Where did you place the BOINC folder? If you would have placed it in your Home folder, as per instructions, you wouldn't need any special permissions.
https://boinc.berkeley.edu/wiki/Installing_BOINC#The_Berkeley_Installer

Here is an example. The archive is downloaded to the desktop. It is then moved to the home directory (~). Finally BOINC is unpacked and installed. All of this can be done within a regular user account; root privileges are not needed.
This creates a directory called BOINC/ under the home directory containing the following files...

Actually, those instructions are a little off. You can simply drag the download to your Home folder, and after the dependency is installed just double click on 'The BOINC Manager' to launch it.
With the BOINC folder in your Home folder, Ubuntu is about the easiest system to configure BOINC on. No Mac permissions, and No Windows hidden folders.

BOINC didn't ask were to go (I don't think) It's living in var/lib/boinc-client at the moment, with 'BOINC core client' as the owner.

I will run out my tasks and try a reinstall and see it I can 'bring it home' :)

Brent, you don't have to install it. Just download from the boinc site, unpack it and move the whole folder to your home directory (or move first and unpack then). Then run the manager from terminal. At least that's how I did it and it runs fine.

---

Hi Folks,
update from my machine with linux special cuda 6 app.
RAC is up to 8900 and still rainsing.
Everything is running fine with this app.

My RAC is still raising too.
Started on 14th dec at 3100cr/day with my i3-2100T now with the 750ti is runs 6800cr/day (reduced to 18h daily)

https://setiathome.berkeley.edu/show_host_detail.php?hostid=1931980

---

Regards from nothern Germany
Roman

SETI@home classic workunits 207,059
SETI@home classic CPU time 1,251,095 hours

Good to hear Rock, I have Ubuntu 14.04 running now and just burning off these stock tasks before attempting the special app.

Let's make the 1070 and 980 smoke :)

Well the 980 is a hybrid, it will be fine :)

Is the Cuda info not showing a driver version a concern before switching to the new apps?
I have been running Cuda60 and SoG on stock apps with 1 new Invalid against SoG

Invalid

Wed 28 Dec 2016 06:12:52 AM CST | | CUDA: NVIDIA GPU 0: GeForce GTX 1070 (driver version unknown, CUDA version 8.0, compute capability 6.1, 4096MB, 3992MB available, 10204 GFLOPS peak)
Wed 28 Dec 2016 06:12:52 AM CST | | CUDA: NVIDIA GPU 1: GeForce GTX 980 (driver version unknown, CUDA version 8.0, compute capability 5.2, 4038MB, 3898MB available, 8245 GFLOPS peak)
Wed 28 Dec 2016 06:12:52 AM CST | | OpenCL: NVIDIA GPU 0: GeForce GTX 1070 (driver version 367.57, device version OpenCL 1.2 CUDA, 8113MB, 3992MB available, 10204 GFLOPS peak)
Wed 28 Dec 2016 06:12:52 AM CST | | OpenCL: NVIDIA GPU 1: GeForce GTX 980 (driver version 367.57, device version OpenCL 1.2 CUDA, 4038MB, 3898MB available, 8245 GFLOPS peak)

Heat is certainly a problem in this mid tower case :(

None of the Linux BOINC Apps posted by Berkeley show the CUDA driver version. It hasn't been a problem yet, so, I don't expect it ever will. The CUDA driver version is shown in the later versions of BOINC, only available from the Linux repositories, it's the same as the OpenCL version anyway and that is shown in the earlier BOINC versions.
The current Windows OpenCL App on Main has a known problem with Overflows, the Baseline CUDA App doesn't have that problem, so, I'd say you were probably Robbed on that Invalid Overflow. Too bad it's too late to grab the task and test it on a CPU. It's an Overflow though, and will Never be used by SETI to look for ET or anything else, so, it doesn't really matter. The Newer Windows OpenCL App handles the Overflows better, but as with the New Mac nVidia OpenCL App, it's still stuck at Beta. Maybe soon the Newer Apps will make it to Main.

Does anybody know why Boinc reports my VRAM with 1024MB on Windows but only 977MB on Linux?

---

When you look at NVIDIA Settings in Windows what does it show for Total Memory and Total Dedicated Memory?
I'm guessing your Linux NVIDIA Settings is showing 1024 for Total and around 977 for Total Dedicated Memory, that's the way it is on my nVidia cards anyway. BOINC is not showing the Total Memory.

Where can I find the AP application files that the sample app_info in the CUDA60 download refer to?

