Thank you for your detailed help. Although it is great for energy conservation, cost less, and costs less to operate, my CPU does not have hyperthreading. Perhaps because of that I saw a decline in WUs processed per day when I operated two per GPU. The CPU %, as indicated by Boinc Tasks, declined from 99.xx to 66.xx. I had to return to one WU per GPU. Nevertheless, I greatly appreciate the time you took. Thanks again.
HyperThreading wasn't even in my equation... Neither of the two systems that I have crunching here have HyperThreading, either. My MAC is Quad Core, and my XP Pro x64 system is Dual Core. I choose NOT to crunch CPU Units because they take SOOOOO long to complete compared to GPU Units.
If you were to utilize my app_config.xml file, (and in Web Settings, select NOT to use CPU), you would be able to crunch 2 Units at a time... It's totally up to you, though. Whatever you choose will be respected.
Good luck, and God bless. Always - Happy Crunching.
[EDIT:]
Also, in looking at your computers, it seems that one of your Xeon CPUs is listed as "3 Core". For that system, (to keep one core free to run the system), I'd make the Max_Concurrent Line "4". This would allow 2 CPU Cores to feed two of the three GPUs in that system 2 Units at a time. The third GPU would NOT crunch.
For the Quad Core CPU system, make Max_Concurrent "6", and ALL three GPUs in that system would crunch 2 Units at a time.
The two 750TI SC cards are, (as always), crunching 2 Units at a time.
Pending Units are 4 with one marked Inconclusive. The Inconclusive Unit shows the first Wingman with a Validate Error... Don't know what to make of that, and how that made my Unit "Inconclusive"...
Invalid Units are now at 1. (MAC OpenCL bug still prevalent here...)
[XP Pro x64 Update:]
The 1.18 Units are crunching away on the 760 card and have 0 Errors, 0 Inconclusives, and 0 Invalids.
[RAC Update:]
RAC has FULLY recovered!!! Actually, RAC has improved over the BRP4G stats that I had. Currently at 107.7 K RAC.
Since RAC has fully recovered, I will cease with RAC Updates.
As always, I will continue monitoring and report on the MAC OpenCL issues with Inconclusives and Invalid Units. I will also continue monitoring and report on 1.18 Units on my Win system; but, for now, those seem bulletproof.
My first 1.18 tasks just finished on my GTX970 and I have to say WOW!
It took 24.5 minutes while 1.17 tasks usually take about 47-51 minutes, that's a 50% speedup! I'm running x2 and the 1.18 task was paired with a 1.17 task.
I also observed that it still uses a full CPU core as support.
If this speedup holds for all FGRP-GPU tasks then I suspect that some adjustments to the estimates needs to be done again or my DCF will start swinging more wildly. My DCF right now is at 0.84 and the FGRP-GPU tasks, both 1.17 and 1.18, are estimated at about 54 minutes.
I posted the above on 16 Jan, about 2 weeks ago.
Runtime for my host is about 28 min running x2 with recent tasks and a DCF of 1.0 would have them estimated @ 1h05m, that drives the DCF down to less than 0.5 with the effect that Boinc requests to much CPU work. As I keep a small cache of work it hasn't had any adverse effects but it would be nice if things where a bit less volatile.
Has there been any though about adjusting the estimated speed of the v1.18 app since it's release?
Server status says: 37 % work still remaining (22 days estimated)
I haven't monitored that constantly, but progress seems to go steadily onward. How will this search continue in near future? Will there be plenty of new work available soon or how is this thing working? I'm curious about the future plans
Server status says: 37 % work still remaining (22 days estimated)
I haven't monitored that constantly, but progress seems to go steadily onward. How will this search continue in near future? Will there be plenty of new work available soon or how is this thing working? I'm curious about the future plans
The server status only knows about the work that is currently in the BOINC system. We are currently preparing new datasets that will be added when they are ready. The search covers a big parameter space where we only do little slices at a time. So there is enough work for the foreseeable future available.
Vasishk-Taneya wrote:
whats up with the credits system?
