Intel Identifies the Future of High Performance Computing. Surprise. It Is Itself

February 29, 2016

I make a feeble attempt to pay attention to innovations in high performance computing. The reason is that some mathematical procedures require lots of computing resources; for example, figuring out the interactions in a fusion plasma test. Think in terms of weeks of calculation. Bummer. Most folks believe that the cloud and other semi magical marketing buzzwords have made super computers as fast as those in a sci fi movie. Wrong, gentle reader. There are computational issues. Big O?

image

I read with interest “The Future of High Performance Computing Has Arrived.” The write up does not do too much with the GPU methods, the brute force methods, or the “quantum” hopes and dreams.

Nope.

The write up points out with a nifty diagram with many Intel labels:

Intel is tightly integrating the technologies at both the component and system levels, to create a highly efficient and capable infrastructure. One of the outcomes of this level of integration is how it scales across both the node and the system. The result is that it essentially raises the center of gravity of the memory pyramid and makes it fatter, which will enable faster and more efficient data movement.

I like the mathy center of gravity lingo. It reminds me of the “no gravity” buzzword from 15 years ago.

Allegedly Moore’s Law is dead. Maybe? Maybe not? But as long as we are geared up with Von Neumann’s saddles and bits, Intel is going to ride that pony.

Gentle reader, we need much more computing horse power. Is it time to look for a different horse to ride? Intel does not agree.

Stephen E Arnold, February 27, 2016

Comments

Comments are closed.

  • Archives

  • Recent Posts

  • Meta