By Dr Beck
If you ever wanted a view of what the future might look like if an articulate, intelligent woman of color was given the reigns of power, that vision was present. If you just were hoping for bits of insight on where technology is headed that too was present. But if you wanted to be inspired by the work of others at AWS and so inspire folks yourself, this Summit was meant for you.
Before being invited to the Summit, I had gotten a free account back in March on Amazon’s Cloud. It was for my own research with Python3’s parallel processing concepts and now AWS’ “pcluster” or parallel cluster integration and coding. I had take a certification in AWS Basics. Imagine that, for a second. I had a free account with a nearly unlimited number of multiprocessors with cached memory. Oh my!
Today, I learned those were not just any processors. They were Amazon’s own ARM design called “the graviton”. It is now in its third generation. The graviton is made in Austin, TX and Kirkland, WA by Annapurna Labs, part of AWS. There are 64 core units within each graviton, all with cached or dedicated memory for each core unit. WoW!
For a parallel processing “geek” like myself that is something. The most efficient parallel code runs with individually cached memory, as opposed to shared or bulk memory. Having 64 core units in one chip makes a tidy, super-fast cluster. Remember, it was not that long ago (2012-2015) we were struggling with 12-24 core Linux clusters, working with 2-3 clusters at a time and waiting 8-10 hours for our calculations to end on classical Maxwell equations/interations of EM propagation using nanometer-sized Yee Cells.
I have attached the sign-up page that’s at AWS below for a free account like I have and a chunk of the Amazon Cloud for your experimentation. It will take you to the Summit page. Scroll up until you see the “Get started for free,” if you are interested in AWS. I am retired and do not work for anybody. I give opinions only.