Credit raise

Alinator
Alinator
Joined: 8 May 05
Posts: 927
Credit: 9352143
RAC: 0

RE: No, it's not about

Message 74065 in response to message 74064

Quote:


No, it's not about lying. It's about getting the relation right between certain properties of the workunit and dynamically calculated parameters in the App that affect the runtime. At the very moment, neither of these is visible to the participants, and even to us only when running under a debugger (or with an amount of debug output that would fill up your harddisk). We see the sinewave over all workunits of a given frequency band, but it's still hard to predict the phase, which would be important for a specific WU.

BM

Hmmm...

It's been a very interesting peek under the hood and the work that goes on in the EAH pits so far on S5R2 and S5R3 so far. :-)

So can we take it that the reasonably good 'linearity' for the earlier runs was because a lot more time (resources) for preliminary testing was available to work out those factors ahead of time for the earlier science runs? IOW's there wasn't as much pressure to have the app ready to 'rock and roll' for the general public, so to speak.

As far as 'lying' goes, pehaps that's too harsh a word. Must come from hanging out too much over at SAH (where it seems like some people would kill to get an extra 'stone or two sometimes). ;-)

Alinator

Richard Haselgrove
Richard Haselgrove
Joined: 10 Dec 05
Posts: 2143
Credit: 3047427338
RAC: 2071937

RE: .... We see the

Message 74066 in response to message 74064

Quote:

.... We see the sinewave over all workunits of a given frequency band, but it's still hard to predict the phase, which would be important for a specific WU.

BM


Two observations which may help you narrow down what is happening.

1) I'm not seeing a sinewave, but sharp peaks of runtime with rounded minima between them - not as extreme as a cycloid, more like a catenary.

2) Not enough evidence yet, but I think you should look at the frequency of the oscillation, rather than the phase. Archae86 reported a frequency of 60 (actually, slightly above 60, looking at his graph), for a data template frequency of 542.70Hz. My first plot showed an oscillation frequency of about 58 for a data template frequency of 522.10Hz. I'm now seeing an oscillation frequency of 45 for a data template frequency of 464.35Hz. Beginnings of a trend there? In all cases, the phase seems to be that result __0 is at or close to a maximum of runtime.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.