It is highly recommended that Beta apps stay on beta.

I think RIchard posted something about consequences if you are caught.

So best to leave testing apps on the testing site.

I think it was about V7 app being used for V8 workunits. I've already used beta apps openly (the OpenCL one) and nobody said I should not.

A few minutes searching a- and here's Richard's warning: http://setiathome.berkeley.edu/forum_thread.php?id=78702&postid=1754529

If I see anyone else damaging the main science database by running v8 tasks using a v7 application, I will have no hesitation in asking Eric to prevent their computer downloading any new work.

Pretty clear, indeed EXTREMELY CLEAR

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

And in this very thread, too.

Running a v7 app as if it were v8 is a definite no-no, and will always remain so - all platforms, all devices. Just don't do it.

Running a Beta app on the Main project is more of a grey area. I can't pretend it's absolutely prohibited, because I'm doing it myself. But I'm doing it for a clear reason, and under controlled circumstances.

First, the apps are pre-tested before I load them here: first in an offline bench environment, and then by running live at Beta for a while. Only then would I move them across.

Secondly, I limit the number of computers I run that way - just two at the moment, one testing CUDA apps, and the other testing intel_gpu apps.

And thirdly, and most importantly, I'm watching them like a hawk. I have a window open permanently displaying my inconclusive tasks - just three for CUDA at the moment, and none for intel_gpu. Every single inconclusive task that shows up gets investigated in depth.

And it actually helps a great deal that most of the tasks I end up paired with are run by the stock apps - it cuts down on the noise considerably. When two Beta apps collide and can't agree, it takes even more investigation to work out which one is at fault.

You need to look at how long it takes to run those work units.

You might be able to run 5 but at 2.5 times the amount of time it would take to run 3 at a time.

So then you are actually being counter productive.

This all goes back to seeing how fast increasing amounts of work units on graphic cards take.

I can confirm this as fact. I was running 3 nvidia and 3 intel GPU jobs on two of my rigs(using the Ivy Bridge/Haswell features of my Intel 4770's)along with 7 cpu tasks (AP7 and S@H 7&8)and watched my RAC steadily decline. I have since disabled the CPU and the Intel GPU features in lunatics and am requesting jobs just for my nvidia GPUs, I have set them to 4 simultaneous threads to see if the efficiency rises. No work received so far but I'll post the results when I've processed some work.

---

"Sour Grapes make a bitter Whine." <(0)>

I suggest that if you really have to run an app from beta here on main to keep your card busy then choose the OpenCL one, because this one is already more tested then the Cuda ones. But you need to keep an eye on your results so you can react when you see any problems (errors, invalids, etc).

---

And here is Richards release post for CUDA on Beta:
https://setiweb.ssl.berkeley.edu/beta/forum_thread.php?id=2266&postid=55789#55789

Notice that he says:

"First availability of Windows SETI MB v8 CUDA applications for NVidia cards.

These are for use on the BETA SITE ONLY".
--------------------------------------------------------------------------
But then of course, should one want to do it anyhow, hiding ones computers would be the first step....
https://setiathome.berkeley.edu/show_user.php?userid=198885

Hmm, Science of RAC, Science or RAC....:-)
Nuff said about that....

I hadn't seen that. I don't really have a problem not using the cuda app on main - I just was not aware of the constraints...

I tried using the OpenCL app for NV but it was just not working for me. I'll just wait for it to go to main.

---

Why can't some people just wait until official releases?

Bad form I guess.

Cheers.

Why can't some people just wait until official releases?

Bad form I guess.

Cheers.

I guess the main problem is the lack of tasks for GPUs atm. Some ppl don't like to wait so they are looking for ways to keep their GPUs busy.

---

Why can't some people just wait until official releases?

Bad form I guess.

Cheers.

I guess the main problem is the lack of tasks for GPUs atm. Some ppl don't like to wait so they are looking for ways to keep their GPUs busy.

Tasks are tasks are tasks. There are plenty of tasks - what we're working on are GPU applications to process the tasks.

[quote]Why can't some people just wait until official releases?

Bad form I guess.

Cheers.

