My wife is a graduate student working on molecular modeling for HIV. So when she isn't using her computers, I sneak into the office and run BOINC. She keeps the computers running all the time, so why not use those extra cycles? Where I am located, most of our electricity comes from nuclear, so its still relatively cheap.

But dont forget. You should run Boinc only on computers, where you got the permission to run it.

---

Just curious for my own info since I've been thinking of limiting my own computers for costs reasons if the feature that limits cpu usage under processor usage

Use at most xxx% of cpu time

is still something that users with multiprocessor systems using 6.6.36 should not use?

This could be another limited way for me to cut back on cost if available along with

On multiprocessor system, use at most xxx% of processors

since it seems to me running at 100% cpu time also adds to the costs regardless of the percentage of the cpu's I use.

---

Just curious for my own info since I've been thinking of limiting my own computers for costs reasons if the feature that limits cpu usage under processor usage

Use at most xxx% of cpu time

is still something that users with multiprocessor systems using 6.6.36 should not use?

I'm not sure if anything in that direction has been done in BOINC, but the problem seems to persist, depending on the project application. Einstein@home seems to be affected mostly.

But there are other tools that achieve the task more elegantly. Look for ThreadMaster or TThrottle at the BOINC add-on software page.

Gruß,
Gundolf

---

Computer sind nicht alles im Leben. (Kleiner Scherz)

SETI@home classic workunits 3,758
SETI@home classic CPU time 66,520 hours

With my 2 P4's getting less and less credit for the wu's and the electric bill is going up it is soon not very exciting to have them running at all.

---

With my 2 P4's getting less and less credit for the wu's and the electric bill is going up it is soon not very exciting to have them running at all.

Don't know why they'd be gettting less & less credit unless there was something wrong with them. Since the end of June my RAC has actually increased from around 1,900 to around 2,200.

---

Grant
Darwin NT

he's means that it's taking them forever to complete a task... that another computer say a quad can knock at 2-3 times faster.. and a gpu that can knock it out 5 times faster.

---

I recommend Secunia PSI: http://secunia.com/vulnerability_scanning/personal/
Go Georgia Tech.

he's means that it's taking them forever to complete a task... that another computer say a quad can knock at 2-3 times faster.. and a gpu that can knock it out 5 times faster.

It's one thing for it to take longer and longer (my system is the same), it's another to be getting less & less credit.

---

Grant
Darwin NT

Recently, my RAC on one P4 dropped from 350 towards 150, judging by the credit received for the last two weeks. On my quads the credit received has gone up like yours. As far as the P4 goes it gets about 28.76 for 17,614 seconds of work.

---

Or, as it seams, an NVIDIA can complete 100 times faster.

---

Recently, my RAC on one P4 dropped from 350 towards 150, judging by the credit received for the last two weeks. On my quads the credit received has gone up like yours. As far as the P4 goes it gets about 28.76 for 17,614 seconds of work.

That's the problem I have with RAC as a rating system. It doesn't reflect the amount of work a machine actually does in a given time frame unless all of the work you are doing is validation work.

This makes it rather hard to benchmark a system for seti and find the best performance per watt configuration.

I have a machine that I stopped running tasks on over a month ago, as it was just for testing, and it still gets some credit a few times a week form pending tasks.

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

The cost of hydro is real. I just set up my first ever dedicated machine tonight. However I did set it to stop crunching between 3 pm and 7pm which are the highest energy use times in my area. I figure there's no point adding to the strain of the grid. However... I do have the windows update run during that time. Too bad I can't program my PC to auto start-up at certain times.

---

Recently, my RAC on one P4 dropped from 350 towards 150, judging by the credit received for the last two weeks.

What about pending credit?
With longer processing times, on slower machine RAC will drop in the short term, but continue to rise in the longer term (till it gets to what's the general "average"). Pending credit will climb sharply.
If after a month or so the RAC hasn't come back up to roughly where it was- there is a problem with your machine no longer processing as much work as it used to.

---

Grant
Darwin NT

Too bad I can't program my PC to auto start-up at certain times.

You can if your bios supports it. Check your bios for a "wake on alarm" or "power on by alarm" setting.

Thank you Grant, I will keep an eye on the RAC before i retire it.

---

Recently, my RAC on one P4 dropped from 350 towards 150, judging by the credit received for the last two weeks. On my quads the credit received has gone up like yours. As far as the P4 goes it gets about 28.76 for 17,614 seconds of work.

I think I see where he is coming from. The recent rash of shorties has lowered his CPU seconds per credit on his P4. A 92 credit WU was taking on average 43,000 CPU seconds, which comes out to under 500 CPU seconds per credit, while the shorties are getting 28.75 credit in roughly 17,500 CPU secs...an average of over 600 CPU seconds per credit.

EDIT: That doesn't account for the 50% drop in RAC for that box, but still...

I just checked my Core2Duo and my ratios are 97 and 83, respectively, so the shorties have actually improved my CPUsec/credit ratio a bit.

---

"Life is a tragedy for those who feel, and a comedy for those who think."

"I never get into an argument that I cannot win."

I wonder if my P4 is running hotter than before and thus less efficient?

---

It's entirely possible. Dust buildup around the fans can prevent them from moving the heat fast enough, thus causing your CPUs to run hotter and less efficient.

Summertime heat can also have an effect if the ambient room temperature is rather warm.

---

It's entirely possible. Dust buildup around the fans can prevent them from moving the heat fast enough, thus causing your CPUs to run hotter and less efficient.

Summertime heat can also have an effect if the ambient room temperature is rather warm.

I changed out a fan on an older system last night as it had started to make a lot of noise. After pulling the fan off I then pulled the dust off of the heat sink in a mat. It was both funny and disturbing. I have to remember to clean out my older machines as often as I do my main one.

As I figured I'd clean other systems while I was at it. My MODT server, which I pretty much haven't opened up since I built it, actually had cobwebs inside of the case.

---

SETI@home classic workunits: 93,865 CPU time: 863,447 hours
Join the [url=http://tinyurl.com/8y46zvu]BP6/VP6 User Group[

It's entirely possible. Dust buildup around the fans can prevent them from moving the heat fast enough, thus causing your CPUs to run hotter and less efficient.

Summertime heat can also have an effect if the ambient room temperature is rather warm.

I changed out a fan on an older system last night as it had started to make a lot of noise. After pulling the fan off I then pulled the dust off of the heat sink in a mat. It was both funny and disturbing. I have to remember to clean out my older machines as often as I do my main one.

As I figured I'd clean other systems while I was at it. My MODT server, which I pretty much haven't opened up since I built it, actually had cobwebs inside of the case.

YUK!! you mean a spider has been living in your server... MY GOD

---

I recommend Secunia PSI: http://secunia.com/vulnerability_scanning/personal/
Go Georgia Tech.

Too bad I can't program my PC to auto start-up at certain times.

Why not?

Many modern BIOSes have settings to power on at a specific time, and most operating systems have scheduler or "chron" daemons that allow programs or events to be run by schedule.

---