I was surprise to see this application this morning, and no thread on the forum!

But after an upload, i don't receive another WU with this application.
So maybe this was an error!

I saw that this appliation use almost 50% of one core also.

GDF, can you tell me why i had this application?

Jim PROFIT

I also had a workunit get done with v6.48 yesterday.
However the other 5 that have downloaded since then are all v6.45.

CPU usage was a lot higher, Task manager showed 20% cpu usage (25% being 100% on 1 core), and also took longer to do the workunit.
Time per step was 54.227 ms compared to the normal of 50.955 ms

---

Down with the Kredit Kops!!!

I have two of them. One is waiting to be processed by a 8800GT, the other one is now being processed by a GTX280 and the CPU load is about 50% of one CPU core (25% total load) (Intel Core 2 Duo 8500 no OC.) I'll report results when it be completed.

EDIT: I have estimated processing speed of the 6.48 unit being procesed (GTX280 OC: clock 697, shaders 1500, mem stock) and it gives me: one WU in 5.7 Hours (about 30% faster).

I had put some out to compute with the new application for testing.
Now, it is the default one. It's also less cpu friendly than the old one but it should faster. Blame Windows, on Linux it is still 2%.

gdf

GDF, thank you for your work :)


..testing 1st 6.48 app-WU - will report after this one...

Just perfect...


http://www.ps3grid.net/result.php?resultid=85504

..same performance as the Linux app ... :-)


CPU Load @ Windows XP w/ SP3 and Q9550 @ 1-2 %


Great!

With a GTX280 on a Phenom Quad 9950 BE at 2.6 GHz under Vista 64 bit the 6.48 is using 15% of the CPU power, thats ca. 60% of the power of one core.

---

With a GTX280 on a Phenom Quad 9950 BE at 2.6 GHz under Vista 64 bit the 6.48 is using 15% of the CPU power, thats ca. 60% of the power of one core.

Same here with an Intel Q9300, GTX 260 and Vista 64.
A few times to compare -

Windows 6.45 - 43-45ms/step
Linux 6.47 - ~32 ms/step
Windows 6.48 - 34.2 ms/step for the first WU

---

pixelicious.at - my little photoblog

Hi ...

First 6.48 WU - ran in a little over 7 hours, with a step time of 30ms on my system - Vista32, Q6600 and GTX260. Lots of CPU usage though - about the same 60% use of a single CPU as seen by other posters.

P.

@GDF

how did you/your team managed the warps? My G92 has just a estimated time of 36700s overclocked 8800GT 256 MB @ 720/1800/800.
Is your new Application detecting the Cuda-Hardware and how does it makes the difference? Because the difference between GT200 and G92 is denotative, concerning the amount of registers and active warps per multicore (set of 8 streamprocessors).

Hi GDF.
The new 6.48 App is just awesome!

The first WU with 6.48 took ~6.6h on my GTX260 (time per step: 27.91)
With the 6.45 it took ~8.5h (time per step: 36.05)

The CPU load is ~30% on one Core - seems OK for me.


Great job!!!



Cu KyleFL

There is no big difference between my GTX260 and GTX280.

GTX280 Vista64 Clock rate: 1404000 khz, Time per step: 28.564 ms, WU: 24279.624 s
GTX260 XPPro64 Clock rate: 1242000 khz, Time per step: 28.580 ms, WU: 24292.703 s

Both running under Boincmanager 6.3.14 the application 6.48.

The GTX280 has 240 shader and is running with 1404000 khz, the GTX260 has 192 shader is only running with 1242000 khz. The difference in time per WU should be bigger.

---

Do you have the new GTX260 with 216 shaders?

gdf

Hi,

i have no luck with application 6.48. Last night i set up a Windows XP64 box with a 9800, newest (cuda)driver from NVIDIA, Boinc 6.3.14.

