On this day 10 years ago my wife called me at SC11

... and told me the results of a biopsy, that began this journey.  I was ready to pack up our booth and come home, but she said "stay".  We had an appointment scheduled with the oncologist a few weeks out (first available).

The (current) epilog.  She is teaching AP Statistics, and Algebra II at a local college prep high school.  So we know how this has turned out to date.  

Cancer has taken too many of our family, and impacted all of us in one way or another (patient, or family). One of my biggest desires/wishes is to see us (collective "us") be able to attack it and win battles using, in part, high performance computing systems (HPC).

Last year, I worked on, what is now called an "urgent" computing problems, provding a mechanism for massively parallel sampling of docking space using codes that were basically poorly written for parallelism (Autodock Vina), and obtaining 250-1000x overall performance relative to the "parallel" version.  This still isn't enough.  The space to sample is simply too huge for brute force methods to work efficiently.  We need better theory to guide simulation and experiment.  

All this said, I am so very happy that our journey (thus far) has turned out the way it did.   I wrote more on the journey here.  I leaned on many people for support (huge shout out to Chris Samuel here).  Randall Munroe of xkcd fame also went through this with his wife.  His writing/comics capture even more of the angst than my writings did.  Highly recommended, especially if you are supporting a loved one going through this themselves.  

Highly recommended reads (and some of them may get you to tear up) are 881, 931, 1141 was especially hard, and I saw it at the 1 year anniversary of our surgery.  1928 gave me hope at the 5-6 year boundary, meaning we were survivors.  I had just walked my daughter around campus of University of Michigan then, as she was starting college, and we brought our eclipse viewer with us.  

In roughly 7 weeks, we hit our 10 years mark.  Which is the subject of 2386.  

In no uncertain terms, FUCK CANCER.  And while we are at it, FUCK COVID.  We need a sustained Manhattan project like program to deal with urgent stuff like this.  Supercomputing is part of it, but not all, or even most of it.  We need better models, better simulation, more science to tackle these, and other emergent, and longstanding threats.

Show Comments