Ha ha, and there I was reading of 9GB models being run at one project. Don't know of a GPU that has this, or the bandwidth to exchange all that data back to RAM and storage, but those with 32GB system ram were commenting that their octo showed jobs being frozen "waiting for memory" when trying to run more than 3 of those [they have not discovered <max_concurrent> yet.] Think multi-threading would be the better solution, to not have devices end up in RAM cram.

+100500 :)

dpenfoldbrown I couldn't agree with you more that it is more useful to dedicate limited resources to tasks that can achieve broad benefit rather than targeting specific diseases or issues.

With that said, I want to double down on my suggestions for the Mole Rat and Jelly fish as those should produce broadly interesting and hopefully useful data not only for aging but other considerations as well. I also thinking aging is a pretty broad area to study as it is the root cause or factor in the many if not the majority of health issues.

In general, it is the role of government, big companies, NGO's and big high risk funding to produce tools and data that will ultimately make it easy and inexpensive for startups and small companies to develop biological tests, treatments, drugs, etc... This is what happened in the computing space. Once big money was spent to develop low cost PCs and development software, the web and app movements took off.

Thanks,

-NGM