I always got this (means that all tasks are marked with a "computation error"):

18-Oct-2008 02:04:27 [PS3GRID] Computation for task ec11646-GPUTEST4-2-10-acemd_2 finished
18-Oct-2008 02:04:27 [PS3GRID] Output file ec11646-GPUTEST4-2-10-acemd_2_1 for task ec11646-GPUTEST4-2-10-acemd_2 absent
18-Oct-2008 02:04:27 [PS3GRID] Output file ec11646-GPUTEST4-2-10-acemd_2_2 for task ec11646-GPUTEST4-2-10-acemd_2 absent
18-Oct-2008 02:04:27 [PS3GRID] Output file ec11646-GPUTEST4-2-10-acemd_2_3 for task ec11646-GPUTEST4-2-10-acemd_2 absent
18-Oct-2008 02:04:27 [PS3GRID] Starting eN15112-GPUTEST3-8-10-acemd_0
18-Oct-2008 02:04:27 [PS3GRID] Starting task eN15112-GPUTEST3-8-10-acemd_0 using acemd version 648
18-Oct-2008 02:04:29 [PS3GRID] Started upload of ec11646-GPUTEST4-2-10-acemd_2_0
18-Oct-2008 02:04:33 [PS3GRID] Computation for task eN15112-GPUTEST3-8-10-acemd_0 finished
18-Oct-2008 02:04:33 [PS3GRID] Output file eN15112-GPUTEST3-8-10-acemd_0_1 for task eN15112-GPUTEST3-8-10-acemd_0 absent
18-Oct-2008 02:04:33 [PS3GRID] Output file eN15112-GPUTEST3-8-10-acemd_0_2 for task eN15112-GPUTEST3-8-10-acemd_0 absent
18-Oct-2008 02:04:33 [PS3GRID] Output file eN15112-GPUTEST3-8-10-acemd_0_3 for task eN15112-GPUTEST3-8-10-acemd_0 absent
18-Oct-2008 02:04:33 [PS3GRID] Starting Ku31408-GPUTEST4-3-10-acemd_0
18-Oct-2008 02:04:33 [PS3GRID] Starting task Ku31408-GPUTEST4-3-10-acemd_0 using acemd version 648
18-Oct-2008 02:04:34 [PS3GRID] Finished upload of ec11646-GPUTEST4-2-10-acemd_2_0
18-Oct-2008 02:04:35 [PS3GRID] Started upload of eN15112-GPUTEST3-8-10-acemd_0_0
18-Oct-2008 02:04:39 [PS3GRID] Finished upload of eN15112-GPUTEST3-8-10-acemd_0_0
18-Oct-2008 02:04:39 [PS3GRID] Computation for task Ku31408-GPUTEST4-3-10-acemd_0 finished
18-Oct-2008 02:04:39 [PS3GRID] Output file Ku31408-GPUTEST4-3-10-acemd_0_1 for task Ku31408-GPUTEST4-3-10-acemd_0 absent
18-Oct-2008 02:04:39 [PS3GRID] Output file Ku31408-GPUTEST4-3-10-acemd_0_2 for task Ku31408-GPUTEST4-3-10-acemd_0 absent
18-Oct-2008 02:04:39 [PS3GRID] Output file Ku31408-GPUTEST4-3-10-acemd_0_3 for task Ku31408-GPUTEST4-3-10-acemd_0 absent

---

I received the same errors on the new 6.48 Application, I rebooted which usually stops any errors but it still got them so I set the Box to NNW for now ... ???

