It seems my own favorite backup project, Einstein@home, can't be considered as reliable.
Second time already it has server crash simultaneous with SETI problems.
What backup projects do you use? (please, with one sentence aim description if possible).

power button

Primegrid, ABC and MilkyWay, however I run them at equal resource shares. PG and ABC because they have efficient 64bit apps, MW because it runs on my ATI card. PG and ABC are reliable.
BR,

---

My first backup is an additional 7 days cache of Seti.
Unfortunately the second backup is Einstein.
Cosmology sits at 3rd, but definitely not that reliable.

I suppose my backup projects tend to be ad-hoc, attached only when turbulence expected at S@H.

---

Primegrid seems pretty relaible. You may get the fame of finding some huge Prime number.

Spinhenge - reliable and nice small work units make for a good backup.

Otherwise, take your pick. You can channel surf and pick what takes your fancy.

Ian

Either Prime Grid or WCG

---

Climate Prediction. Tasks take hundreds of days, with points received for intermediate reports. Plus, the graphics are kind of cool.

---

Einstein, Rosetta and WCG on all, and for the fastest computer also includes CPDN.

---

"I make so many mistakes. But then just think of all the mistakes I don't make, although I might."

AQUA, QMC, CPDN and CPDN Beta, Einstein, LHC. This last is practically dead.AQUA simulates a quantum computer, QMC is about quantum chemistry via MonteCarlo method, CPDNs are climate models, Einstein searches gravitational waves but uses also Arecibo radio data to hunt double pulsars. LHC used to model particle trajectories in the CERN collider. AQUA also uses CUDA.
Tullio

---

Einstein, primegrid and poem..

---

I recommend Yoyo and primegrid.
If you are going to do Primegrid I would suggest picking projects that have short WU's Some of their WU's take longer than an Astropulse WU! SO be careful what you pick. They do show the average run time for each project so it's pretty easy to see which are long and short

Yoyo is a great project and always has available work. I recommend the ECM project on YoYo. Its WU's can run from a couple minutes to 5-6 hours.

---

In a rich man's house there is no place to spit but his face.
Diogenes Of Sinope

Climate Prediction. Tasks take hundreds of days, with points received for intermediate reports. Plus, the graphics are kind of cool.

Yep, gotta love those graphics in CPDN!

---

Overclock with the MSI G31M3-L and Intel E8600 3.33Ghz
Intel D865GLC Socket 478 Motherboard
~How To Overclock The Eee ASUS 1005HA Netbook To 1.9Ghz~

The set I work on for times such as this:
Cosmology
Einstein
Milkyway
Rosetta

I can't tell you when the last time I ran out of work with this set of applications.

I do set the shares to favor SETI above all else so that when SETI is down, I still keep all four cores running flat out.

YMMV

John

---

Climate Prediction. Tasks take hundreds of days, with points received for intermediate reports. Plus, the graphics are kind of cool.

Yep, gotta love those graphics in CPDN!

Of course, it depends on your PC´s performances and capabilities to calculate the time needed for each chunk of data.....
but mine takes about one week to crunch the packages, when crunching all the time (24/7).
The report/deadline is quite long, usually one year ahead, so you have plenty of time if you have a slow PC or not too much time available.
One thing that makes me wonder is the date shown on the screensaver:
it shows a past date (two or three decades ago), which makes not too much sense, since this project tries to figure out climate predictions (of the past?)....

---

Milkyway.

---

One thing that makes me wonder is the date shown on the screensaver:
it shows a past date (two or three decades ago), which makes not too much sense, since this project tries to figure out climate predictions (of the past?)...

If I recall correctly, yes, they predict the past (at least partly) to see if the calculations match reality.

Gruß,
Gundolf

Of course, it depends on your PC´s performances and capabilities to calculate the time needed for each chunk of data.....
but mine takes about one week to crunch the packages, when crunching all the time (24/7).
The report/deadline is quite long, usually one year ahead, so you have plenty of time if you have a slow PC or not too much time available.
One thing that makes me wonder is the date shown on the screensaver:
it shows a past date (two or three decades ago), which makes not too much sense, since this project tries to figure out climate predictions (of the past?)....

I think that it checks the correctness of the model. The input data are known from the past meteorological records. If the model is correct it should give as a result today's data and then predict the future data.
Tullio

---

Raistmer, as skildude noted, yoyo; has cool things going on (golumb ruler, evolution, particle accel design, and elliptic curve factorization), very dependable as backup, and if you care about these things you get neat virtual badges; good forum support as well. http://www.rechenkraft.net/yoyo/

and in case you have a ps3 looking for work, their cruncher ogr utilizes the SPEs. those running their ps3 247 have rac @ ~2500. I get ~2200 rac. pretty good for $399.

---

One thing that makes me wonder is the date shown on the screensaver:
it shows a past date (two or three decades ago), which makes not too much sense, since this project tries to figure out climate predictions (of the past?)....
[/b]

Climate Prediction is my alternate project as well. Depending on the models run they can range from 1 to 2 year analysis to 180 year analysis. I think the dates shown on some of the models are used to compare to historicals actuals to see how the models compared to actual history. This way models prove there viability in future forecasting? (Dont quote me on that - I may be way off base there) Other Models run a 45 year spin up with very little changes in enviromental conditions then the model increases the CO2 amounts (or other pollutants) and run another 45 years to see the temperature and precipitation differences. I'm currently running one that stretches over 180 years and also includes ocean temperature variances. It has 1887:49:15 elapsed time and 877:25:10 to completion. Credits are awarded virtually daily based on amount completed and this particular model will probably award between 45000 and 50000 credits when done.

Edit: Oh and BTW: They to are working through some server issues but hopefully are near the end of them as I understand a new server is on order.

---

Rick
**************************

I've not found a project that interested me enough to be a backup project. I crunch SETI and SETI Beta and that's about it.

I'm also more like the average cruncher than the majority of posters: I don't turn computers on just for crunching, I run BOINC on computers that are on because they're needed for other purposes.

If you do have some project as a backup, you do have a small obligation to that project to do some work -- and complete what you get.

"Reliable" is relative. From the project side, one of the goals is to avoid the high costs of high-reliability (redundant servers, redundant connections, etc.). The BOINC client gets around that by supporting multiple projects.

If two projects aren't enough to keep you crunching full-time, add a third (or a fourth).

---