CareerPlan

AI Compiler & Kernel Engineer

Beijing, ChinaFull-time2026-10-13 → 2026-11-16

Core

Develop and optimize high-performance AI operators, kernels, and compilers for large-scale cloud inference workloads on Qualcomm Cloud AI platforms.

Role type

Senior IC AI compiler and kernel engineer

Builds

Optimized inference software stack for LLMs, MoE, and multimodal models

Domain

Cloud AI inference, compiler technology, parallel computing

Deliverable

production ML models

Required skills

C++, Python, Linux, Transformer architectures (LLMs, MoE), compiler frameworks (MLIR, Triton, TVM), graph transformation, operator lowering, scheduling, kernel auto-tuning, code generation, memory scheduling, parallel computing, vectorization, tiling, memory hierarchy optimization, workload scheduling

Preferred skills

Optimizing transformer operators (Attention, FlashAttention, KV Cache, Embedding, GEMM, Softmax, MoE routing), custom operator development for AI accelerators (GPUs, NPUs, TPUs)