18-Oct-2008 05:20:06 [PS3GRID] Started download of yHI3867-GPUTEST4-3-LICENSE
18-Oct-2008 05:20:06 [PS3GRID] Started download of yHI3867-GPUTEST4-3-COPYRIGHT
18-Oct-2008 05:20:06 [PS3GRID] Started download of yHI3867-GPUTEST4-3-yHI3867-GPUTEST4-2-10-acemd_1
18-Oct-2008 05:20:06 [PS3GRID] Started download of yHI3867-GPUTEST4-3-yHI3867-GPUTEST4-2-10-acemd_2
18-Oct-2008 05:20:06 [PS3GRID] Started download of yHI3867-GPUTEST4-3-yHI3867-GPUTEST4-2-10-acemd_3
18-Oct-2008 05:20:06 [PS3GRID] Started download of yHI3867-GPUTEST4-3-grama.ionized.pdb
18-Oct-2008 05:20:06 [PS3GRID] Started download of yHI3867-GPUTEST4-3-grama.ionized.psf
18-Oct-2008 05:20:06 [PS3GRID] Started download of yHI3867-GPUTEST4-3-parameters
18-Oct-2008 05:20:06 [PS3GRID] Started download of yHI3867-GPUTEST4-3-GPUTEST4
18-Oct-2008 05:20:06 [PS3GRID] Started download of ss30338-GPUTEST3-5-LICENSE
18-Oct-2008 05:20:06 [PS3GRID] Started download of ss30338-GPUTEST3-5-COPYRIGHT
18-Oct-2008 05:20:06 [PS3GRID] Started download of ss30338-GPUTEST3-5-ss30338-GPUTEST3-4-10-acemd_1
18-Oct-2008 05:20:06 [PS3GRID] Started download of ss30338-GPUTEST3-5-ss30338-GPUTEST3-4-10-acemd_2
18-Oct-2008 05:20:06 [PS3GRID] Started download of ss30338-GPUTEST3-5-ss30338-GPUTEST3-4-10-acemd_3
18-Oct-2008 05:20:06 [PS3GRID] Started download of ss30338-GPUTEST3-5-grama.ionized.pdb
18-Oct-2008 05:20:06 [PS3GRID] Started download of ss30338-GPUTEST3-5-grama.ionized.psf
18-Oct-2008 05:20:07 [PS3GRID] Finished download of yHI3867-GPUTEST4-3-LICENSE
18-Oct-2008 05:20:07 [PS3GRID] Finished download of yHI3867-GPUTEST4-3-COPYRIGHT
18-Oct-2008 05:20:07 [PS3GRID] Finished download of yHI3867-GPUTEST4-3-yHI3867-GPUTEST4-2-10-acemd_3
18-Oct-2008 05:20:07 [PS3GRID] Finished download of yHI3867-GPUTEST4-3-GPUTEST4
18-Oct-2008 05:20:07 [PS3GRID] Finished download of ss30338-GPUTEST3-5-LICENSE
18-Oct-2008 05:20:07 [PS3GRID] Finished download of ss30338-GPUTEST3-5-COPYRIGHT
18-Oct-2008 05:20:07 [PS3GRID] Started download of ss30338-GPUTEST3-5-parameters
18-Oct-2008 05:20:07 [PS3GRID] Started download of ss30338-GPUTEST3-5-GPUTEST3
18-Oct-2008 05:20:08 [PS3GRID] Finished download of ss30338-GPUTEST3-5-ss30338-GPUTEST3-4-10-acemd_3
18-Oct-2008 05:20:08 [PS3GRID] Finished download of ss30338-GPUTEST3-5-GPUTEST3
18-Oct-2008 05:20:09 [PS3GRID] Finished download of yHI3867-GPUTEST4-3-parameters
18-Oct-2008 05:20:09 [PS3GRID] Finished download of ss30338-GPUTEST3-5-parameters
18-Oct-2008 05:20:12 [PS3GRID] Finished download of yHI3867-GPUTEST4-3-yHI3867-GPUTEST4-2-10-acemd_1
18-Oct-2008 05:20:12 [PS3GRID] Finished download of yHI3867-GPUTEST4-3-yHI3867-GPUTEST4-2-10-acemd_2
18-Oct-2008 05:20:12 [PS3GRID] Finished download of ss30338-GPUTEST3-5-ss30338-GPUTEST3-4-10-acemd_1
18-Oct-2008 05:20:14 [PS3GRID] Finished download of ss30338-GPUTEST3-5-ss30338-GPUTEST3-4-10-acemd_2
18-Oct-2008 05:20:15 [PS3GRID] Finished download of ss30338-GPUTEST3-5-grama.ionized.pdb
18-Oct-2008 05:20:23 [PS3GRID] Finished download of yHI3867-GPUTEST4-3-grama.ionized.pdb
18-Oct-2008 05:20:24 [PS3GRID] Finished download of yHI3867-GPUTEST4-3-grama.ionized.psf
18-Oct-2008 05:20:25 [PS3GRID] Starting yHI3867-GPUTEST4-3-10-acemd_0
18-Oct-2008 05:20:26 [PS3GRID] Starting task yHI3867-GPUTEST4-3-10-acemd_0 using acemd version 648
18-Oct-2008 05:20:27 [PS3GRID] Computation for task yHI3867-GPUTEST4-3-10-acemd_0 finished
18-Oct-2008 05:20:27 [PS3GRID] Output file yHI3867-GPUTEST4-3-10-acemd_0_0 for task yHI3867-GPUTEST4-3-10-acemd_0 absent
18-Oct-2008 05:20:27 [PS3GRID] Output file yHI3867-GPUTEST4-3-10-acemd_0_1 for task yHI3867-GPUTEST4-3-10-acemd_0 absent
18-Oct-2008 05:20:27 [PS3GRID] Output file yHI3867-GPUTEST4-3-10-acemd_0_2 for task yHI3867-GPUTEST4-3-10-acemd_0 absent
18-Oct-2008 05:20:27 [PS3GRID] Output file yHI3867-GPUTEST4-3-10-acemd_0_3 for task yHI3867-GPUTEST4-3-10-acemd_0 absent
18-Oct-2008 05:20:32 [PS3GRID] Finished download of ss30338-GPUTEST3-5-grama.ionized.psf
18-Oct-2008 05:20:33 [PS3GRID] Starting ss30338-GPUTEST3-5-10-acemd_0
18-Oct-2008 05:20:33 [PS3GRID] Starting task ss30338-GPUTEST3-5-10-acemd_0 using acemd version 648
18-Oct-2008 05:20:34 [PS3GRID] Computation for task ss30338-GPUTEST3-5-10-acemd_0 finished
18-Oct-2008 05:20:34 [PS3GRID] Output file ss30338-GPUTEST3-5-10-acemd_0_0 for task ss30338-GPUTEST3-5-10-acemd_0 absent
18-Oct-2008 05:20:34 [PS3GRID] Output file ss30338-GPUTEST3-5-10-acemd_0_1 for task ss30338-GPUTEST3-5-10-acemd_0 absent
18-Oct-2008 05:20:34 [PS3GRID] Output file ss30338-GPUTEST3-5-10-acemd_0_2 for task ss30338-GPUTEST3-5-10-acemd_0 absent
18-Oct-2008 05:20:34 [PS3GRID] Output file ss30338-GPUTEST3-5-10-acemd_0_3 for task ss30338-GPUTEST3-5-10-acemd_0 absent

