Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Dataman]

The project is now listed as a BOINC project. However I woud not even call it Beta yet but more like Alpha. About 10% of the wu's fail due to a bug in the science where a return code is interperted by BOINC as an error when it is not. They "say" they are working on it.
The credits are very good and it is a single quorum project with immediate credit. I did ~21K in about 4 days even with a 10% error rate. Wu's run in about 5 minutes but some go as high as 20 minutes. The status bar has some problems too. They have a VERY slow download/upload which is reaking havoc with my bandwidth. :hbang:
You have to join the forum seperately (bummer) and there is almost NO activity including information from the project scientist. ???

I am probably going to suspend it soon if I do not get some answers in the forum but I thought I would let everyone know what is happening there.

Cheers!

 :aus1:

[kashi]

Thanks for letting us know about it.  

Hmm, it's only granting about 19-20 cr/hr for me, seems quite a low rate. Perhaps I have higher expectations because I usually don't crunch lower credit projects or it is unsuited to my computer. The runtime under reports to blazes too, at least on the 2 4 minute tasks I've completed. Ah well perhaps they will fix it up one day. I'll run it for a little while until I finish a few monster PrimeGrid Challenge tasks then probably switch to something else. Seems to be an inefficient application in that CPU time is sometimes much less than runtime, this is good for hot days because it runs very cool. Well compared to PrimeGrid LLR tasks at least.

Edit: Just completed some longer tasks and they granted 23-24 cr/hr so that's a little better and the longer ones reported the runtime correctly too.

Because it is now eOn2 the url has changed from that listed in the Project Overview and is now http://eon.ices.utexas.edu/eon2/

[Dataman]

Yes, I am getting about 23-24 credits/hour too. Heck, I spent three weeks at WCG badge hunting so 24 looks good after 12/hr at WCG.  I moved most cores back to Orbit.  Using 26 cores at Eon used so much bandwidth due to constant upload/downloads it was affecting my browser

 .

[kashi]

12 an hour at WCG, wow only project I've done recently that was lower than that was AlmereGrid, can't remember exactly what it was but it was abysmal.

Upload/download is OK for me, currently have HT disabled so only 4 cores and one is still on a 2.5 day PrimeGrid task.

[Dataman]

WCG has always been around 12/hr. I got about 16/hr with the water project. So what is the highest paying CPU project currently? Probably a math project.  biggrin I'm not much interested in them. I tend to stick to Chemistry and Astronomy lately.
I tend to divide projects into two categories ... those that do something and those that hope to do something. Currently at Milkyway on the GPU side and Orbit on the CPU side. I've had enough protien in my diet over the years.

Thanks for the feedback on eOn. 

 :aus1:

[kashi]

I thought I was getting more than 12/hr when we did some AAs on WCG a while ago, must have remembered incorrectly or perhaps they have dropped it again, it must use a reduce towards zero scheme because I was getting about 12-16 on my old A64 Clawhammer @ 2GHz a few years ago.

Don't know what the highest credit CPU project would be, probably AQUA I suppose. QMC has the highest credit tasks I have experienced in the last 12 months but only on certain beta tasks that are rarely issued. AQUA is too unstable for me when running GPU projects, plus it's a nuisance to configure to use less than all cores so I no longer run it.  Next would possibly be the PrimeGrid Sieve tasks, I don't run them anymore either except for Challenges because I already have the badges.

Might give the ATI OpenCL beta a go when ATI fixes the HD 5970 bug. Credit is very poor for a GPU project but it may help the developer to give it a trial for a while.

eOn is going along fine, I had a lockup when I first started but it has been OK since then. I see what you mean about the uploads, I'm only running it on 3 cores and each task has a decent sized upload. If I was running it on 8 cores it would be going to town on the uploads. :)