Are these short or long tasks generally?

Maybe off-topic, but when is it planned to start exactly?
____________
--------
"MiauiKatze" is german and means as much as "mewing cat"!
--------

Will it use the conventional double-checking system, i.e. two identical tasks to two users?

Will we be starting from where the last PG search started?

That would mean that we trust the work that other people have done. Do I seem like a trusting person? :)

I predict we will find two Wieferich primes in work unit 0. /JeppeSN

The one with the first WU will get the prime found badge :)
____________
--------
"MiauiKatze" is german and means as much as "mewing cat"!
--------

Or the very first WU that surpasses the previously searched range is prime.

Looks like preferences are now available just waiting on actual units.

Rick, come join the Discord.

Are these short or long tasks generally?

Around 980 - 1000 sec for a Radeon VII. (16min)

2730 sec for Vega 56.(45min) Ewww.
____________



Слава Україні!

4911 sec (83min) R9 280X
____________



Слава Україні!

Running on a Gtx 1050, I've got a task on 6420s, but most fell in the 6750~6950s range.

Greetings

Running a GTX 1660 Super - averaging between 860-910 sec depending on workunit.

Regards

Ryzen 5950X 16c 122 minutes, 2*8c 2*257 minutes
RX 5700XT 2100 seconds

830 seconds using a free Google Cloud virtual PC running Linux on Intel(R) Xeon CPU @ 2.20GHz and a Tesla T4 (4095MB) GPU.
____________

For a CPU time, my Broadwell i5 2.9 GHz had a 136k second run time on 2 cores, 246k second CPU time
____________
Just a poor Mac user with no fancy GPUs

Tested GFN CPU transforms

GT 710 - 107000 seconds

Yes, I run it as kind of joke. Running this thing on 4-core CPU was more then 3 times efficient (31000 seconds).

GT 710 - 107000 seconds

Yes, I run it as kind of joke. Running this thing on 4-core CPU was more then 3 times efficient (31000 seconds).

I've still got a GTX 285 running. I was going to see if it could handle a task.

I am pretty sure that it won't run it. I think you will need a Fermi card or newer to run WW as the elements used in the app require a compute capability that was introduced in those cards.

I am have trouble figure out my 1660Ti. It is running about 1600 sec a task, this is about twice the time that Scott Brown has stated for his 1660Ti. Looking at GPU-Z it shows that my card is running at half capacity. I tried updating my driver, but that has not helped.

Am I missing something that I should be looking at to solve my dilemma?
____________
Werinbert is not prime... or PRPnet keeps telling me so.
Badge score: 12x3 + 1x4 + 2x6 + 2x7 + 1x8 + 1x9 + 1x10 = 93

Thanks for the response, unfortunately I am not moving away from win 7 anytime soon. I will just have to live with the increase in WU_Prop hours. :-)
____________
Werinbert is not prime... or PRPnet keeps telling me so.
Badge score: 12x3 + 1x4 + 2x6 + 2x7 + 1x8 + 1x9 + 1x10 = 93

Thanks for the response, unfortunately I am not moving away from win 7 anytime soon. I will just have to live with the increase in WU_Prop hours. :-)

830 seconds using a free Google Cloud virtual PC running Linux on Intel(R) Xeon CPU @ 2.20GHz and a Tesla T4 (4095MB) GPU.

Well a solution to my problem that works is to run two tasks simultaneously. The result is one task every 800 seconds or so. I no longer feel sad due to GPU envy. Eventually I will need to look into why my GPU was under-performing, but I will leave that for another day.
____________
Werinbert is not prime... or PRPnet keeps telling me so.
Badge score: 12x3 + 1x4 + 2x6 + 2x7 + 1x8 + 1x9 + 1x10 = 93

830 seconds using a free Google Cloud virtual PC running Linux on Intel(R) Xeon CPU @ 2.20GHz and a Tesla T4 (4095MB) GPU.

Is that just the free trial or like free-free?

When the project was new, I downloaded one workunit to try as I was interested in the new project. According to the "task status" page (if I am reading the graph correctly), one workunit should take about 0.271 hours to run on GPU (= 16 minutes). However, after running the task for 20 minutes using my computer GPU, only about 1% of the task was completed, and the estimated remaining time was over 1 day.

As I was not running any other CPU or GPU tasks or programmes during that time, I was not sure what caused the task to run so slowly, so I aborted the task. Does anybody know what is causing this 100-fold increase in task duration? Also, how should I fix this problem?

I see...

