News

The Chan Zuckerberg Foundation is developing a GPU cluster for medical research

The Chan Zuckerberg Initiative Foundation, an NGO created in 2015 by Priscilla Chan and Mark Zuckerberg, has announced an initiative related to health and generative AI. For this they are ffinancing and developing a GPU cluster latest technology, that will use Artificial Intelligence to generate predictive models of cells healthy and sick.

The foundation hopes that it will help researchers better understand the cells of the human body, as well as their reactions. And they believe that with the collection of computers that will make up the cluster they will help them achieve the goal of curing, preventing or managing all types of diseases before the end of the century.

The system will use a cluster of around a thousand GPUs to train large AI language models with human cells. This NGO hopes that AI models can extract information and conclusions that go beyond what a team of human experts in the field can discover. But also, do it much more quickly, and reduce research processes from years to several weeks. It will be the time it takes for the models to find the information they need.

AI models could become very useful in biomedicine, and for example, predict how an immune cell responds to an infection, what happens at the cellular level when a child is born with a rare disease, or check how the body responds. a patient to a new medication.

This system will be one of the first to work with human cell models available on a large scale, suggesting that the investment made by the NGO could help researchers with good ideas but insufficient funding.

Among the data with which the models will be trained are those integrated into the genetic tool Chan Zuckerberg Cell, which has a database of about 50 million cells. They will also do so with the resources of the CZ Science research institutes, as well as with publicly available data sets. One of its goals, apparently, is to create a virtual biology simulator.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *