Senior Compiler Performance Engineer (LLVM/C++)

Location: On-site in Shanghai, China

Company

hirify.global develops processor platforms and software technologies for heterogeneous computing, including compiler technologies for CPUs, GPUs, and accelerators.

What you will do

• Lead performance analysis of real-world applications, AI/HPC workloads, and industry benchmarks; identify bottlenecks across the compiler, runtime, libraries, and microarchitecture.
• Design and productize compiler optimizations and code generation features that expose hirify.global CPU and GPU capabilities.
• Own performance for a compiler component or workload domain, track targets through continuous integration, and manage regressions and gains.
• Advance heterogeneous and offload compilation across SYCL, OpenMP, Triton, MLIR, and DSL paths.
• Collaborate with hardware architecture teams and open-source communities through designs, RFCs, upstream patches, and code reviews.
• Influence technical direction, review designs and code, mentor junior engineers, and represent the team's work to partners, customers, and the broader ecosystem.

Requirements

• Master's or PhD in Computer Science, Computer Engineering, or a related field with 5+ years of relevant industry experience; or a Bachelor's degree with 8+ years.
• Strong knowledge and hands-on experience in modern C++, data structures and algorithms, and compiler theory.
• Production experience in at least one compiler-stack area: middle-end optimization, backend code generation, language front end, or runtime and offload support.
• Demonstrated experience analyzing and tuning the performance of real-world applications on CPU and/or GPU.
• Proven ability to analyze and solve complex technical problems and deliver features through a full product cycle.
• Excellent written and oral communication skills for explaining technical details to engineers, architects, and stakeholders.

Nice to have

• Substantial experience with LLVM compiler infrastructure, including upstream contributions; MLIR experience is a strong plus.
• Deep knowledge of hirify.global x86 instruction set architecture and Advanced Vector Extensions, or GPU architecture and execution models.
• Experience with heterogeneous and parallel programming models such as SYCL, OpenMP offload, Triton, OpenCL, or CUDA.
• Experience optimizing AI/ML workloads or integrating compiler technology with PyTorch and inference engines.
• Experience with CI and performance-tracking infrastructure at scale.
• Experience leading projects across geographies and mentoring engineers.

Culture & Benefits

• On-site work model in Shanghai, China.
• Full-time employment with Shift 1 China scheduling.
• Work connected to open-source compiler communities and hirify.global hardware architecture teams.
• Commitment to ethical hiring and Responsible Business Alliance compliance; no recruitment or hiring fees are charged.