Intelligence artificial generative is gaining more and more traction, with solutions like ChatGPT becoming more and more popular. Intel, one of the world leaders in the field of technology, has recently launched its own model called Aurora January. This model is destined to revolutionize scientific research in various fields.
Topics of this article:
Aurora genAI: a new step forward for science
The new genAI Aurora model was announced during ISC23, a landmark event for the industry. The model includes several training sessions to work on various types of texts, scientific data and much more. This means Aurora genAI will be able to handle a wide range of data and information, making it one valuable tool for scientific research. Aurora genAI was created in collaboration with theArgonne National Laboratory e HPE. This model will work with the new Aurora supercomputer, which has more than two exaflops of performance. This extraordinary computing power allows the supercomputer to manage up to 1 trillion parameters to generate results.
Read also: Artificial Intelligence can read and translate thoughts: opportunity or threat?
Aurora genAI and its impact on scientific research
Rick steves, associate director of the Argonne laboratory, commented on Aurora genAI's announcement, emphasizing the importance of this innovation for science. According to Steves, the project aims to harness the full potential of the Aurora supercomputer to produce a resource that can be used for science downstream in Department of Energy laboratories and in collaboration with others.
Le Aurora system specs are impressive: One HPE Cray EX supercomputer, 63.744 GPUs, 21.248 CPUs, 1024 DAOS storage blocks, and HPE Slingshot high-performance Ethernet networking. These specifications indicate that the Aurora system is capable of delivering outstanding performance, even outperforming AMD MI250 GPUs and delivering 20% more efficiency in QMCPACK quantum mechanical applications than the H100 model.
Aurora genAI applications
Aurora genAI model can work with “general text, code, scientific texts and structured scientific data from biology, chemistry, materials science, physics, medicine and other sources“, because he is trained for it in a generative way. Intel has provided some application examples for the model, which include the design of molecules and materials, the synthesis of knowledge from millions of sources to suggest new and interesting experiments in systems biology, polymer chemistry and energetic materials, climate science and cosmology. It will also be able to accelerate research by identifying biological processes related to cancer and other diseases, using its generative potential to suggest targets and even drug development.
