Hi,

AP on Titan V vs GTX 1080. I run 2 at a time (gpu_usage = 0.5).

http://setiathome.berkeley.edu/workunit.php?wuid=2915879436


Petri

A question about this - your APs are running Anonymous, as are mine on 1080s. But mine run ~100% CPU of elapsed time in addition, whereas yours are running much less CPU. I would like to know what I am doing wrong, if anyone would hazard a guess. Typical APs seem to run around 1/2 hour on mine; I run 3 WUs/GPU, with no explicit controls on how many APs or MBs, and with no explicit command line, since APs have been so rare until recently.

Thanks!

Hi,
I'm running with
<cmdline> -verb -st -nog -unroll 80 -sbs 2048 -ffa_block 2304 -ffa_block_fetch 1152 -oclFFT_plan 256 16 256 </cmdline>

The unroll is high because I run the VOLTA and 1080 with the same cmdline. For 1080 I'd suggest 20 or 28 for the unroll.
I have not tested the ffa_* parameter values with my current cards. The must date back to 980's or 780's.

One reason why I have so low cpu usage is that I run a special library file to replace Yield() with nanosleep. It is called libleep.so and it is using linux LD_PRELOAD feature. The source code and instructions are buried here in the forums a couple of years back. Maybe the -sleep does the same.

Petri

---

To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones

I managed to get 44 for the 30th March UTC.

Cheers.

15 greeted here this morning and strangely 11 of them are CPU tasks, usually those numbers are reversed in favour of the GPU's.

Cheers.

16 in the end arrived here for the 31st March UTC.

Cheers.

I acquired this card just for the
a) fun,

Don't need to say anything else, havin fun is all.
Hope you have a lot of fun with it.

+1

---

25 for the 1st UTC and 7 so far for the 2nd UTC.

Cheers.

I acquired this card just for the
a) fun,

Don't need to say anything else, havin fun is all.
Hope you have a lot of fun with it.

+1

The a) was the right choice and I'm certainly having fun.
With no WUs to process I have had time to test and optimize MB CUDA code.

---

To overcome Heisenbergs:
"You can't always get what you want / but if you try sometimes you just might find / you get what you need." -- Rolling Stones

19 arrived here for the 4th March UTC.

Cheers.

Unless another file gets loaded I've got 58 for the 5th April UTC.

Cheers.

18 arrived here for the 6th April UTC so far.

Cheers.

Picked up 46 in the last 2 days though some were very minuscule...

---

"Sour Grapes make a bitter Whine." <(0)>

starting crunching AP till yesteday

23 done, 8 to do dl from 5th to 6th of april

---

Wow, I managed to snare a few AP's & this is the result?

5 Apr 2018, 15:29:12 UTC 9 Apr 2018, 6:55:31 UTC Completed and validated 2.59 0.58 1.42 AstroPulse v7 v7.03 (sse) windows_intelx86

I wound up with 2 resends on the 7th, 3 on the 8th, 3 on the 9th along with 8 fresh jobs and 7 so far for the 10th April UTC.

Cheers.

Likewise, 8 yesterday and 2 so far today.

---

SETI@home classic workunits 4,019
SETI@home classic CPU time 34,348 hours

I'm up to 25 26 now for today.

Cheers.

I'm up to 37 now.

Cheers.

I'm up to 37 now.

I wound up with 39 for the 10th April UTC as the count has begun for today.

Cheers.

Just the 3 AP's so far today for me :)

245 channels of AP yet to be processed. Could be a record haul for some.

---

Grant
Darwin NT