2027校招-推理框架工程师（上海/成都/北京） 北京、上海、成都 校招 正式 职位描述 分布式推理系统开发1、设计并实现大规模AI模型（如LLM、多模态模型）的分布式推理框架，优化吞吐量（Throughput）和延迟（Latency）。2、开发模型并行（Tensor/Pipeline Parallelism）、动态批处理（Dynamic Batching）和连续批处理（Continuous Batching）策略。3、解决多节点、多GPU环境下的负载均衡和通信瓶颈问题。4、优化计算图执行（如使用TensorRT、ONNX Runtime、vLLM等工具）。5、实现量化（FP8/INT8）、KV Cache优化、Flash Attention等加速技术。 职位要求 精通 Python/C++，熟悉PyTorch/TensorFlow推理生态。深入理解分布式系统（NCCL/RPC通信、GPU Direct RDMA）和 并行计算（CUDA、OpenMP）。熟悉推理加速技术（量化、剪枝、算子融合）和框架（TensorRT、vLLM 、Sglang）。 投递...
Company: Qualcomm China Job Area:Engineering Group, Engineering Group Software Engineering General Summary: You will design, implement, optimize, and maintain GStreamer plugins for streaming, hardware-accelerated video/audio encoding & decoding, muxing/demuxing, camera integration, GPU/NPU AI inference, and end-to-end
Our Purpose TERADYNE, where experience meets innovation and driving excellence in every connection. We are fueled by creativity and diversity of thought and in our workforce. Our employees are supported to innovate and learn something new