Among the advantages are reduced upload files.

Among the advantages are reduced upload files.

Thank GDF :-)

PS: I live in the USA. We haven't discovered high speed internet yet...

Ditto here is Canada. It's not just the speeds it's the caps on monthly bytes transferred. Angela Merkel for Prime Minister!

---

BOINC <<--- credit whores, pedants, alien hunters

How do we identify WUs that have reduced upload times?

---

How do we identify WUs that have reduced upload times?

You can identify by the the application's version number. 6.18 is the latest in the long queue (the previos version was 6.17)

How do we identify WUs that have reduced upload times?

You can identify by the the application's version number. 6.18 is the latest in the long queue (the previos version was 6.17)

However, I don't see any reduction in the upload size with the 6.18 client.

Two of my machines are failing every v6.18 WU:

�� - exit code -1 (0xffffffff)

Both ran perfectly with v6.17.

One is a GTX 460, the other a 650 TI. Any ideas?

Two of my machines are failing every v6.18 WU:
�� - exit code -1 (0xffffffff)
Both ran perfectly with v6.17.
One is a GTX 460, the other a 650 TI. Any ideas?

Installed NVIDIA v310.90 drivers and reset project. Still every v6.18 WU fails. Isn't anyone else seeing this? Of my GPUs, 3 work with v6.18 and 2 don't. All were perfect with v6.17.

DUH, nevermind, found the problem: Comodo AV.

That's good news you found the problem, one thing I noticed is that my GPU usage is all over the place. On my GTX670s and 680 it dips as low as 67%, on the lone GTX560 I'm still using it stays right around 99%. Anyone else notice this? It doesn't seem to effect the run times too much though.

Mine are all pretty steady at 88-89% so far.

Two of my machines are failing every v6.18 WU:

�� - exit code -1 (0xffffffff)

Both ran perfectly with v6.17.

One is a GTX 460, the other a 650 TI. Any ideas?

Same here, but it has only happened on my xp computer, so far. Windows 7 seems to be ok right now. See link:

http://www.gpugrid.net/result.php?resultid=6549840

http://www.gpugrid.net/result.php?resultid=6552217

Here is the log information:

2/28/2013 1:32:41 AM | GPUGRID | Computation for task App120_r1-TONI_AGGd8-12-100-RND4947_1 finished
2/28/2013 1:32:41 AM | GPUGRID | Output file App120_r1-TONI_AGGd8-12-100-RND4947_1_1 for task App120_r1-TONI_AGGd8-12-100-RND4947_1 absent
2/28/2013 1:32:41 AM | GPUGRID | Output file App120_r1-TONI_AGGd8-12-100-RND4947_1_2 for task App120_r1-TONI_AGGd8-12-100-RND4947_1 absent
2/28/2013 1:32:41 AM | GPUGRID | Output file App120_r1-TONI_AGGd8-12-100-RND4947_1_3 for task App120_r1-TONI_AGGd8-12-100-RND4947_1 absent
2/28/2013 1:32:54 AM | GPUGRID | Starting task Ann090_r1-TONI_AGGd8-14-100-RND3960_0 using acemdlong version 618 (cuda42) in slot 2
2/28/2013 1:32:55 AM | GPUGRID | Started upload of App120_r1-TONI_AGGd8-12-100-RND4947_1_0
2/28/2013 1:32:55 AM | GPUGRID | Started upload of App120_r1-TONI_AGGd8-12-100-RND4947_1_4
2/28/2013 1:33:05 AM | GPUGRID | Finished upload of App120_r1-TONI_AGGd8-12-100-RND4947_1_0
2/28/2013 1:33:05 AM | GPUGRID | Started upload of App120_r1-TONI_AGGd8-12-100-RND4947_1_7
2/28/2013 1:33:06 AM | GPUGRID | Finished upload of App120_r1-TONI_AGGd8-12-100-RND4947_1_7
2/28/2013 1:38:09 AM | GPUGRID | Sending scheduler request: To report completed tasks.
2/28/2013 1:38:09 AM | GPUGRID | Reporting 1 completed tasks, not requesting new tasks
2/28/2013 1:38:13 AM | GPUGRID | Scheduler request completed