I received the same errors on the new 6.48 Application

The output of Stoffelstasks shows the "incorrect function" error, whereas you get "<message> - exit code -1073741819 (0xc0000005)</message>]]>". Don't know what that means though.

MrS

---

Scanning for our furry friends since Jan 2002

CPU Time down from about 3000 seconds to 40

GPU time up from 31.000 up to 35.000

NOT Acceptable.

GTX 260, Vista 64 Bit SP1, Boinc 6.3.14

see here for an example result
http://www.ps3grid.net/workunit.php?wuid=60790

Do you have the new GTX260 with 216 shaders?

gdf

Yes, you're right. Just checked with GPU-Z, it's a GTX260² with 216 Shaders.

---

Well, the new 6.48 WUs are fine for me:

GTX280 OC: clock: 697; shaders: 1500; mem: stock
Windows Vista Home Premium 32 bits
CPU: Intel Core 2 Duo 8500 3.16 (stock clock)

Before (6.45) (a fast one):
# Time per step: 28.998 ms
# Approximate elapsed time for entire WU: 24648.684 s

Now (6.48)
# Time per step: 23.334 ms
# Approximate elapsed time for entire WU: 19833.684 s

As you can see, the new one goes 24% faster and you have to take into account that I have posted one of my best speeds in 6.45. If I use a slower one:

Before (6.45) (a slow one):
# Time per step: 30.281 ms
# Approximate elapsed time for entire WU: 25738.960 s

Then, the new one goes 30% faster.

CPU Time down from about 3000 seconds to 40

GPU time up from 31.000 up to 35.000

NOT Acceptable.

GTX 260, Vista 64 Bit SP1, Boinc 6.3.14

see here for an example result
http://www.ps3grid.net/workunit.php?wuid=60790

The WU you're linking to was crunched with 6.45, not 6.48. You're currently getting times between 37 and 41 ms/step.. which will surely improve once you get 6.48-WUs.

Regarding CPU-usage: you had a unit with ~40s before. On my machine CPU time increases to ~1h if I use it interactively, otherwise I get times in the tens of seconds.

MrS

---

Scanning for our furry friends since Jan 2002