2018 SETI-Wow! Event

You get a free pass... this time.
squint.gif
 
So i am starting to bunker GPU tasks now on my 1080ti's Trying out the app config and it seems to make things worse... went from 6 min times to 30 mins... (basically 1 every 15 mins)
GPU usage is all over the place anyone else running Nvidia cards on this right now that can chime in? doing .5 .5 cpu/gpu

Edit: also testing 1 / .5 and not much difference.
 
motqalden said:
So i am starting to bunker GPU tasks now on my 1080ti's Trying out the app config and it seems to make things worse... went from 6 min times to 30 mins... (basically 1 every 15 mins)
GPU usage is all over the place anyone else running Nvidia cards on this right now that can chime in? doing .5 .5 cpu/gpu

Edit: also testing 1 / .5 and not much difference.
Click to expand...

I ran it on my gtx760 and I noticed an increase in time as well on the anon platform.
 
hmm. it might just be the particular WU i am using. some interesting reading here:
https://setiathome.berkeley.edu/forum_thread.php?id=80113#1808582

Seems kinda complicated TBH

Snippet i was reading:

"You initially enquired about the GPU task times being all over the place.
Since early spring 2016, S@h is now processing data from a second telescope (Green Bank) other than Arecibo. The current GPU apps were made for Arecibo data.
The Green Bank tasks all have: blc...guppi...vlar in the filenames, and are generally referred to in the forum as: guppi

There are roughly 4 types of tasks that have different run times on nVidia high-end and mid-range GPUs.
1. shorties (mostly Arecibo but also some guppi) are usually less than a minute in duration.
2. Arecibo VLARs: these account for less than 10% of tasks and have the worst times on the CPU and nVidia GPUs.
These usually give the perception of having the worst credit/time because of #3&4 below.
3. guppi: somewhere between 35-50% of tasks received in a day (by my experience during the last few months). These will negatively impact the throughput of your GPU when running multiple tasks in parallel with any of the Cuda## apps from Lunatics v0.44.
4. the other Arecibo tasks (that don't have VLAR in their task/filename): these I refer to as nonVLARs (but the gurus use other labels that is more confusing for people who aren't up-to-date on the S@h forum techno-monbo-jumbo).
These currently account for 50%+ of daily tasks received (by my experience).

To optimize overall throughput of rigs with nVidia GPU(s), the ideal scenario is to process all Guppis on CPU cores and all nonVLARs on GPU, and that is why there is Mr. Kevvy's guppiRescheduler.exe available (and I made a Windows7to10 front-end script to automate the other steps before and after his file runs)."
 
Are you running the SoG (OpenCL) optimized app or the CUDA optimized app? I see better performance from the SoG app on all my Nvidia GPU’s, however you must reserve a CPU thread for each running GPU WU. With this setup, I consistently see WU completion times of ~4:00min on my 1080ti and ~5:40min on my 1070ti’s.
 
RFGuy_KCCO said:
Are you running the SoG (OpenCL) optimized app or the CUDA optimized app? I see better performance from the SoG app on all my Nvidia GPU’s, however you must reserve a CPU thread for each running GPU WU. With this setup, I consistently see WU completion times of ~4:00min on my 1080ti and ~5:40min on my 1070ti’s.
Click to expand...

Ok thank you i got this now. huge difference.
 
Almost finished bunkering my GPU tasks. Almost a month into this and i finally figure out better easier ways to do it and get the kinks worked out.... I COULD have downloaded most of these a month ago and i am kicking myself as i will be fighting for wingman after the first 6k or so are uploaded.... Good news is i learned alot about this process and BOINC in general. Just when you think you know alot you find out that you are only scratching the surface.....

This is what i have now, but only about 7k of this is completed.


State: All (19320) · In progress (18671) · Validation pending (140) · Validation inconclusive (1) · Valid (257) · Invalid (0) · Error (251)
Application: All (19320) · AstroPulse v7 (0) · SETI@home v8 (19320)
 
Last bunkered instance just starting up. Everything should be complete by competition start.
 
RFGuy_KCCO said:
Still waiting on the CPU. Supposed to arrive today (great price, but slow shipping), but I am out of town until Monday. I plan to build it Monday evening and have it crunching SETI by Tuesday morning at the latest.
Click to expand...
TR 1950X #3 is now online and currently crunching a BURP CPU task and a SETI GPU task. BURP should run out of work by tomorrow, so all my boxes will then be crunching SETI 100% of the time by then.
 
From the chat box on seti

Aug. 16, 08:52 zombu2: oh wow changing the mem clock and gpu clock to p0 state speeds changed my time from 150-170 seconds to 107​

Not sure how that will be done safely in Linux; not being in overclocking too much; but seems even better results ... with all the risks attached
 
That's because a lot of teams are still bunkering. lol

But yeah, 3rd place right now is pretty good.
 
You know they're not participating in Formula-BOINC...

Anyone thinking what I am thinking?
 
You must log in or register to reply here.
Back
Top