Differences between revisions 6 and 7
Revision 6 as of 2008-06-20 19:40:34
Size: 2428
Editor: was
Comment:
Revision 7 as of 2008-11-14 13:42:15
Size: 2430
Editor: anonymous
Comment: converted to 1.6 markup
Deletions are marked like this. Additions are marked like this.
Line 15: Line 15:
4. [http://pyprocessing.berlios.de/ Pyprocessing] -- learned about it, used it, integrated with @parallel with Mike Hansen. Very interesting technology. Is '''very''' likely to finally make it so we can start using parallel computing in the core of the Sage library. This is, imho, major. 4. [[http://pyprocessing.berlios.de/|Pyprocessing]] -- learned about it, used it, integrated with @parallel with Mike Hansen. Very interesting technology. Is '''very''' likely to finally make it so we can start using parallel computing in the core of the Sage library. This is, imho, major.

Stein -- what I did at Dev Days 1

1. Made a start on sage lite. It didn't go anywhere, but I have some ideas...

2. Implemented and tested the pickle jar. http://trac.sagemath.org/sage_trac/ticket/3482

It turns out that *all* 465 pickles in sage-3.0.3 made on opteron 64-bit linux *unpickle* fine on 32-bit osx intel.

3. Basic decorator and primitive for parallel computing in sage: http://trac.sagemath.org/sage_trac/ticket/3467 Actually used this to compute 1.6GB of modular symbols spaces in parallel. Works well. There are a few details that it would be nice to add, but already this is a very nice useful thing to have. I deleted more code than I wrote.

4. Pyprocessing -- learned about it, used it, integrated with @parallel with Mike Hansen. Very interesting technology. Is very likely to finally make it so we can start using parallel computing in the core of the Sage library. This is, imho, major.

5. Refereed a lot of patches.

6. Discussed notebook database schema with Tom Boothby.

7. Helped a little bit with the coercion rewrite

8. Rode a Segway.

---

Modular Forms

At the modular forms workshop, I:

1. Gave an intro talk with challenge problems.

2. with Citro and Butt: Estimated time to solve them.

3. with Citro and Butt: Wrote code and computed 1.6GB of data (weight 2 and levels up to about 2200). Very surprising timing results. Here's the data, actively being computed:

Tasks: 479 total,   9 running, 469 sleeping,   0 stopped,   1 zombie
Cpu(s): 10.3%us,  3.3%sy, 36.7%ni, 49.7%id,  0.0%wa,  0.0%hi,  0.0%si,  0.0%st
Mem:  65993220k total, 63870396k used,  2122824k free,  5090292k buffers
Swap:  2931820k total,   364768k used,  2567052k free, 17267892k cached

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND                    
10080 was       39  19 1139m 785m 7880 R  100  1.2  23:08.39 sage-ipython                
10069 was       39  19  894m 540m 5820 R  100  0.8  23:08.89 sage-ipython                
10071 was       39  19 2523m 2.1g 8144 R  100  3.4  23:05.69 sage-ipython                
10076 was       39  19 2290m 1.9g 8056 R  100  3.0  23:00.41 sage-ipython                
10077 was       39  19 2003m 1.6g 7992 R  100  2.6  22:59.63 sage-ipython                
...

dev1/stein (last edited 2008-11-14 13:42:15 by anonymous)