Well, guys, I FINALLY got my rear in gear about doing some number crunching and here we go: a a 2.2GHz P4 and a 2.8 HT P4 running 24-7, about (no lie) 20Ghz spread across 8 different machines which run about 8 hours a day, 5 days a week, and another 5 machines totalling about 4Ghz which run for 8 hours about once a week in addition to all the others. It’s good to be the head of IT at work. :twisted:
lmao. Good Work 8).
DarkWyrm wrote:
Well, guys, I FINALLY got my rear in gear about doing some number crunching and here we go: a a 2.2GHz P4 and a 2.8 HT P4 running 24-7, about (no lie) 20Ghz spread across 8 different machines which run about 8 hours a day, 5 days a week, and another 5 machines totalling about 4Ghz which run for 8 hours about once a week in addition to all the others. It's good to be the head of IT at work. :twisted:
AWESOME!
DarkWyrm wrote:
Well, guys, I FINALLY got my rear in gear about doing some number crunching and here we go: a a 2.2GHz P4 and a 2.8 HT P4 running 24-7, about (no lie) 20Ghz spread across 8 different machines which run about 8 hours a day, 5 days a week, and another 5 machines totalling about 4Ghz which run for 8 hours about once a week in addition to all the others. It's good to be the head of IT at work. :twisted:
Sounds great, except for I don’t think that you actually harness all that power you have there at the moment. Check it.
I did 11.69 M last night.
You did 7.66 M last night.
Here’s my gear
1 AMD 3800+ box
1 AMD 4400x2 box
1 Intel P4 2ghz box
1 Intel P4 1,8ghz box
1 Intel P4 1,9ghz box.
In total I run 6 clients (the X2 gets 2 clients).
Now I’m sure your power there outruns mine easily. So I’m curious if you’ve setup something wrong. Cheers.
Nah. As I have been finding out, it’s just not as reliable a setup when it comes to how many are on at a time as I originally had thought. I’m a musician, too, so I can only count to 4, and I tend to screw up math that goes above that. ;^) I also forgot that I needed to add up clock speeds for AMD chips and not speed ratings. I have a P4 3.06 HT laptop at work and I work with 9 other teachers. Of them, 8 of them have machines that have Internet connectivity and I’m working on resolving the issues for the other one. One of them is a P4 3.06 HT and the rest are XP 2000’s. The crunching power I get out of them depends on my coworkers turning on and actually logging in. None of these ladies are super technical, so sometimes this is an issue, but my boss has specifically said that they need to do so when they get to work in the morning, these things should be starting up in the range of 7:30-8:30 AM and running until around 4PM or so. My boss’ desktop (P4 1.7Ghz) is on all day most days, too. The secretarie’s machine is a P4 2.8 Ghz and runs 24-7. The network server is a 2.8 Ghz P4 HT and also runs 24-7. Once a week, I teach computer and while those machines aren’t super powerful (1 P4 1.7Ghz and 4 PIII 800Mhz), I fire them up when I get there in the morning and let them run all day long. That’s all. :^)
SoB’s stats are somewhat misleading. When a machine that has been off is started back up, and the client starts processing a test it checked out previously, it averages the stats over the period of time since the test was checked out.
This means that a machine that processes at 2M for only 12 hours a day time will appear to have a 1M rate - and obviously it gets worse if the machine is only on for 1/3 day or whatever.
This also makes past graphs change - and it works in reverse of course. At the end of the weeke, Serpentor dumps in his MONSTER STATS and the entire team’s rate increases for the last week by ~20M or more.
I really don’t like the way SoB does the stats, but I understand why - because not every machine is set up to submit every block - sneakernet, dialup, etc. all get in the way.
On the other hand, DW, you can use sobsvc.exe on any HT machine to run 2 clients as a service. As an added bonus, this allows them to run even when the machine is not logged in (sitting at Windows login screen)… I set mine up with a -k so that it shows the clients int he system tray whenever a user logs in (it stops/restarts the clients to facilitate this)
On an HT machine, a second client doesn’t double production, but it does get another 25-50% out of it it seems.