What is HPC?

Trying to define what HPC means in a post cloud world.

What is HPC?

Words are important. The names we give things matter.

Without a shared understanding of what things mean it becomes difficult, if not impossible, for us to discuss those things. To improve them. To move the game on. 

Take DevOps and SRE for example. If I had a penny for every time I see the two terms conflated. Used Interchangeable. Or worse, used to refer to a traditional ops team. Well, I wouldn’t be sat here writing posts on LinkedIn. Or at the very least I’d be sat on a beach in the Maldives writing them.

Yesterday I was in a meeting with two highly experienced veterans of HPC. During the conversation we briefly touched on whether what we are talking about is in fact HPC. Or would it be more correct to call it grid computing, or indeed HTC. This isn’t the first time I have had such a conversation with those in the industry either. It a semi regular  fixture of many a meeting.

So, what is HPC in a post cloud world? What is a supercomputer in a world where something that easily breaks into the top 10 list of most powerful supercomputers in the world can be created, used and destroyed in less than a day? Do we need a new name?

Dear reader, I’m not even going to attempt to define the terms but I invite and welcome you to do so in the comments.