AI Compiler

The AI Compiler team is central to developing the means by which cutting-edge AI models and workflows achieve world-record results on our unique hardware architecture.We are seeking a talented AI compiler expert to form the new team in Belgrade. In this high-visibility, hands-on role, you and your team will be building the AI compiler for next generation platforms.Requirements6+ years of experience in software engineering3+ years of experience developing a backend for a large-scale, open-source-based compiler (e.g. LLVM)2+ years leading a software group of 3 or more full-time software engineersAt least one of the following:Demonstrable open-source contribution to relevant projectsExperience developing state-of-the-art neural network architecturesSpecialization experience in ML accelerators, high performance computing, GPU and/or CPU architecturesBackend experience with MLIR: an advantageResponsibilitiesDevelop a state-of-the-art MLIR-based AI compiler stack targeting our next generation hardware and delivering superior acceleration to AI applicationsInterface with platform teams to define IRs leveraging unique our hardware architectureCollaborate with runtime teams to shape the future of our runtime stack, from driver level up to integration into popular AI frameworksCollaborate with toolchain teams to deliver a holistic toolchain for our customers, supporting static and dynamic development workflows