Looking for:
astropulse_7.08_x86_64-pc-linux-gnu__opencl_nvidia_100
AstroPulse_Kernels_r2751.cl

I found the CPU version, but not for GPU

EDIT: Found them, I should have searched the forum, Duhh

What was I thinking? why didn't I put ny motherboard in the new case which has much much better cooling, and would have been easier to do - tomorrows project.

Got the code purring away GPU temps up 5C, 980@43C, 1070@83C

Runtimes are nice ~ 6:20 on 980, 7:20 on 1070.

Everything really went well other than all the problems with my i7, the i5 works great. And new motherboard ordered, likely Tuesday at the earliest for that.

Now for more reading to try and find some fine tuning.

Nicer on Arecibo tasks 4:20 on 980, 4:40 on 1070.

And haven't see the CPU over 7% for 2 tasks.

When you look at NVIDIA Settings in Windows what does it show for Total Memory and Total Dedicated Memory?
I'm guessing your Linux NVIDIA Settings is showing 1024 for Total and around 977 for Total Dedicated Memory, that's the way it is on my nVidia cards anyway. BOINC is not showing the Total Memory.

Hm, I would have to stop crunching on Linux to boot into Windows. Don't wanna do that right now but from your answer it looks like it's not uncommon and therefore not a problem.

Small Update after a few days with the new app: Everything's fine, have crunched 800+ tasks sucessfully, no Error, no Invalid. RAC was ~5800 in the last days running openCL, currently it's ~7700 and still rising. Nice work, TBar & Petri!

---

My Inconclusives seem just a tad high :(
State: All (504) · In progress (271) · Validation pending (59) · Validation inconclusive (14) · Valid (159) · Invalid (0) · Error (1)

That's ok, I had 22 for some time, all valid until now. [/url]

---

Yea it's OK in the end, but if 10-20% of task need to be rechecked that does seem a little counter productive project wise.

Sure, but since our tasks are valid it's not a problem with this app here.

I think a problem exists with the apple darwin apps, I see those a lot in my inconclusives.

---

Sure, but since our tasks are valid it's not a problem with this app here.

I think a problem exists with the apple darwin apps, I see those a lot in my inconclusives.

Don't forget that credit granting and ultimate validation is quite "forgiving". Task will be counted as valid even if some signal it reports isn't. Provided there are many other valid signals + invalid one in "best" area.
So, ultimate validation doesn't mean that all is OK.

Actually, we not in the stage to decide if app is OK or not. This stage already passed. Jason confirmed missing parts in Pulse signal reduction stages. So we know app is broken in that part.
Low invalids ratio just ensures that bug manifests itself in quite low number of cases. Actually it will show up only if there are few reportable Pulses in same PoT. Cause even one such pulse is rare event having 2 in same PoT and in such order that incorrect one will be reported - even more rare event.

---

SETI apps news
We're not gonna fight them. We're gonna transcend them.

Sure, but since our tasks are valid it's not a problem with this app here.

I think a problem exists with the apple darwin apps, I see those a lot in my inconclusives.

Don't forget that credit granting and ultimate validation is quite "forgiving". Task will be counted as valid even if some signal it reports isn't. Provided there are many other valid signals + invalid one in "best" area.
So, ultimate validation doesn't mean that all is OK.

Actually, we not in the stage to decide if app is OK or not. This stage already passed. Jason confirmed missing parts in Pulse signal reduction stages. So we know app is broken in that part.
Low invalids ratio just ensures that bug manifests itself in quite low number of cases. Actually it will show up only if there are few reportable Pulses in same PoT. Cause even one such pulse is rare event having 2 in same PoT and in such order that incorrect one will be reported - even more rare event.

A bit deeper digging verified a race on the power of 2 folds, which are running in parallel onto a single result set. So when a reportable pulse is detected in one fold, and higher score in the other, without synchronisation it's a coin toss as to which pulse is recorded (as observed).

The traditional (serial) algorithm treats the higher score as a refinement of the prior detection. Needs a bit more walkthrough, but probably i'll just separate the pulse results, and see if adding the reduction CPU side is too costly or not. My feeling is that they're rare enough to not have much noticeable performance impact, though if the increased bus consumption becomes a problem, then adding a GPU side reduction step shouldn't be a huge step.

Either way I'll probably end up undoing the (long time ago) nvidia coded triplets-first rearrangement along the way, restoring to CPU serial final order. The spikes, gaussians, and auto-correlations are somewhat paralleled already anyway, so the synch tap points for
these needed reductions are already there. Just probably need a bit of juggling and an intermediate store.

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

Since a new revision is needed, Downloads have been suspended until the next revision is ready.
I think 35 is enough testers for now. If you find something other than the Pulsefind Race problem please report it here.