[Beowulf] computational cost (was TeX/LaTex for Ubuntu)

Joe Landman landman at scalableinformatics.com
Mon Jun 24 17:53:19 PDT 2013


[not trying to hijack the thread, just thinking about an interesting 
variant based upon Chris' comment ... and as I am supposed to be on 
vacation, I shouldn't be reading/commenting right now ...]

On 06/24/2013 08:11 PM, Christopher Samuel wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> On 25/06/13 08:23, Max R. Dechantsreiter wrote:
>
>> Does anyone have advice on installing TeX (with LaTex) under
>> Ubuntu?
>
> Not really on topic (unless you plan to do embarrassingly parallel
> document generation on a Beowulf cluster), but you can do:

Funny about that ... I built my thesis with 'make' and LaTeX.  There 
were enough independent parts (plots generated with gnuplot and others 
driven from the makefile), that it could have generated it in parallel...

I remember it taking ~10 minutes on my laptop, and about 30 seconds on 
my SGI Indy.

Not exactly a beowulf class problem ... or ...

...

Think about what Netflix does.  It makes some large number N of 
different formatted files for every asset it streams.  Each formatted 
file is for a different bit rate/resolution.  I am guessing that the 
economics favor storage streaming versus computational cycles to 
rescale/bitrate adjust.

Why this is interesting, and the entre into this from the LaTeX bit, is 
that when I was starting out on my thesis research topic, the cost of 
precomputing, storing, and later retrieving specific data (integrals) 
was low compared to the computational cost of generating that data when 
needed, even though the code fundamentally favored "in-situ" generation 
(made for a cleaner design).

Previously, in undergrad, I was working on a research project for a 
professor, and the cost of storage was gigantic as compared to the cost 
of computing (in-situ again), even though the latter took somewhat more 
time.

What I find interesting is how, throughout my career on the non-vendor 
side of the fence, costs of one aspect or another significantly 
influenced code and code implementation.

LaTeX is a compiled language.  The compilation results in a device 
independent binary that is mapped to a device dependent state via 
various tools.  In LaTeX, we see a similar "echo" of this pattern where 
font generation can occur beforehand, or "in-situ" when needed.

In the Netflix case, its likely that the economics of compute in-situ 
for rescaling/resampling versus streaming pre-computed from storage is 
(currently) strongly favoring the streaming.

Honestly, I think Beowulf as a concept is all about gaining some measure 
of control over the economics of this situation.  Lower the cost of 
computing (and storage) to the point where the decisions about what to 
do with the code are driven by the needs of the application and effort 
moreso than the costs of a set of actions.

Just some random thoughts on a warm Monday night in NY.

Back to your regularly scheduled cluster ...

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


More information about the Beowulf mailing list