Please visit donation page to help the project cover running costs for this month

Toggle Menu

Join PrimeGrid

Returning Participants


Leader Boards



1) Message boards : Problems and Help : What does "Work done for this project" mean? (Message 136348)
Posted 12 days ago by Profile mikey
Obviously, it has to do with what you've accomplished on your computer with PrimeGrid. For example mine says "Work done for this project: 32,934". But what does that mean? 33,000 what? In what units is it measuring in?

I downloaded BOINC yesterday so please excuse my lack of knowlegde

One more thing each Boinc Project has their own amount of Cobblestones they award us for the work done for them. Meaning Project A will give you X amount of Cobblestones/Credits for each workunit you finish and gets validated while Project B will award Y Cobblestones for the same amount of work on it's workunits. What that means is only compare Cobblestones/credits within a Project or sub-project not from one Boinc Project to another.

Credits/Cobblestones can be a way of enticing people to crunch here or there so if one Project wants lots of people they may award 100 times as many Credits for the same amount of work that another Project will. One billion Cobblestones/Credits and $10 will get you a cup of coffee at Starbucks, but without the $10 you are NOT getting that cup of coffee.

Essentially Credits show the commitment a user has to a particular Project, for instance if you look at Honza's numbers they have 2.6 Billion credits and their RAC is1.1 Million. RAC is like a speedometer that adjusts depending on your contribution over the short term.
2) Message boards : Number crunching : 2020 Challenge Schedule! (tentative) (Message 136103)
Posted 26 days ago by Profile mikey
Hi, As big hitters always tend to achieve higher scores, much by the sheer power brought to bear, it could be good to have a graded competition. Thus smaller, or individual crunchers would compete in their group size. I.E. 1-5 pcs, 6-10, 11+ You get the idea? This is what we do with our pedigree Aberdeen Angus cattle herd competions.

This is a great idea!

Regardless of whether or not this is desirable, it seems very difficult to implement.

Does a 10 year old 1c/2t Atom-based netbook and a brand new 18 core XE-series AVX-512 CPU both count as "1 PC"?

Does a computer that runs 1 task and then is turned off, and an identical computer that runs 500 tasks during the challenge both count as "1 PC"?

If you fire up a 32-core cloud server and that counts as "1 PC", what if you fire up 8 cloud servers with the same CPU cores as the 32-core server, but each of the 8 servers only has 4 cores. Does that count as "8 PCs" or "1 PC"? It's the same 32 cores as the big server.

I could go on and on with questions like that. Pandora would be proud of this suggestion!

Feel free to chat about this as much as you want, but understand that this idea isn't going anywhere. There's no way to implement it fairly.

What if instead of pc's you used cores? ie under 10 cores, 11 to 35, 36 to 80 etc? That way my hundred plus cpu cores wouldn't affect someone crunching just as diligently with 2 quad core machines. Now I have no idea how this could be monitored or even IF it could be.
3) Message boards : Problems and Help : Using more than one core when specifying one thread for LLR (Message 136063)
Posted 27 days ago by Profile mikey
In an effort to be nice to my wife's ears and decrease the fan noise from one of my computers, I have been trying to run less on it than it is capable of. The computer has a 9700k in it, and I have set ncpus=6 in the cc_config.xml.

I've noticed that when running PPSE with a single thread set in the PrimeGrid preferences that the Windows Task Manager shows each of the primegrid_cllr.exe processes using 1/6 of the CPU. I would expect each one to only be using 1/8 of the CPU since I have set them to use a single thread in the preferences.

Does anyone know where the extra CPU time is going?

They are idling in the background loading webpages or whatever else you have your pc doing, I do this on my main laptop to prevent it from over heating and still make it usable while I crunch with it. Lots of people do this to give their gpu's a full cpu core so it can crunch at full speed, they would have it set to only use 7 cpu cores instead of your 6.
4) Message boards : Number crunching : The Science Cloud - donate CPU time on the cloud (Message 136056)
Posted 28 days ago by Profile mikey
When I saw this post, I hoped it was the reverse of what it actually is. I was hoping it was a way for someone else to use my home computer to run their projects for a small fee (which would have to be less than cloud providers to not be exploited). As someone who lives in an area with very low electricity cost, this would be attractive to me and then I would feel less bad about leaving my computer on 24/7 crunching things. Of course, you would also have to have the fee be associated with families of processors, and have requirements around stolen CPU time, etc.

You could do this on your own thru PayPal or something with a user giving you their user name and password and you crunching for them for x number of days for y cost. Alot of trust between people going on though would probably prevent some people from doing it.

