Dears, we are starting to take full advantage of the "long" queue. I've sent a batch of 700 WUs named "AB*", which should be 8-12h on the fastest cards.

I started a (still preliminary) batch of experiments to study the behavior of a protein likely implied in Alzheimer's disease.

By the way, if you happen to see "MDIO warnings" on redefined parameters, they are harmless.

I've sent a batch of 700 WUs named "AB*", which should be 8-12h on the fastest cards.

I must have faster-than-fastest cards then, since these "AB*" WUs are completing around 17000 sec (4h 43m) on my GTX 580 @ 850MHz, and 19000 sec (5h 17m) on my GTX 480 @ 800MHz.

[ I've sent a batch of 700 WUs named "AB*", which should be 8-12h on the fastest cards]
Hi
When can I get some more? 17000sec for 11k credits standard, 27000secs for the long [7.5hrs] for the long ones 57k credits. I love them.
Cheers
Ross*

---

I received 4 of these units so far. Three ran on my windows 7 machine with GTX 480 cards, with no problem. One ran on my windows xp machine with GTX 285 card. This one gave me a lot of problems. It caused my computer to shut down numerous times. I had to reboot the machine again and again, but the work unit kept running. I repaired my register, did a check disk, ran a virus scan and malware scan. I even under clocked the video card. Nothing. The unit finally finished sucessfully. Since I never had this problem before and the subsequent unit is working fine, I would discount the video card going bad. This unit must have had some issues.

Thanks guys for the data. We are still tweaking WU sizes and credit factors. When things settle, do expect 8-12h and a credit factor of 1.5-2 (in addition to the one for fast return, if any). Maybe I'm optimizing WUs too much ;)

Concerning machine crashes, sorry to hear. BTW, I normally run the WUs on 275s in addition to Fermis.

Just wantedd to say I do like the new long wu's. I have completed 16 with no problems at all. Most run in 15 hours on GTX260's. I have no runtime yet for GTX460's. Great job folks! Keep up the good work.
Cheers!

---

i got a acemdlong and its running at 32 hours long, im surprised at how long its taking, is it me or what, help please.

A have crunched several "AB" WU's, run time about 9:52; XP 64bit + GTX560Ti factory OC to 900MHz, driver 26726, host ID 31329. Used features: Swan_Sync=0, free core out of Boinc, priority of ACEMD & ACEMDLONG precess set to normal. GPU load about 97%.

Hi Ben.
Your longtask times seem to be all over the place; 19K, 22K, 55K, 36K.
I see a fairly constant 26000sec +/- 5% on my GTX470's. Some of this may be explained by different task types, but 19K and 22K for two TONI_AB1 tasks is quite far apart and 55K for a TONI_SMDTRYP7LONG10 sounds a bit long, I ran one and it was only slightly longer (under 28Ksec).

As you have a 500 series Fermi, I think it might be throttling itself (dropping speed if it gets too warm). If not perhaps your Boinc settings or system usage is causing such high variation in times.

Note that when you are using swan_sync and configure Boinc to only use 99% of the CPU cores (threads in your case), the GPUGrid task will use a full thread so your CPU usage will remain at 100% overall. After you set swan_sync you should restart the system. Might be an idea to check the last checkpoints of your tasks first.

Check the temperature of your system and GPU especially. You should be using temp controlling software to keep the card at a reasonable temperature (60-70deg). I suggest MSI Afterburner, as this allows you to configure the fan speed acording to the temperature (it increases as the temperature rises, and decreases when it should). A simpler but less controllable program is EVGA Precision. Both can be downloaded from freeware sites such as http://majorgeeks.com/ - restart required.

This IBUCH_2_mutEGFR task type is taking about 12h on a GTX470 (XPx86, swan_sync, free i7 thread). I think this is a good size given that it just creeps into the estimated 8-12h range and the card is at the lower end of the top Fermi cards.

Hey,

skgiven is right.
All IBUCH_?_??EGFR_110315 should take about that. They are the same length and size. 52K atoms for 2,500,000 simulation steps which is a simulation of 10 nanoseconds.

Many thanks for running acemdlong guys!

ignasi

Hi, on IBUCH EGFR my gpu is used at only 40% and it take almost 18h too finish a WU,it seems really long, on other long WU my gpu take 9h too finish a WU . Is this normal ? I have a GTX 570 thanks.

Without Swan_Sync enviroment variable and free CPU core yes.

As nenym said, use swan_sync=0 and free up a CPU core in Boin Manager.

For how to see the FAQ: Best configurations for GPUGRID thread.

Some of the IBUCH tasks tend to be a bit less resource intensive than the other tasks, but 40% utilization is too low, so improve your settings.

One other thing, Vista and Win7 tend to be about 11% slower, due to the Windows Display Driver Model (WDDM), than XP; and that's after the system is optimized for GPUGrid crunching.

Thank you very much for your support, I added the variable SWAN_SYNC and I effectively increase the load of the GPU to 62%.

Did you also configure Boinc to only use 50% of the CPU?
I get 83 to 84% GPU utilization on my GTX470s when running the long IBUCH tasks.

Points per second are very very different between the different long WU types.
These numbers come from a GTX295 on WinXP with a slight OC of shaders running at 1368 and were all returned within 24 hours. CPU = i7-920 @3.8 GHZ HT = ON, 2 threads free.

TONI_KKAL3 = .538 points per second
IBUCH_2_mutEGFR = .927 points per second point
TONI_AB1 = 1.292 points per second

elapsed times are also very different, you could make the KKALs twice as long again :-)

TONI_KKAL3 = ~7 hours
IBUCH_2_mutEGFR = ~19 hours
TONI_AB1 = ~12.5 hours

---

Thanks - Steve

I've got one acemdlong task on nVidia G105M - I've noticed the task after having run for 5 hours and reaching (IIRC) less than 3%, BOINC promised to surely finish it in less than a couple of weeks (with 4 days due-time) :-D

So I've suspended the task until I'll find this thread...

[edit]After resuming the task a few hours later, it crashed weirdly (error 0x3, <message>"System cannot find the given path"</message> + "MDIO ERROR: cannot open file 'restart.coor'").

I've updated my prefs to exclude "ACEMD for long runs". But, could the server try to estimate, whether the target machine is at all capable to finish a given task until deadline?

Peter

The G105M is not powerful enough to run GPUGrid tasks.
It only has 8 shaders, you need about 100.

The G105M is not powerful enough to run GPUGrid tasks.
It only has 8 shaders, you need about 100.

I'd say, bad wording: not powerful enough to keep given deadline. Sure, therefore I've already excluded such tasks from mu prefs.

But still, as the CPUs do not receive tasks, which will supposedly not be done in-time, (why) does the same not (yet) happen for other resources (GPUs in this case)? BOINC knows the GPU's speed estimate (in my case it's merely "39 GFLOPS peak", I just can not see it anywhere in cleartext in sched_request_www.gpugrid.net.xml), project knows the task's FLOPs cost...

Or is this just to be implemented yet? I may stand corrected...

Peter