A quick question, why does the app report that it uses 41% of the CPU? If it not too much trouble, may you look into perhaps lowering the number? perhaps closer to the magic number of 27% ?

While the new GPU app says it makes a high CPU usage on the task list, it is actually using far, far less. Almost all of the work is done on the GPU per Gipsel's impressive original optimization. A quick test shows me that it's using less than 1% of my CPU.

Oh I know, I would like to use 3 GPU's without taking out a CPU task. My system crunches only Milkway. Cheers to future stability.

Well I run 4 x 5970 GPU's for a hit of 2 - 3% for ALL system uses. I have no problem running eight climate models on my i7. Boinc reports it is using .05% CPU per GPU task.

---

A quick question, why does the app report that it uses 41% of the CPU?

Mine is showing 1.02 CPUs. Ouch.

---

A quick question, why does the app report that it uses 41% of the CPU?

Mine is showing 1.02 CPUs. Ouch.

Meh. It should be a purely cosmetic misrepresentation that can be corrected later. Are you not currently getting full use of all your CPU cores for other BOINC tasks?

A quick question, why does the app report that it uses 41% of the CPU?

Mine is showing 1.02 CPUs. Ouch.

Meh. It should be a purely cosmetic misrepresentation that can be corrected later. Are you not currently getting full use of all your CPU cores for other BOINC tasks?

It *should* be?

---

Mine says it is using 0.31 of the CPU but is using none.

I was getting no tasks when I was using the app_info, so I crunched down my old units, returned them and reset the project. Shut down BOINC, removed the files and then started back up and got 24 along with the new app and .dll file.

---

.....I was getting no tasks when I was using the app_info, so I crunched down my old units, returned them and reset the project. Shut down BOINC, removed the files and then started back up and got 24 along with the new app and .dll file.

Yes the app_info.xml is now recognised as anonymous platform and treated by the MilkyWay scheduler as a CPU. So to receive ATI tasks now with app_info.xml, you need to tick "Use CPU" in MilkyWay preferences. If you also untick "Use ATI GPU if present", BOINC will stop asking for GPU tasks needlessly. You may also need to adjust your resource share for MilkyWay to a much higher level than other CPU projects or BOINC may stop requesting MilkyWay ATI tasks if the total cache of work for all projects gets full. It still wasn't getting work properly all the time so I had to reduce my script interval from 30 minutes down to 10 minutes. That seems to have fixed it for now. Even if I get 48 short 13.7 second tasks they will still take over 10 minutes.

I also tried to fix it by running my ATI cache down to zero tasks then adding <plan_class>ati13ati</plan_class> to the app_info.xml file but it has no effect other than changing the name of the Application on the BOINC Manager Tasks tab to milkyway 0.20 (ati13ati). The MilkyWay server must be configured slightly differently to the Collatz server.

Well I run 4 x 5970 GPU's for a hit of 2 - 3% for ALL system uses. I have no problem running eight climate models on my i7. Boinc reports it is using .05% CPU per GPU task.

Yes but you are still using the 0.20 app_info.xml version.

The people who are having trouble with too great an allocation of CPU resources to support ATI processing are using the new automatically downloaded 0.21 (ati13ati) version. If they have 2 or more ATI cores doing MilkyWay then BOINC may start "stealing" CPU cores from their CPU projects.

OK then, thanks for the clarification, I had better stay where I am until it is all sorted out.

---

lol, I couldn't resist detaching and reattaching to see what happened.

Sure enough I get the .38 CPU message. However, actual usage is now bouncing between 9 and 15%. That is on two 4850 GPU's. On my extensive sample (2 WU's) the finishing times seem to be the same. At least they are downloading regularly so I will let it run.

---

The Windows 32 Bit version is now available on BOINC, please report any issues in this thread.

hello,

I have switched to serveur application 0.21 on Windows 7 64 bits + boinc 6.10.25 + catalyst 9.12.
Except serveur problems today, all wus were validated with around same cpu/gpu times.

So stay like this for long term test.

lol, I couldn't resist detaching and reattaching to see what happened.

Sure enough I get the .38 CPU message. However, actual usage is now bouncing between 9 and 15%. That is on two 4850 GPU's. On my extensive sample (2 WU's) the finishing times seem to be the same. At least they are downloading regularly so I will let it run.

It seems to be about 30 seconds slower on my 4830 from the 0.20b app.

---

The version on BOINC is actually 0.20b.

It must be the b-1 tag I was using then as I have slowed down quite a bit.

Okay, I will probably go back to the app_info and turn on CPU usage.

---

Boinc just download the new ati13 ati apps but then immediately crashed.
It managed to mangle several work units before i stopped it.

Here's the error i got. Any ideas?

running Boinc: 6.10.25
GPU: HD4850


http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=27087886

<core_client_version>6.10.25</core_client_version>
<![CDATA[
<message>
- exit code -1073741757 (0xc0000043)
</message>
]]>


Thanks Mike P.

Thanks Travis, now everything works great!!
He acknowledged the version of the BOINC client, the GPU ATI and downloaded the application and has already ground 9 WU without errors.
Very good and good work.

Boinc just download the new ati13 ati apps but then immediately crashed.
It managed to mangle several work units before i stopped it.

Here's the error i got. Any ideas?

running Boinc: 6.10.25
GPU: HD4850


http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=27087886

<core_client_version>6.10.25</core_client_version>
<![CDATA[
<message>
- exit code -1073741757 (0xc0000043)
</message>
]]>


Thanks Mike P.

What Catalyst Driver version and Operating System?

Boinc just download the new ati13 ati apps but then immediately crashed.
It managed to mangle several work units before i stopped it.

Here's the error i got. Any ideas?

running Boinc: 6.10.25
GPU: HD4850


http://milkyway.cs.rpi.edu/milkyway/result.php?resultid=27087886

<core_client_version>6.10.25</core_client_version>
<![CDATA[
<message>
- exit code -1073741757 (0xc0000043)
</message>
]]>


Thanks Mike P.

What Catalyst Driver version and Operating System?

Let me answer this.
It's Vista 32bit and Catalyst 9.12. It was running with the app_info.xml before. Maybe you have set up a wrong brook*.dll file on the server?