I can see the OP's way of doing it though for challenges and other short term things. I could boost my crunching power for 2 weeks for example.
5) Message boards : Number crunching : GTX 1650 SUPER vs GTX 1660 (Message 136011)
Posted 29 days ago by Profile mikey
Good day All,

I have received an i5-760 CPU/motherboard/RAM combo with 2 PCIEx16 slots and I wanted to add a couple of GPUs to it. Browsing local computer stores it seems that I can get a GTX 1650 SUPER or a GTX 1660 for about the same price due to the sales on the 1660s. Memory bandwidth is the same for both cards, and core configuration is relatively similar (1280 vs 1408), however the memory speed in the 1650 SUPER is 50% higher than memory speed in the 1660.

In your observations, would I be better off with the higher memory speed of the 1650 SUPER, or the higher core count of the 1660? Primary use would be GFN.

Thank you,

Higher memory speed is better as the data gets processed faster so you have a better chance of being first. If you do other Boinc Projects too then the higher core count would be better.
6) Message boards : Problems and Help : not allowing tasks (Message 135697)
Posted 41 days ago by Profile mikey
Sorry it does not even sks for that.

I don't think mine asks either it just comes up with a blank line that I type the password into, press enter and off it goes. I use Linux Mint though.
7) Message boards : Problems and Help : No work available to process & Processing suspended (Message 135624)
Posted 43 days ago by Profile mikey
Thank you. I had to go into advanced view and go into computing preferences to disable these restrictive options, and that fixed the problem. GPU computing is now enabled.

Windows Task Manager's Performance tab doesn't show GPU usage (shows 2%), but GPU-Z shows a 100% load.

I went from using 180 watts to 351 with my monitor off. Ouch :x.

Yes that's exactly what is supposed to happen! There are less computationly intense Projects for your gpu but that kinda defeats the purpose for most people.
8) Message boards : Number crunching : Why different sent times? (Message 135556)
Posted 44 days ago by Profile mikey

I did actually have it set up this way initially, and my main PC (#662) is still set up this way, but I found that it would cause some issues with the PrimeGrid CPU tasks that I was also running. My CPU tasks would all complete, and then I would download another batch of however many cores my CPU had (ie. 4 or 6 more tasks), even though they were supposed to run multi-threaded. That sent my percentage of firsts down pretty low... I had not though of it until you mentioned this and read dthonon's post above, but with the new functionality to control the number of tasks and threads through the preferences page this problem would no longer exist as I can specify the number of tasks per PC from PrimeGrid.

I did this and the problem with the zero resource share on LLR units is that you want it to use multiple cpu cores and if your other project uses even one cpu core then you might not ever get an LLR unit to run until it's near it's deadline. Cache size doesn't matter here it's workunit deadline that takes priority and you could run Project X on 1, 2 or even 3 cpu cores and still have one, or more, cpu cores open if they send you limited numbers of tasks on a quad core cpu. It depends on your LLR settings of course but LLR units are most efficient when using multiple cpu cores, if you set it to only use 1 cpu core it will work just fine as a zero resource share backup project.
9) Message boards : Number crunching : Why different sent times? (Message 135466)
Posted 47 days ago by Profile mikey
The best, but totally unsatisfying answer, is that sending out the matching tasks simply is not a priority written into the BOINC scheduler.

(Full disclosure: I'm not an expert on the internal workings of how this part of the BOINC software operates, but I've never seen anything in the software that attempts to send out tasks in the same workunit sequentially.)

Given the 'race' for being first here at Prime Grid I wonder if someone could ask for that to be coded into an upcoming version of Boinc with the ability to enable or disable it as the Admins choose?

I've had the same question myself, being on both sides of the situation in the past and it just seems unfair as the OP says when one person gets the workunit, 39 hours in this case, before the 2nd person. A few minutes or even afew hours could be overcome by hardware but not 39 hours unless someone is using a multi day work cache for example.

It would also help Projects that do double precision, but no race for first, as everyone gets the workunit at about the same time.

I think it would involve just keeping track of how many workunits are sent out of a given kind and to whom and then send out the 2nd copy of workunit ????,,,??? to the very next different person asking for a workunit of that kind.
10) Message boards : Number crunching : Google Colab and its free GPUs (Message 134849)
Posted 74 days ago by Profile mikey
I noticed the terms of service for colab say no cryptocurrency mining.

Please note that using Colaboratory for cryptocurrency mining is disallowed entirely, and may result in being banned from using Colab altogether.

I'm not sure they'll like people using BOINC either.

Since Prime Grid distanced itself from GridCoin I'd hope helping to find Prime Numbers would be a good thing.

Next 10 posts
[Return to PrimeGrid main page]
Copyright © 2005 - 2020 Rytis Slatkevičius (contact) and PrimeGrid community. Server load 3.06, 2.33, 2.53
Generated 26 Jan 2020 | 2:00:45 UTC