Jobs

Inferact is seeking an AMD GPU Performance Engineer to optimize vLLM for the AMD accelerator ecosystem, making it a first-class inference engine. This role involves building and optimizing AMD GPU backends, kernels, and benchmarking infrastructure using tools like ROCm, HIP, and Triton. The ideal candidate will have hands-on experience with AMD GPU workload optimization and a deep understanding of GPU execution and kernel performance. You will work at the intersection of inference systems, compilers, and hardware architecture to improve critical performance paths and ensure vLLM delivers frontier inference performance on AMD GPUs.