Along with the introduction of the Aurora supercomputer, Intel unveiled its Aurora genAI artificial intelligence model A new actor is added to the scene. Of course, this one is not like other artificial intelligence that we have been introduced to in recent times, but Aurora genAI is a completely new model of generative artificial intelligence for scientific applications.
| Huawei is working on its own artificial intelligence, NetGPT |
Aurora genAI, up to 1 trillion parameters
Intel’s 2 Exaflops Aurora supercomputer can be considered a living being, and exactly this system will be used to power the Aurora genAI artificial intelligence model. Announced in today’s ISC23 keynote, the Aurora genAI model is intended to be trained on general text, scientific texts, scientific data, and domain-relevant code. This will be a purely science-focused generative AI model, with potential applications including:
- Systems Biology
- Cancer Research
- Climate Science
- Cosmology
- Polymer Chemistry & Materials
- basic science
The foundations of Intel’s Aurora genAI artificial intelligence model are Megatron and DeepSpeed. Most importantly, the target size considered for the new model is 1 trillion parameters. For comparison, just note that the free and public version of ChatGPT is only 175 million, and the Intel model is 5.7 times higher in the number of parameters.
Argonne National Laboratory, in collaboration with Intel and HPE, announced plans in a press release to create a series of generative AI models for the scientific research community:
“This project aims to use the full potential of the Aurora supercomputer to produce a resource that can be used for downstream science at Department of Energy laboratories and in collaboration with others,” said Rick Stevens, director of Argonne Laboratories.
These generative AI models for science are trained on general text, code, scientific texts, and structured scientific data from biology, chemistry, materials science, physics, medicine, and other sources.
The resulting models (with about 1 trillion parameters) in various scientific applications, from the design of molecules and materials to the synthesis of knowledge in millions of sources to propose new and interesting experiments in systems biology, chemistry of polymers and energy materials, climate science, and cosmology. will be used. This model will also be used to accelerate the identification of biological processes related to cancer and other diseases and to suggest targets for drug design.
Considering what we have seen in recent times from public and free artificial intelligence models, we can expect very high from this very powerful model to speed up scientific research. Surely, in the future, we will hear a lot of news about solving many complex scientific problems in the shortest possible time thanks to the Aurora genAI artificial intelligence model.
Related posts:
- Airplanes in the future will be piloted by artificial intelligence pilots
- Developing an artificial intelligence to read the human mind with fMRI
- Bill Gates: Artificial intelligence will teach children to read in the next 18 months
- Elon Musk founded his own company in the field of artificial intelligence
