Siemens Selected by RIKEN for AI Device Research

Siemens Digital Industries Software has been chosen by Japan’s national research agency, RIKEN, to support its next-generation AI device research. RIKEN will leverage Siemens’ emulation and High-Level Synthesis (HLS) platforms to advance their AI accelerator devices.

Siemens Selected by RIKEN for AI Device Research

Enhancing AI Accelerator Development

RIKEN, famous for developing the K computer and Fugaku supercomputer, will use Siemens’ Veloce Strato CS emulation and Catapult HLS platforms. These technologies will help RIKEN explore AI accelerator device architecture and design space.

RIKEN’s Strategic Research Goals

Dr. Kentaro Sano, Team Leader at RIKEN’s Processor Research Team, emphasized Siemens’ role in supporting their AI for Science research. “With Siemens’ tools, we can enhance our computational infrastructure to develop generative AI models for scientific discoveries,” he said. Siemens’ platforms are crucial in maintaining RIKEN’s leadership in computational science.

Siemens Supports Groundbreaking AI Research

Jean-Marie Brunet, Vice President at Siemens Digital Industries Software, expressed enthusiasm for supporting RIKEN’s AI device research. He highlighted that the Veloce Strato CS and Catapult HLS platforms provide comprehensive tools to help RIKEN achieve its research goals.

Strengthening AI Research in Japan

Yukio Tsuchida, Vice President for Japan at Siemens, echoed the sentiment. “We are proud to support RIKEN’s research and development efforts,” he said. Siemens looks forward to assisting RIKEN in pushing the boundaries of AI in scientific research.

Advanced Technology for AI Accelerator Development

Siemens’ Veloce Strato CS and Catapult HLS platforms provide advanced solutions for large-scale Application Specific Integrated Circuits (ASICs). These tools assist RIKEN in evaluating the performance of optimized AI circuits and system-on-chips (SoCs).

AI Accelerators for Future Supercomputers

The AI accelerator developed through this collaboration could play a key role in the next generation of supercomputers after Fugaku. This next-generation system is expected to be operational by 2030.