2/28/2013 6:27:38 AM | GPUGRID | Computation for task Ann090_r1-TONI_AGGd8-14-100-RND3960_0 finished
2/28/2013 6:27:38 AM | GPUGRID | Output file Ann090_r1-TONI_AGGd8-14-100-RND3960_0_1 for task Ann090_r1-TONI_AGGd8-14-100-RND3960_0 absent
2/28/2013 6:27:38 AM | GPUGRID | Output file Ann090_r1-TONI_AGGd8-14-100-RND3960_0_2 for task Ann090_r1-TONI_AGGd8-14-100-RND3960_0 absent
2/28/2013 6:27:38 AM | GPUGRID | Output file Ann090_r1-TONI_AGGd8-14-100-RND3960_0_3 for task Ann090_r1-TONI_AGGd8-14-100-RND3960_0 absent
2/28/2013 6:27:51 AM | GPUGRID | Starting task pp165_r2-TONI_AGGd8-21-100-RND4169_1 using acemdlong version 618 (cuda42) in slot 2
2/28/2013 6:27:52 AM | GPUGRID | Started upload of Ann090_r1-TONI_AGGd8-14-100-RND3960_0_0
2/28/2013 6:27:52 AM | GPUGRID | Started upload of Ann090_r1-TONI_AGGd8-14-100-RND3960_0_4
2/28/2013 6:28:03 AM | GPUGRID | Finished upload of Ann090_r1-TONI_AGGd8-14-100-RND3960_0_0
2/28/2013 6:28:03 AM | GPUGRID | Started upload of Ann090_r1-TONI_AGGd8-14-100-RND3960_0_7
2/28/2013 6:28:06 AM | GPUGRID | Finished upload of Ann090_r1-TONI_AGGd8-14-100-RND3960_0_7
2/28/2013 6:33:08 AM | GPUGRID | Sending scheduler request: To report completed tasks.
2/28/2013 6:33:08 AM | GPUGRID | Reporting 1 completed tasks, not requesting new tasks
2/28/2013 6:33:12 AM | GPUGRID | Scheduler request completed

From windows 7 computer:

http://www.gpugrid.net/result.php?resultid=6549243

Check to see if your antivirus or antispyware is blocking the new app, That was my problem, but didn't happen on every machine, just 2.

Only new WUs will have reduced upload sizes. The old ones will be the same.
We need to tell the code to output in a different format.
Soon all will be using the compressed format.

gdf

Not sure if there's more work being done by the new app or the WUs are different or maybe v6.18 is less efficient; but so far my WU times are 5-7% slower on all 5 GPUs.

it should be the same on the same workunits.

gdf

Check to see if your antivirus or antispyware is blocking the new app, That was my problem, but didn't happen on every machine, just 2.

What did you actually do with your antivirus / antispyware that fixed this problem?

And the more important question is that version 6.17 was working just fine, I didn't do anything different to this computer, along comes version 6.18, and I get errors on one of my computers, so what is the problem with 6.18?



http://www.gpugrid.net/results.php?hostid=30790


My other computer is working just fine.

What did you actually do with your antivirus / antispyware that fixed this problem?

I marked the file as trusted (Comodo AV). You can also mark rhe whole directory as trusted.

And the more important question is that version 6.17 was working just fine, I didn't do anything different to this computer, along comes version 6.18, and I get errors on one of my computers, so what is the problem with 6.18?

http://www.gpugrid.net/results.php?hostid=30790

My other computer is working just fine.

Your problem might be different as in my case the WUs would fail immediately. Yours are failing with the same exit code but are running for a good length of time first.

Your problem might be different as in my case the WUs would fail immediately. Yours are failing with the same exit code but are running for a good length of time first.

That might be the difference between an 'on access' AV check and a scheduled scan. You often have to manage the two separate sets of exclusions independently.

it should be the same on the same workunits.

gdf

OK, I have more data now, and found a commonality. All 5 have 1 NV GPU and 1 ATI. On the 3 machines running POEM on the ATI GPU:

Machine 1, GTX 460: 8-9% slower with v6.18
Machine 2, GTX 460: 3-9% slower with v6.18
Machine 3, 650 TI: 7-8% slower with v6.18

On machines running Collatz or Moo on the ATI GPU:

Machine 4, GTX 460: the same with v6.18
Machine 5, GTX 460: the same with v6.18

Interesting, the difference may be that POEM is much more CPU intensive than either Collatz or Moo. For some reason there seems to be more contention between GPUGrid & POEM (fighting for CPU time?) with v6.18.