I'm running 3 tasks on a GTX 770 with 2gb ram, same as I run under Seti. I gave up all overclocking and all 6 cores of the 970 are running around 70C at 100% load with a water cooler. I have 4gb ram which is the max for 32bit Win7 OS.

Bob B.

Running 3 Einstein tasks on that GPU and only using a 32-bit OS with your limited system memory likely is a bad combination and a possible cause of your problem, turn the GPU down to 2 tasks and reserve at least 1 CPU core if you already havn't (two maybe needed though).

Why didn't you go with a 64-bit OS and at least twice the memory?

Cheers.

hmm, yes, this mix of 32 Bit Windows, and this Hardware is problematic for a host of accumulated reasons. These point toward the described problems/symptoms, before necessarily considering specific application(s).

Some bullet points (probably not comprehensive)
- Windoiws Vista/7 onwards moved to the WDDM ( Windows Display Driver Model) that 'virtualises' video memory.
- the described system will not be able to utilise the full 4GiB host RAM, and 2GiB VRAM, since the 32 bit address space won't address that much combined (physically) [ note no physical address extension here... ]
- something (probably both host memory and the GPU VRAM) will be being substantially capped.
- The Kernel portion will be reserving around half of whatever's left, including for various driver crash recovery & caching purposes.
- When whatever VRAM is actually allowed becomes close to filled, this will rely on the virtual memory and page to disk
- This in turn will induce delays prone to expose system resources limits/timeouts and produce various kinds of failures.

All that considered, I would venture to guess that even running 2 instances of a demanding application may become unreliable. Most likely just switching to a 64 Bit version of Windows would resolve the problem. Even still, 2GiB VRAM is a lot to mirror (As WDDM does) with only ~2GiB kernel space (leaving 2GiB for user applications, minus overheads).

Most likely, IMO, these mounting limitations go a long way toward the GPU Vendors' gradual moves toward mostly 64 bit support for the newer cards. There just simply aren't enough system resources available under 32 bit to support all the extra Driver+OS heavy lifting in the background.

How many tasks at a time on your GPU? FWIW, each GPU task requires 300-350mb of GPU RAM. So for example, if you have 1gb of GPU RAM, You should limit it to just 2 or maybe 3 tasks at a time on your GPU.

Aside from that, The things I would look into:

1) Perhaps Einstein runs your CPU hotter, and your cooling is not up to the task. Or your thermal connection to the heatsink is poor.

2) Perhaps Einstein uses more RAM, and one of your DIMMs is marginal. With the extra usage, it is exposing the issue.

So why haven't you asked at the Einstein forums? Since it's their applications that run and give you so much trouble, wouldn't their forums be the first place to ask for help on this, instead of the Seti forums?

I did reset the project a couple times with no change. After the 6th reinstall the messages about missing files finally disappeared and it downloaded 38 tasks. Perhaps things will return to normal.

Instead of uninstall/reinstall, did you try the repair option? Don't uninstall, just run the installer. When the installer sees that you are trying to install the same version over the top, it will offer a repair option. It is necessary when reverting to an earlier version, but might work on problems like this.

I uninstalled/reinstalled 7.0.64 and Lunatics 0.41 several times

This doesn't do anything of worth, you know?
Uninstalling BOINC will just remove the BOINC worker applications from the C:\Program Files\BOINC\ directory. It won't reset anything project specific, or else you would never be able to update BOINC with work in cache... and you can.

So all Reinstalling does is add those BOINC worker applications again to the C:\Program Files\BOINC\ directory. It also does nothing to reset the project or the contents of the data directory.

Uninstalling and reinstalling Lunatics does change for a moment what science applications will be used, but if you aren't running BOINC at any moment between uninstalling Lunatics and reinstalling it, it doesn't change anything either. Aside from that, the work caching and scheduling capability is something that BOINC does, not the science application, not even when they're third party. So uninstalling and reinstalling Lunatics is a complete waste of time.

So then, what does matter?
1. Always check the messages that BOINC gives as to why you don't get work.
2. Always check your account->Computers on this account->Details of the computer that has the trouble->Application details to see if there are any problems there.
3. If you can't make heads or tails of it, post about it on the forums.

Now, optionally:
4. You could try to reset the project. Although in the case of a 'you have too many errors and a quota of 1', this doesn't matter. The quota is enforced on the server, not locally.
5. You can try (backing up and) removing your Data directory. But before you do so, always better ask on the forums if this matters. In this case, I don't think so. As BOINC will just make the hostID again and after contact with the server, you'll still be on a quota of 1.

And for those who really don't want to do any of the above:
6. Just give up in disgust and wait it out for a week, do not touch the system in that time, or look at it. BOINC will always recover on its own.

Interesting - that computer (5203915) has now made contact and you have 38 tasks without the daily limit increasing. You must be on someone's good side.

Number of tasks completed 647
Max tasks per day 2
Number of tasks today 38
Consecutive valid tasks 1
Average processing rate 11.38523663292
Average turnaround time 28.86 days

After uninstall/reinstalling BOINC, you lost all custom settings, so make sure your settings for minimum work buffer are okay and the max additional work buffer is very small, like .01. Since network activity is infrequent, check that preference setting also.

Why do you recommend additional work buffer being so small? I usually specify 10 days going back to the old days when sometimes no WU's would be available for a week or longer.

And that worked well prior to BOINC v7. With v7, the definitions are different and the first number (minimum work buffer) is the one you want to set high to hedge against work shortages. The second number (max additional) influences how often it asks for work. That's supposed to be after you reach the minimum work buffer, but it doesn't always work that way. A large number there with BOINC v7 could be why this computer isn't bugging the server for work - now more than 24 hours since last contact. Just trying to get you those 2 tasks / per day since every validation will increase that temporary limit. The value of .01 is pretty widely used by forum members for their more active crunchers - a higher value might be okay for a notepad but don't use more than a day.

After uninstall/reinstalling BOINC, you lost all custom settings, so make sure your settings for minimum work buffer are okay and the max additional work buffer is very small, like .01. Since network activity is infrequent, check that preference setting also.

You have two computers showing errors. Comp 5203915 has 15 the last 4 from 18n sept. Comp 7050508 show 49 errors. last 13 from 22sept.

I have no idea what could be wrong.