I guess the main problem is the lack of tasks for GPUs atm. Some ppl don't like to wait so they are looking for ways to keep their GPUs busy.

Tasks are tasks are tasks. There are plenty of tasks - what we're working on are GPU applications to process the tasks.

I know Richard. Ok, let me try again: I guess the main problem is that ppl are not receiving tasks which they can process on their GPUs because there are no official V8 GPU apps available right now and there are rarely any APs. That's why some ppl, who do not like to wait, look for ways to keep their GPUs busy.

I hope it's understandable now, otherwise blame my bad english but pls keep in mind that it's not my main language.

---

I think we've cleared it up between us! Now for my morning chase around the inconclusives...

Why can't some people just wait until official releases?

Bad form I guess.

Cheers.

I guess the main problem is the lack of tasks for GPUs atm. Some ppl don't like to wait so they are looking for ways to keep their GPUs busy.

The answer is very simple, do a backup project or 2 (like me) to keep you warm instead of whinging about the current state of things (it'll be fixed in time & no sooner). ;-)

Cheers.

...if you have your "quick sup" raise one in Richard's direction on my behalf.

---

Bob Smith
Member of Seti PIPPS (Pluto is a Planet Protest Society)
Somewhere in the (un)known Universe?

Ey up, that is Yorkshire talk for a quick jimmy, then I'm off down 't pub for a sup :-))

A "Jimmy" here is regarded as a Jim Beam & Coke in a stubbie, can or glass Chris, now what do you mean by it? (hopefully not a tool for for break'n'entry) ;-)

Cheers.

Now for my morning chase around the inconclusives...

Ey up, that is Yorkshire talk for a quick jimmy, then I'm off down 't pub for a sup :-))

errr - no, he was really checking his inconclusive lists for suspicious results to DL and run offline.

we are at x41zi and waiting for alpha testing to conclude before the builds get handed to Eric to deploy at beta for wider testing. it's a RC but so was x41zh - we won't know until enough tasks have run through to pronounce it good enough.

Those who grabbed x41zh from beta and are running it here, please standby to switch to x41zi when it becomes available. [and do report any anomalies you see with beta builds!]

edit: corrected version number to switch to...

---

A person who won't read has no advantage over one who can't read. (Mark Twain)

1 thing that I've noticed since the intro of V8 is that my 2500K rig that is running as an "anonymous platform" hasn't received any V7 MB resends at all, but my 3570K rig that had a failed data HDD that BOINC was running on that has now been replaced and is now running as stock gets quite a few of them.

Is it just a coincidence for me or has anyone else noticed this behaviour?

Cheers.

Why can't some people just wait until official releases?

Bad form I guess.

Cheers.

I guess the main problem is the lack of tasks for GPUs atm. Some ppl don't like to wait so they are looking for ways to keep their GPUs busy.

The answer is very simple, do a backup project or 2 (like me) to keep you warm instead of whinging about the current state of things (it'll be fixed in time & no sooner). ;-)

Cheers.

Sure, but in the end it's everyones very own and personal decision. If the ppl in charge don't want crunchers to use any unofficial apps here then they have to make that absolutly clear. If this was my project the I guess I would do that. But I can't remember any statement on that matter.

---

Those who grabbed x41zh from beta and are running it here, please standby to switch to x41zi when it becomes available. [and do report any anomalies you see with beta builds!]

or any other versions that may be announced - though at this stage it doesn't look as if they will be necessary.

But I can't remember any statement on that matter.

The project - and especially Eric - have always made it clear that they want to be as inclusive as possible, and have welcomed independent developers porting to other hardware/operating systems - ever since the launch of the Classic version.

The official statements are on this website, at Porting and optimizing SETI@home and Other sources of BOINC client software. Though note that the latter page says that all the SETI versions listed are out-of-date, because of the v8 transition.

Can't find anything here that says "Don't run any other app then the official ones". Looks rather like they encourage ppl to run whatever they like. Seems a bit risky to me but I guess they know what they are doing.

---

A few of us clearly said on different threads what can be used and what not.

---

With each crime and every kindness we birth our future.