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

Job Title

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

Role Summary

Develop and optimize NVIDIA's deep learning inference stack (TensorRT, TensorRT-EdgeLLM, Torch-TensorRT) to improve model performance across datacenter GPUs and edge SoCs. Work on benchmarking, performance analysis, graph compiler algorithms, frontend operators and code generation for inference workloads.

This role collaborates with internal and external teams to integrate and scale inference solutions for generative AI, recommender systems, vision, speech and embedded use cases.

Experience Level

Entry-level / New College Grad. Typical experience guidance: 0–3 years; this posting requests ~2 years of relevant software development experience.

Responsibilities

Primary responsibilities include:

• Design and implement performance benchmarking methodologies and analysis workflows for NVIDIA inference software.
• Contribute code and features to NVIDIA and open-source inference frameworks (TensorRT, TensorRT-EdgeLLM, Torch-TensorRT, etc.).
• Develop and optimize model pipelines focusing on quantization, scheduling, memory management, and distributed inference.
• Analyze performance issues and identify optimization opportunities across GPU and edge accelerator architectures.
• Collaborate with cross-functional teams (workflow, kernel, modeling, product teams) to drive inference performance improvements.
• Scale model performance across different NVIDIA accelerators and deployment targets.

Requirements

Must-have qualifications:

• Approximately 2 years of relevant software development experience.
• Strong C++ and Python programming and software engineering skills.
• Experience with deep learning frameworks (PyTorch, JAX, TensorFlow, ONNX) and inference libraries (TensorRT, TensorRT-LLM, vLLM, etc.).
• Experience performing performance analysis and implementing performance optimizations.

Nice-to-have:

• Strong understanding of GPU architecture and low-level performance considerations.
• Deep knowledge of modern DL models and workloads (Transformers, recommenders, ASR, TTS, vision).
• Experience with domain-specific languages and libraries for DL acceleration (CUDA, Triton, Cutlass, TileIR, CuTeDSL).
• Prior contributions to LLM inference projects or experience with graph compilers (TorchDynamo, TorchInductor) and LLM frameworks (vLLM).
• Experience optimizing low-latency or resource-constrained/embedded AI systems (e.g., Jetson or other edge accelerators).

Education Requirements

Bachelor's, Master's, or PhD in Computer Science, Computer Engineering, Electrical Engineering/CS (EECS), AI, or a closely related field - or equivalent practical experience.

About the Company

Company: NVIDIA

Headquarters: Santa Clara, California, USA

NVIDIA is a global leader in accelerated computing, renowned for its innovative solutions in AI and digital twins that transform diverse industries. The company specializes in networking technologies, providing end-to-end InfiniBand and Ethernet solutions for servers and storage that optimize performance and scalability. NVIDIA serves sectors such as high-performance computing, enterprise data centers, and cloud computing, constantly reinventing its products and services to stay ahead in the market.