LLM inference • 大模型推理系统优化:通过系统手段来提升TensorRT-LLM的性能、易用性和功能丰富度。包括但不限于TensorRT-LLM的API改进、LLM runtime优化、服务层优化、定制优化算子、分布式优化等等。 • 大模型推理模型优化: 通过模型优化的手段进一步提升TensorRT-LLM的性能表现，包括低比特量化，KV Cache量化，稀疏化等等，也包括类似于Speculative decoding, Streaming-LLM这类模型系统联合优化的工作内容。 AI Complier System • AI编译器: 主要包括如何使用AI编译器技术，在NV的当代硬件，以及未来硬件上，以自动化的方式生成高效的计算密集/访存密集的算子kernel。

LLM inference • 大模型推理系统优化:通过系统手段来提升TensorRT-LLM的性能、易用性和功能丰富度。包括但不限于TensorRT-LLM的API改进、LLM runtime优化、服务层优化、定制优化算子、分布式优化等等。 • 大模型推理模型优化: 通过模型优化的手段进一步提升TensorRT-LLM的性能表现，包括低比特量化，KV Cache量化，稀疏化等等，也包括类似于Speculative decoding, Streaming-LLM这类模型系统联合优化的工作内容。 AI Complier System • AI编译器: 主要包括如何使用AI编译器技术，在NV的当代硬件，以及未来硬件上，以自动化的方式生成高效的计算密集/访存密集的算子kernel。

职位介绍： 1- 为NV产品提供散热解决方案。 2- 利用热仿真软件，结合模拟测试预测产品温度分布情况。 3- 搭建测试平台，验证散热方案。

该岗位需要onsite实习1年，有转正机会！ 我们特别需要以下关键领域的专业知识： 使用深度学习 (DL) 和人工智能 (AI) 技术实现游戏自动化，重点关注游戏行业的研究和应用。候选人应精通以下深度学习技术： • 深度强化学习 (DRL) • 卷积神经网络 (CNN) 和循环神经网络 (RNN) • 生成对抗网络 (GAN) 具备 AI 驱动机器人、游戏内移动自动化和玩家行为预测方面的经验或研究经验者优先考虑。 具备使用 AI 开发工具创建测试计划、开发测试用例和自动化测试用例的知识。

· Direction 1: Deep Learning Software Architect (Fast Kernals/TensorRT/ TensorRT-LLM/ Architecture) · Direction 2: GPU Graphics Performance Architect (Ray Tracing and Neural Rendering/ Real-time rendering/ SoC platform) · Direction 3: GPU Instruction Processor & Work Distributor Architect · Direction 4: System Software Engineer What you’ll be doing: · Design and verify next generation GPU architecture for computer graphics or high performance computing for cutting edge technology and features. · Develop production and core infrastructure software for key NVIDIA product line. · Optimize performance for gaming, ray tracing and deep learning applications at full stack from high level application, software stack to hardware architecture.

· Direction 1: Deep Learning Software Architect (Fast Kernals/TensorRT/ TensorRT-LLM/ Architecture) · Direction 2: GPU Graphics Performance Architect (Ray Tracing and Neural Rendering/ Real-time rendering/ SoC platform) · Direction 3: GPU Instruction Processor & Work Distributor Architect · Direction 4: System Software Engineer What you’ll be doing: · Design and verify next generation GPU architecture for computer graphics or high performance computing for cutting edge technology and features. · Develop production and core infrastructure software for key NVIDIA product line. · Optimize performance for gaming, ray tracing and deep learning applications at full stack from high level application, software stack to hardware architecture.

What you'll be doing: Performing in-depth analysis and optimization to ensure the best possible performance on current and/or next-generation NVIDIA GPUs. Understanding and analyzing the interplay of hardware and software architectures on core algorithms, programming models, and applications. Actively collaborating with the hardware design, software engineering, product, and research teams to guide the direction of accelerated computing. Diving into accelerated computing applications to facilitate software-hardware co-design. Write up and present your work by writing white papers, conference publications, official blog posts, patent applications, etc.

· Direction 1: Deep Learning Software Architect (Fast Kernals/TensorRT/ TensorRT-LLM/ Architecture) · Direction 2: GPU Graphics Performance Architect (Ray Tracing and Neural Rendering/ Real-time rendering/ SoC platform) · Direction 3: GPU Instruction Processor & Work Distributor Architect · Direction 4: System Software Engineer What you’ll be doing: · Design and verify next generation GPU architecture for computer graphics or high performance computing for cutting edge technology and features. · Develop production and core infrastructure software for key NVIDIA product line. · Optimize performance for gaming, ray tracing and deep learning applications at full stack from high level application, software stack to hardware architecture.