I take it that boinc calculates your credit per trickle by using the benchmark results. If your benchmark result is higher does that mean you get more credit for each trickle completed?

I'm a whore for stats, <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/forum_thread.php?id=14">reference</a>.

Hi,
On CPDN the credit per trickle is always the same: 94.52 credits
A finished Wu will give you 6805.26 credits.
AFAIK, Benchmarks results don't affect credits.
Cheers...

---

Arnaud

Why do benchmarking then?

AFAIK benchmarking doesn't determine your CPU work load because you set that yourself in the general preferences via no. of CPUs to use and amount of processing usage. Moreover, I don't think it confirms your computer suitability for the CPDN when such low spec computers are reported to be crunching these huge WUs and taking months. Just image the high probability of failure with time. So why benchmark?

CPDN is not benchmarking, Boinc is. :-)
All other projects do use the benchmark to calculate your credit, so if you run multiple projects on one PC the benchmark is needed.

I thought I would calculate my credit per trickle and I get 114.28... for this <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/show_host_detail.php?hostid=86671">computer</a>.

This 2.6 GHz P4 HT computer is crunching its first 2 models with a total credit of 6994.30 to date. It is half way through both models i.e. in phase 2.

total trickles = ((1stPhase = 259248 trickles) + (2ndPhase = 140426 trickles)) x 2 models

total trickles for computer = 799348

&gt; total trickles = ((1stPhase = 259248 trickles) + (2ndPhase = 140426 trickles))
&gt; x 2 models
&gt;
&gt; total trickles for computer = 799348

Not quite. That's your number of timesteps. The trickle interval is every 10802 timesteps, which means you've made 74 trickles (you could find that out by following the link for each result, which show that each has made 37 trickles).

Your credits have actually been rounded up from 6994.295. Dividing that by 74 gives the actual credits per trickle of 94.5175

---

"The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer

Thanks for that. I thought I must have mis-calculated and I probably made the most silly mistake, lol. Then I tried to correct it by dividing the timesteps by the credit ... oh dear ... when I should have been doing credits per trickle.

For a long time , I had only participated in CPDN-Boinc.
After some time I added SETI-Boinc with only 10% daily usage.
During that time, I did not pay attention to any difference about credits given
by each project. I figured they were about the same for equal amounts of CPU usage.

Now that LHC and Einstein are up and running, I have joined them also. I placed Einstein 60% and divided the other three among the remaining 40%. To my surprise, I notice after a few days that the graphs showing my total BOINC credits was climbing at a higher rate than when I only had CPDN &amp; SETI.

That when I started looking at the average hourly rate of credit for each of my four projects. I found that SETI, LHC, and Einstein had about the same rate but CPDN's rate was about 69% of the others.

Does anyone know why CPDN grants credits at such a low rate?

---

<img src="http://www.boincsynergy.com/images/stats/comb-793.jpg">

&gt; Does anyone know why CPDN grants credits at such a low rate?

Yes. But the answer is on the 'dead' forum. It was posted by Carl Christensen, the programmer who was attempting
to synthesize a credit similar to the other projects.

Here, the only goal is a completed model, with 72 trickles along the way. So Carl worked out a value to 'award'
for each trickle. Unfortunatly, when some models were completed, a large number of extra credits were awarded.
This upset the credit hunters who didn't get the extra. Carl attempted to fix this, (forget how), and things got worse, so previous values were restored from a backup.
In the end, just before his contract expired, Carl wrote a 'mickey mouse' program which runs about every 4 hours, and re-calculates total credits. So you MAY get a huge number of credits on completing a model, only to lose them again a few hours later.

At one point also, Carl realised that the total credits for a job was 'off' a little, compared with other projects, and increased the increment a little.

Or, to misquote Siegfried from "Get Smart" (in a loud voice to a henchman):
'This is NOT SETI! VE do NOT hunt crrredits here!'

Les