there are so many cores

Just another site

Build the right thing the first time

I am unsure if this will work but will try.

  1. shorter iterations, monthly public mini-releases
  2. beta quality release this summer
  3. production quality release by end of this year

If you are interested in this work, please influence it. Tell me what to build. The biggest risk for this project is that I fail to assess value correctly and build the wrong thing.

An example of what I mean is stencil kernels. It was Korneliussen’s comment to this blog and then seeing Christen’s presentation at SIAM PP12 on stencil kernels that pointed out a glaring omission. Part of this was also Robert Rossi’s presentation at the HPC & GPU Supercomputing Meetup in Silicon Valley.

Really, I have done some computer vision before. So I understand the importance of stencils for non-linearity and feature detection. Yet, it was not until I was prompted by all three of these that it became a project priority.

It is very easy to miss what is important. That’s o.k. if you have a lot of resources. I am very resource constrained. So I must build the right thing the first time.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: