorch, optimize state-of-the-art models across NVIDIA accelerators, and implement graph compiler algorithms, frontend operators, and code generators. The role involves collaboration across workflow improvement, performance modeling, performance analysis, kernel development, and inference software development teams.

Responsibilities

Establish performance benchmarking methodologies and analysis workflows for NVIDIA's inference ecosystem, including TensorRT, TensorRT-EdgeLLM, and Torch-TensorRT.
Identify performance issues and optimization opportunities.
Contribute features and code to NVIDIA and open-source inference frameworks.
Develop optimized model pipelines involving quantization, scheduling, memory management, and distributed inference.
Collaborate with teams working in generative AI, automotive, robotics, image understanding, and speech understanding.
Scale the performance of deep learning models across different NVIDIA architectures and accelerator types.

Requirements

Bachelor's, Master's, or PhD degree, or equivalent experience, in Computer Science, Computer Engineering, EECS, AI, or a related field.
Two years of relevant software development experience.
Strong C++ and Python programming and software engineering skills.
Experience with deep learning frameworks such as PyTorch, JAX, TensorFlow, or ONNX.
Experience with inference libraries such as TensorRT, TensorRT-LLM, vLLM, SGLang, or FlashInfer.
Experience with performance analysis and optimization.

Preferred Qualifications

Strong foundation and architectural knowledge of GPUs.
Deep understanding of modern deep learning models and workloads, including Transformers, recommenders, automatic speech recognition, text-to-speech, and visual understanding.
Proficiency in CUDA, TileIR, CuTeDSL, CUTLASS, Triton, or another deep learning domain-specific programming language.
Contributions to major large language model inference frameworks such as vLLM.
Experience with deep learning inference graph compilers such as TorchDynamo or TorchInductor.
Experience optimizing low-latency, resource-constrained systems or embedded AI pipelines, such as Jetson systems or other edge AI accelerators.

Compensation and Benefits

The base salary range is USD 124,000–195,500 for Level 2 and USD 152,000–241,500 for Level 3. Compensation is determined based on location, experience, and the pay of employees in similar positions. The role also includes eligibility for equity and benefits.

Applications will be accepted at least until October 16, 2026. NVIDIA is an equal opportunity employer.