C[omp]ute

Welcome to my blog, which was once a mailing list of the same name and is still generated by mail. Please reply via the "comment" links.

Always interested in offers/projects/new ideas. Eclectic experience in fields like: numerical computing; Python web; Java enterprise; functional languages; GPGPU; SQL databases; etc. Based in Santiago, Chile; telecommute worldwide. CV; email.

Personal Projects

Lepl parser for Python.

Colorless Green.

Photography around Santiago.

SVG experiment.

Professional Portfolio

Calibration of seismometers.

Data access via web services.

Cache rewrite.

Extending OpenSSH.

Last 100 entries

NYT Has More Details; Obvious Question; Advertising Low Cost Routes?; Similar Analysis Here; My Current Take On Surveillance Scandal; Last.fm is Hiring; How I Am 2; The back-wards compatibility fallacy; Wiggle The Mouse To Pass The Test; Python Enums on Crack, Part II; Multiple Monitors with Linux; What Is Happening In Turkey; A Simpler Enum; John Fogerty on IAmA; And You May Well Ask...; Progress on a Better Enum; I'm a MACHIIIIIIIIIIIIIIINE; Those little tab things on the side of jet engines; Re: Python's sad, unimaginative Enum; Some explanation; Printing binary trees sideways; About "Python's sad, unimaginative Enum"; Atoms in python; Some good feedback here; Frustration Understood; I agree with you #nt; What would be imaginative?; Re: Enum; this is fucking useless; Enum; Python's sad, unimaginative Enum; Possible Fix; Work, Exhaustion, Vacation; VirtualBox with Centos 6.3 to 6.4, client; Matasano - Programming Lessons Learned; PDF to HTML; Alternate Substitution; Why RSA Works; Trigger; Dreaming of Death; Example: Tracing; Using Coroutines In Protocol Simulations; Python 3.3 Only; Pure Python SHA1 and MD4 Implementations; Ubuntu on VirtualBox; Starting TOR as a service on OpenSuse 12.3; 1001 Albums; Using fail2ban on OpenSuse 12.3; PPPoE on OpenSuse 12.3; Good Article on Unified Physics; It's Police (Carabineros); Linux Software for Listening to and Exploring Music; Android is Pretty Bad; Lucky Number; 3D Printing for Casting; Cover Art for MPDroid; Who'd a thought the French were so bigoted?; PS Input Signal; Small Problem with Roksan K2 Amp; Roksan K2 Amp + ATC SCM7 Speakers; Do What Makes Sense; Re: Arguing About Tests, Still; Arguing About Tests, Still; Images; Good Article on NY Drummers; Related Bug Report; Getting Python 3.3 and Virtualenv Working in OpenSuse 12.3; How I Am; Awesome video about digital audio; The Difference Between Dimensional and Normalized Databases; The rise of the new Chinese bogeyman; Updated Syntax; Very First Steps to C-ORM; The Ideal User Interface For Music Exploration; Can The Republicans Be Saved?; Rate Limiting Calls to EchoNest; Mods to Cache; Comparing UYKFG and UYKFD/E/F; Someone Else is Concerned; EchoNest-based Playlist Generator for MPD; Example Voting Results; A Heavyweight Python Cache; Identifying Artists with EchoNest; Notes on Pregalex / Pregabalina / Lyrica; The Neil Cowley Trio; Drake - Make for Data; A Reliable Python Web Service; Useful Python Date/Time Library?; Need to Sleep, But this is Good; Command Line Set Difference; Little Details...; Linux Command Line Tricks; AutoTools Tutorial; Hangman Tactics; A Tor Proxy Embedded In A Web Page; Tree (Nested Dicts) in Python; Sleeping at Parties; I Know Someone Who Hurts Other People; Light and Tea; Description of the LCS35 Time Capsule Crypto-Puzzle; Re: I can relate to that ...

Matlab on NVidia GPUs

From: andrew cooke <andrew@...>

Date: Sat, 17 Oct 2009 07:35:17 -0300

I can't find any notes of what used to exist, but there are now 2
companies - one appears to be offering freeware (but not yet on
Linux).

http://gp-you.org/index.php (free)

http://www.accelereyes.com/ (commercial)

Andrew

Calling OpenCL Directly

From: andrew cooke <andrew@...>

Date: Mon, 28 Dec 2009 21:21:53 -0300

Instead of using those libraries, I ended up writing OpenCL (which is
very similar to CUDA, but cross-platform) directly.  The final speedup
was a factor of 100.

The reason I didn't use the libraries is that I couldn't see how they
could work efficiently for our case.  There were two reasons for this.

First, I tried vectorizing the Matlab code so that it might use
multiple cores on the CPU.  This was a failure, which made me suspect
that the code didn't have the form necessary to be auto-parallelized
(in the very core of the loop is an indirection which makes things
non-trivial).

Second, I was dealing with large data sets and needed to have qute
detailed control over what data was moved onto the GPU at what time
(again related to the indirection).

The entire project (including rewriting code in plain C as well as in
OpenCL) was 1-2 months, and got a 100x speedup.  An awful lot of that
time was spent (1) reverse engineering the algorithm I was porting and
(2) learning the details of OpenCL/CUDA.  Knowing what I know now, I
could repeat the work in a week.

The 100x speedup sounds impressive, but plain C was 6x faster than
Matlab, so the GPU only adds another factor of 17.  This is, I think,
because of the indirection I discussed earlier - the memory accesses
cannot be coallesced, which makes the entire calculation memory bound
(although I have not yet profiled the app to check this).

On the other hand, the original code was running on a rather
expensive, new dual Xeon box.  The GPU was a sub-$200 generic card
from newegg.  So "speed per $" is over x1000....

Andrew

Matlab/OpenCL Cross Reference

From: andrew cooke <andrew@...>

Date: Mon, 28 Dec 2009 21:22:43 -0300

PS See http://www.acooke.org/cute/Experience0.html

Comment on this post