Nothing. The claimed credit column is a remnant of BOINC that we don't use. Just ignore it. It was based on CPU time which obviously doesn't work for GPU tasks.
CElliott
)
HyperThreading wasn't even in my equation... Neither of the two systems that I have crunching here have HyperThreading, either. My MAC is Quad Core, and my XP Pro x64 system is Dual Core. I choose NOT to crunch CPU Units because they take SOOOOO long to complete compared to GPU Units.
If you were to utilize my app_config.xml file, (and in Web Settings, select NOT to use CPU), you would be able to crunch 2 Units at a time... It's totally up to you, though. Whatever you choose will be respected.
Good luck, and God bless. Always - Happy Crunching.
[EDIT:]
Also, in looking at your computers, it seems that one of your Xeon CPUs is listed as "3 Core". For that system, (to keep one core free to run the system), I'd make the Max_Concurrent Line "4". This would allow 2 CPU Cores to feed two of the three GPUs in that system 2 Units at a time. The third GPU would NOT crunch.
For the Quad Core CPU system, make Max_Concurrent "6", and ALL three GPUs in that system would crunch 2 Units at a time.
TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join SETI Refugees
[MAC Update:] The two 750TI
)
[MAC Update:]
The two 750TI SC cards are, (as always), crunching 2 Units at a time.
Pending Units are 4 with one marked Inconclusive. The Inconclusive Unit shows the first Wingman with a Validate Error... Don't know what to make of that, and how that made my Unit "Inconclusive"...
Invalid Units are now at 1. (MAC OpenCL bug still prevalent here...)
[XP Pro x64 Update:]
The 1.18 Units are crunching away on the 760 card and have 0 Errors, 0 Inconclusives, and 0 Invalids.
[RAC Update:]
RAC has FULLY recovered!!!
Actually, RAC has improved over the BRP4G stats that I had. Currently at 107.7 K RAC. 
Since RAC has fully recovered, I will cease with RAC Updates.
As always, I will continue monitoring and report on the MAC OpenCL issues with Inconclusives and Invalid Units. I will also continue monitoring and report on 1.18 Units on my Win system; but, for now, those seem bulletproof.
TL
TimeLord04
Have TARDIS, will travel...
Come along K-9!
Join SETI Refugees
TimeLord04 wrote:I will cease
)
Warmly appreciated.
-----
Holmis wrote:My first 1.18
)
I posted the above on 16 Jan, about 2 weeks ago.
Runtime for my host is about 28 min running x2 with recent tasks and a DCF of 1.0 would have them estimated @ 1h05m, that drives the DCF down to less than 0.5 with the effect that Boinc requests to much CPU work. As I keep a small cache of work it hasn't had any adverse effects but it would be nice if things where a bit less volatile.
Has there been any though about adjusting the estimated speed of the v1.18 app since it's release?
Server status says: 37 % work
)
Server status says: 37 % work still remaining (22 days estimated)
I haven't monitored that constantly, but progress seems to go steadily onward. How will this search continue in near future? Will there be plenty of new work available soon or how is this thing working? I'm curious about the future plans
whats up with the credits
)
whats up with the credits system?
Richie wrote:Server status
)
The server status only knows about the work that is currently in the BOINC system. We are currently preparing new datasets that will be added when they are ready. The search covers a big parameter space where we only do little slices at a time. So there is enough work for the foreseeable future available.
Nothing. The claimed credit column is a remnant of BOINC that we don't use. Just ignore it. It was based on CPU time which obviously doesn't work for GPU tasks.
I promoted the 1.18 out of
)
I promoted the 1.18 out of Beta status (for GPUs w. >=1GB) which should solve the work shortage.
I plan to soon release 1.20, which is an updated version 1.17 for "smaller" GPUs.
BM
Hi Bernd, Any plans for Mac
)
Hi Bernd,
Any plans for Mac updates on the horizon? Still working harder than others while 1.17 is current on macos...
Many thanks for your energy!
Kailee.
Bernd Machenschalk wrote:I
)
DCF should also be more stable with the new plan classes.
BM