Martin wrote:

stanmc wrote:
[...]

... if I convert the times given by SETI at the completion of a WU I get a best time of 10,071.8 seconds. 2 hours 47 minutes 51.8 seconds. I am using the Windows Screensaver version with the screensaver set to blank the screen.

The CLI is much faster. With the screen saver, the data for the graphics is still generated although you do get a slight speed-up by not subsequently drawing the graphics.

You can also use various add-ons that very efficiently show pretty graphics to let you see what's happening with the CLI data.

My computer is an ASUS a7n8x dlx v2.0 running a Barton 2500+ o'clocked to 2.08GHz.

Very nice too. Checked out ok with memtest86 and also the Prime95 torture test?

How accurate is the SETI timekeeping function?

Now that depends... on many things... including what OS you're using.


Regards,
Martin

My recollection is that there is a very big improvement with the screensaver set to other than the SETI one. My speed with it set to the SETI screensaver function was like half again as much (150%). I haven't used it in months. In a non-scientific test (observation) I just let it run for 2 minutes with graphics display up and it completed 1.343% and with the graphics inactive it ended the next two minutes with 3.394 for a difference of 2.051. (1.52+%)

I think the timekeeping function is reasonable. I note my start times and find that the difference is miniscule. I've heard complaints that the clock on the Asus a7n8x can be inaccurate enough to display several seconds wrong each hour.

System passes Memtest86 and runs stable.

Someday, when I get more adventurous I'll look at the CLI. At the moment I'm content to let the screensaver version execute at start up and not have to be bothered with until I decide to check progress or shut it down to do other things.

Thanks for the response.

Stan