Originally I thought it was a problem with my computer running this specific task because my GPU can run PPS-Sieve tasks (3000 credits) in ~1 hour 40 minutes, so I thought WW tasks (4000 credits) would take ~2 hours.

But what you said makes sense, my computer is nearly half as old as myself so it does not have modern fast processors... just that I did not expect it to be 100 times more slow than the new processors (that's like a tortoise (1 km/h) vs a car!)

I have downloaded and began to run one PPS-Sieve task using GPU. Progress is 3.725% at the 4 minute mark (there was a small delay before the percentage started to increase). I have work to do on my computer so the task may proceed slightly slower, but right now the estimated time from start to finish is ~1 hour 47 minutes.

Resources used are 0.0116 CPU + 1 Nvidia GPU (default settings, I did not change anything)

Ok, the PPS-Sieve task is finished. Elapsed time: 1 hour, 43 minutes and 58 seconds; CPU time of 13 minutes and 34 seconds

Runtimes and credit are intended to be similar to AP27.
Credit will be 4000 credits per task, unless we change the task size.

Now that WW tasks have been completed, are you considering modifying the credit?
- AP27 --- 27.5 hours average CPU time --- 4,043 credit
- WW --- 77.4 hours average CPU time --- 4,000 credit
- ESP --- 78.3 hours average CPU time --- 11,575 credit

Decided to run one on my old laptop with 2 cores - currently been running for 4d 23:56 and at 85% complete. :) Yeah, I know it's kinda silly, but wanted to see how long it would take.

I am have trouble figure out my 1660Ti. It is running about 1600 sec a task, this is about twice the time that Scott Brown has stated for his 1660Ti. Looking at GPU-Z it shows that my card is running at half capacity. I tried updating my driver, but that has not helped.

Am I missing something that I should be looking at to solve my dilemma?

Most likely, it is because you are on Windows 7 rather than Windows 10. Windows 10 allows for certain aspects of the WW app to work more efficiently than Windows 7 (indeed, this was causing significant screen lag on Windows 7 before the app went fully public and coding adjustments had to be made).

Is there minimum requirements for GPU?

It seems to be that of my old GT 620's and 710 doesn't get any WW gpu workunits.

In order to be sent WW tasks for an Nvidia GPU, all of the following must be true:

* The compute capacity (i.e., the features of the GPU, not its speed) must be at least CC 2.0

* The CUDA version (this is part of your video driver) must be at least 3.2

* The video driver version must be at least 175.0

* The amount of video ram on the GPU must be at least 1.5 GB

It appears to be the last one that is preventing you from getting tasks, because both of your GPUs have 1 GB of memory.

There's a good chance the CPUs in those computers are faster than the GPUs. Nvidia GPUs ending in "10" or "20" are bargain devices designed to provide basic video for CPU/motherboard combos that lack built in video, especially for office computers. They're very slow.

I've lowered the requirement down to 1 GB. You should be able to run it now.

It turns out that the amount of memory you need is proportional to the speed (really, the number of cores) of the GPU. So while larger, faster GPUs *do* need 1.5 GB, slower/smaller GPUs like yours do not. So it's safe to lower the limit. GPUs that will actually need that much VRAM because they have so many cores will have enough VRAM to handle it.

Until someone inexplicably produces a GPU with 20,000 CUDA cores and half a GB of VRAM. :)

EDIT: I've removed the memory requirement for WW altogether.

Okay, first wu in progress, have to look tomorrow how long it will take, afk.

Thanks :)

Job cache is too small for this project. I am using an RTX 3090 FE GPU and each work unit takes about 4 minutes and 45 seconds to complete. Ever thought about greatly increasing the job cash for us high end GPU owners?

Job cache is too small for this project. I am using an RTX 3090 FE GPU and each work unit takes about 4 minutes and 45 seconds to complete. Ever thought about greatly increasing the job cash for us high end GPU owners?

What exactly do you mean by "job cache"? Most of the queue and buffer controls that affect job flow are actually set by the user rather than by the admins.

If your computer isn't requesting work, that's a setting on your side. You may need to look at the messages in your BOINC log to see whether or not your computer is requesting work when it contacts the server.

If it is requesting work and you're not getting any, unless you're trying to grab an ungodly number of tasks (which is a terrible strategy and you shouldn't be doing that), then the problem is likely that you have run out of wingmen and all the available tasks are from workunits where you've already run a task, and are thus ineligible to get another task. If this is the problem, there's little that can be done about it other than convincing other people to run WW or to diversify your task selection.