Machine Learning System Software Architect (AI)

Location: Sunnyvale, California, United States; work at hirify.global's Sunnyvale office

Company

hirify.global is developing AI silicon and systems around its Kunlun AI SoC and accelerator platform.

What you will do

• Architect and create high-performance machine learning system software and distributed AI training systems connecting thousands of Kunlun Accelerators and servers.
• Develop architectural innovations for the Kunlun AI SoC roadmap and design, simulate, and evaluate machine learning solutions.
• Build system-level ML architectures optimized for performance, power, and latency.
• Evaluate the power, performance, and cost of architectures and subsystems, and develop scalable modeling and performance-evaluation tools.
• Collaborate with system, application software, SoC design, verification, and validation engineers to optimize the full hardware/software stack.
• Monitor AI industry and academic trends and build partnerships with ML researchers and application developers.

Requirements

• 5+ years of experience architecting software solutions for machine learning, acceleration, and optimization, especially distributed training systems and HPC.
• Knowledge of machine learning markets, technology trends, software ecosystems, and emerging applications.
• Experience with TensorFlow, PyTorch, or PaddlePaddle, as well as CPUs, GPUs, memory systems, and accelerators.
• Experience with performance simulation and modeling in C++.
• MS or PhD in Electrical Engineering or Computer Engineering.
• Excellent communication skills in both English and Chinese.

Culture & Benefits

• Work in a highly technical AI hardware and systems environment.
• Collaborate in small, fast-moving teams.
• Self-directed work with opportunities to learn new skills and grow.
• Contribute to a mission focused on advancing AI silicon and machine learning systems.