[Beowulf] Degree

[Joe Landman]

On 10/24/2012 08:58 AM, Andrew Holway wrote:
> Hello,
>
> I have no education. I left school at 16 and, in my mid 20's, somehow
> got into supercomputing and now am doing all kinds of silly stuff.

I hate to put it like this, but a degree is, in effect, a union card.  I 
know I am gonna get hate mail for this.

Very bright people have made great contributions without having the 
higher forms of the union card.  c.f. 
http://en.wikipedia.org/wiki/Robert_Pound (not the only example)

>
> Thing is, I need some kind of degree in this stuff to do the kind of
> work I really want to do. Especially in Germany, organisations
> involved in HPC usually strictly require an advanced degree if you
> want to be paid. They can hire you but can't pay you due to union
> restricted pay grades that are based on your level of education.

Look at the degree as I indicated above.  It means you've paid your 
union dues by other means, working your way from apprentice (B.S.) to 
craftsperson/master craftsperson (M.S./Ph.D).  This said, some of the 
more enlightened departments and schools may not have a problem with 
accepting experience in the field as evidence of passing various levels 
of competency.  Here in the US, I seem to remember some of the CUNY 
schools (City University of New York) doing this.

OTOH, when I transfered institutions between MS and PhD, what I learned 
from the process (on the administrative side) was that the only 
hard/real requirement was that I *pay* for 30 credit hours of research 
study.  Call this a cautionary tale, in that you need to check out their 
academic AND administrative requirements ahead of time.  More than 
likely, there will need to be some expenditure.

>
> I've managed to do some interesting stuff freelance but it was a real PITA.
>
> Im currently working as a system architect for a Berlin firm,
> designing and testing a new large scale cloudy type infrastructure. I
> am the project lead and am completing pretty much all the work myself.
> Playing with technologies such as Infiniband, ZFS, iSCSI, Databases et
> al. The problems are complex and require a very special understanding
> of the whole stack.
>
> Would it be possible to turn this kind of work into a masters degree
> or perhaps even a PHD? Its been suggested that the scale of the
> projects that I am working on now are equivalent to a PHD.......

MS, certainly, but you'd need to get with a good uni/prof to do this. 
And you'd need to worry about the administrivia as indicated.

You might be able to attach yourself to a uni in the US or Canada, with 
facilities at CERN or others, and work on not just HPC but applications. 
  I had lots of grad school buddies spend weeks/months at CERN.  If you 
are more into the hardware/software side than the app side, computer 
engineering with HPC specialization should be possible (though its a 
function of the advisor/group/department/...).

Friendly word of advice on this:  find an advisor (Masters or Ph.D) who 
publishes with a frequency greater than 1 paper per year.

>
> I have the resources, time, intent and lots of complex problems to solve.
>
> Would it be possible for a friendly professor to supervise me in this
> kind of circumstance?
>
> If this is possible, what kind of costs would I incur?
>

See above, each uni is different.  Most likely won't be cheap/free. 
Look at it as an investment in your future.  You are assuming that the 
return will justify the expenditure.  I personally don't do much physics 
these days.  I do lots of HPC storage, and related things.

Focus on your goals, lay out a plan with a timeline, and see if you can 
negotiate your way into a good group.  There are a number of good profs 
here on the list, and many more attending SC** and likely ISC**, so you 
should be able to find/connect with a good group.  Though it might take 
a while to find exactly the right fit.

-- 
Joseph Landman, Ph.D
Founder and CEO
Scalable Informatics Inc.
email: landman at scalableinformatics.com
web  : http://scalableinformatics.com
        http://scalableinformatics.com/sicluster
phone: +1 734 786 8423 x121
fax  : +1 866 888 3112
cell : +1 734 612 4615