HostID 1001562 - Richard Haselgrove's Q6600 Quad Core

archae86
archae86
Joined: 6 Dec 05
Posts: 3,008
Credit: 4,847,877,910
RAC: 3,347,963

RE: But I got this one nice

Message 78967 in response to message 78966

Quote:
But I got this one nice run. Not enough to nail down the maxima and minima, but plenty of food for thought for all you curve-fitters out there.

My working estimate would put the cycle length for 1167 Hz at 400, which would mean your data extended up close to, but not reaching the third cycle peak at 1200, and that your big wiggle section was somewhat less than midway in sequence between that peak and a preceding minimum at 1000.

The wiggles are certainly prominent. Unless we learn to predict wiggles in detail (not something I've even thought about trying to do), it emphasizes that for performance comparisons one needs enough samples to average away the wiggle effect, and to use matching methods which don't get driven wildly off course by them.

If convenient could you post or send me the linecount for the skygrid file for 1170 Hz? I'd initially estimate it to be about 335560 lines. I already see from the web site that you see 841 skypoints, which is no surprise. (at higher frequencies all I've seen is 840, 841, or 842)

Richard Haselgrove
Richard Haselgrove
Joined: 10 Dec 05
Posts: 2,040
Credit: 686,928,874
RAC: 927,719

RE: If convenient could you

Message 78968 in response to message 78967

Quote:
If convenient could you post or send me the linecount for the skygrid file for 1170 Hz? I'd initially estimate it to be about 335560 lines.


Who's a clever boy? Linecount is 335,619

archae86
archae86
Joined: 6 Dec 05
Posts: 3,008
Credit: 4,847,877,910
RAC: 3,347,963

RE: Linecount is

Message 78969 in response to message 78968

Quote:
Linecount is 335,619

Thanks, that means the Bikeman estimate method directly applied to linecount and skypoints specific to this case and my current best algorithmic estimate agree on an expected cycle length of 399 (I was rounding off before in saying 400). So your data got really close to the peak, it seems. Selfish oaf that I am, I'd have much preferred to see both sides.

Holmis
Joined: 4 Jan 05
Posts: 1,118
Credit: 1,005,268,993
RAC: 991,139

RE: RE: Edit: I do have

Message 78970 in response to message 78964

Quote:
Quote:
Edit: I do have task with sequence numbers from 132 down to 94 in cache if it will help, but it will take a few days to process them...
In that frequency I'm already pretty confident that the "real" cycle length is 57, but it would be good to review your host's data when results from about 108 through 120 are available, as a check.

You can find the data for frequency 434.30 and sequence numbers 151 - 101 here.

archae86
archae86
Joined: 6 Dec 05
Posts: 3,008
Credit: 4,847,877,910
RAC: 3,347,963

RE: You can find the data

Message 78971 in response to message 78970

Quote:
You can find the data for frequency 434.30 and sequence numbers 151 - 101 here.

Thanks, I'd forgotten to look. As with my own hosts, there is a little more noise than convenient, but squinting a bit, the second peak fits about 113, which makes the cycle length estimate from this observation 56.5.

Comment viewing options

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