Amazon recently announced that it will spend US$110 million to develop its self-developed Trainium chips, a move aimed at reducing its dependence on Nvidia and improving the competitiveness of its cloud computing platform AWS. This program called "Build on Trainium" will focus on supporting universities and scientific research institutions to use Trainium chips to conduct in-depth research in the field of artificial intelligence and promote AI technology innovation and development. Amazon has built a supercluster containing up to 40,000 Trainium chips to provide researchers with powerful computing resources.
Amazon recently announced that it will inject $110 million into AI research to support the development of its self-designed Trainium chip. This move is aimed at reducing Amazon's dependence on Nvidia and promoting the technological advancement of its own cloud computing platform AWS.
Picture source note: The picture is generated by AI, and the picture authorization service provider Midjourney
Trainium is a machine learning chip tailored for deep learning training and inference tasks. Amazon hopes that through this investment, it will encourage the academic community to use Trainium chips to conduct research on new AI architectures, machine learning libraries, and performance improvements.
The initiative, called "Build on Trainium," will provide university researchers with the opportunity to use Trainium for large-scale distributed computing. Amazon says it has built a research supercluster of up to 40,000 Trainium chips that are optimized for AI’s unique workloads and computing structures. Through this plan, Amazon hopes to attract all types of AI research, including algorithm improvement and research on large-scale distributed systems.
Amazon also promises that all AI results achieved through the program will be shared in an open source manner so that researchers and developers can further advance the technology. In addition, the program will provide funding for new research and student education, and plans for multiple rounds of research awards, with selected projects receiving AWS training credits and access to large Trainium superclusters.
Todd C. Mowry, computer science professor at Carnegie Mellon University's Catalyst research group, who is involved in the program, said AWS's Build on Trainium program provides their faculty and students with large-scale access to modern accelerators that will help Expand research on tensor program compilation, machine learning parallelization, and language model services and tuning.
This investment by Amazon not only shows its ambition in the field of AI, but also further promotes cooperation with academia, striving to maintain a competitive advantage at the forefront of AI technology.
Highlight:
Amazon invests $110 million to support university AI research and reduce dependence on Nvidia.
The "Build on Trainium" program offers researchers the opportunity to utilize 40,000 Trainium chips.
All research results will be open source, and financial support and AWS training points will be provided to selected projects.
Amazon’s move not only demonstrates its strategic layout in the field of artificial intelligence, but also provides valuable resources and opportunities for the academic community, indicating that the field of AI will usher in new development opportunities in the future and promote broader cooperation and innovation. This investment will have a profound impact on the development of artificial intelligence technology.