You don't just seem to understand the fact that since you only have 2 cores, if you do more than one project then there will be times that the system will be processing one project & one project only in order to meet your Resource share settings between all the projects that you have chosen to do work for.
That is simply a basic fact- you have 2 cores, so only 2 Tasks can be processed at any given time. The more projects you have, then the longer it takes to do enough work for each project so that your Resource share settings can be met.

i participated in S@H because I admired its goals of finding life on other planets. I crunched Asteroids because it made significant contributions to solar system science. I did citizen water quality monitoring because I was concerned about the Chesapeake Bay ecosystem. I do Globe at Night because it raises awareness of the problem of light pollution. I do Milky Way because it improves understanding of our galaxy.

I do WCG and Rosetta because they both do important research in finding cures and treatments for devastating diseases

OK, maybe I'm too dense and can't quite fathom the workings of BOINC. When I put this computer together from a bare bones kit with its dual-core AMD processor, I naively believed it would be substantially faster than my previous computers.
And it is.

just not advanced enough to do BOINC as well as other here.

Maybe when this computer fails, I'll get a CPU with more cores, a graphics processor and more memory. Of course, by that time, my new system will be outdated and inadequate compared to others.

Anyway, the original message asked for our "thoughts on granting credit for long-running tasks." That's what I had, long-running tasks and so attempted to respond to that request.

My primary complaint was about the three-day deadline, which you said should be acceptable. For me, it isn't.

Doesn't mean i'll stop doing Rosetta. I was trying to respond to the request for thoughts.


S. Gaber

Yes. Seriously. Did you read the history of my participation in citizen science projects that I engaged in without expectation of reward or recognition?
Yes i did. And in the very same paragraph you them complain about your perceived lack of recognition here at Rosetta.
And you still don't respond to the question i keep asking you every time you make the statement that recognition isn't important, then complain about the lack of it- if we don't use Credit for recognition of work done, then what do we use?

I was making a comparison between Rosetta and other projects I participated in.

It would be better if the credit system were on a level playing field.

S. Gaber

OK, maybe I'm dense and can't quite fathom the workings of BOINC.

Think of an old single core computer, it can do one thing at a time quickly, or a a few things at a time really, really, really slowly if they all need a lot of CPU time to do their work.
So when it comes to doing computational work, you can only do 1 thing at a time.

If you have only 1 Project, then it can spend all it's time doing that project. If you have 2 projects, it needs to spend it's time divided between both projects, 3 projects- time split 3 ways, 4 projects- time split 4 ways etc.
One project's Task may take 5 days. Another projects Task may take 5 minutes. Another project's may take a month. Even in a single project, not all Tasks may take the same amount of time.
So because you only have 1 core you have to run 1 Task for a project, then another Task for another project, then another Task for another Project etc till you get back to the first project again. You can only run run Task at a time, so if you have 5 projects, as you need to process 10 Tasks minimum for each Project -that Validates- before BOINC gets an accurate idea of processing times, not to mention how much time the system is on & when it is on how much time it is able to process BOINC work, it takes a very long time for BOINC to settle down & figure out how much work to download - and which Project to get it from- at any given time in order to meet your Resource share settings.

And there may be two projects that both have Tasks that run for 5 minutes. One project has excellent programmers & their code is extremely well optimised and efficient. The other project has lazy programmers that just used the default compilation flags & their code is extremely inefficient.
Although both projects Tasks run for the same, the first project's much more efficient application actually does 50 times more work in that 5 minutes than the lazy programmer's application.

So using the time spent doing work for each project is an extremely poor way of allocating computing time to multiple projects. Better to use the work actually done by each project's application(s) to determine what Project to process at any given time.
And the work actually done is what Credit (the Cobblestone) is based on.


And once again you haven't answered my question.
What would you use?

Anyway, the original message asked for our "thoughts on granting credit for long-running tasks." That's what I had, long-running tasks and so attempted to respond to that request.

You do realise this thread was started early last year?
Those Tasks came & went over 12 months ago.

---

Grant
Darwin NT

It would be better if the credit system were on a level playing field.

As has been pointed out by many people (including myself) endlessly for over 20 years...

---

Grant
Darwin NT

My primary complaint was about the three-day deadline, which you said should be acceptable. For me, it isn't.

Why?
What is wrong with a 3 day deadline?

If something is important you want a quick response. If it's not, then whenever is good enough. As for 3 days to do 8 hours of work being unreasonably short- Why do you feel that way?

---

Grant
Darwin NT

I was making a comparison between Rosetta and other projects I participated in.

It would be better if the credit system were on a level playing field.

S. Gaber

It would be massively better and in an ideal world we would have a way of enforcing it but in the real world where Boinc is open source software we cannot.

"And once again you haven't answered my question.
What would you use?"

Anyway, the original message asked for our "thoughts on granting credit for long-running tasks." That's what I had, long-running tasks and so attempted to respond to that request.

"You do realise this thread was started early last year?
Those Tasks came & went over 12 months ago."[quote]

Well, it was still out there and I still had the problem, so I thought it was an open question.

My computer can run two projects at a time.
Although you think I'm stupid, it seems to me that a 5-day deadline would be reasonable.

As for which project I would chose to concentrate on, I dunno. Rosetta and WCG are doing the most immediate good for society. But I wouldn't want to quit my other projects.

Guess I'll either have to plod along the way I'm doing or get a faster, spiffier machine. Can't afford that right now, so I'll just have to stand pat and stop complaining.

My computer can run two projects at a time.

And you are attached to 5, 4 of which are active.
So the fact is there will be times that your system can work on 2 different projects at a time, or just one project at a time. Even if all the deadlines were the same & their runtimes were similar, that would still occur on occasions.
That is just the basic fact of running more than one project on a system with very few cores/threads.

Although you think I'm stupid, it seems to me that a 5-day deadline would be reasonable.

And i keep asking- Why? Why is 5 days reasonable if 3 days is unreasonable?
If it were a 1 day deadline i would agree it's unreasonable to do only 8 hours worth of work. Even 2 days. But 3 days to do only 8 hours of work? 8 hours is just over 11% of 3 days. ie bugger all. It's 1/3 of of the time available if the system is only running for 8 hours a day. Significant, but still less than half.
So why do you consider it a problem?

The models run here have the potential to save or at least improve the lives of billions of people. Why wouldn't want that work back as soon as is possible?
Why do you consider that to be unreasonable?

Guess I'll either have to plod along the way I'm doing or get a faster, spiffier machine. Can't afford that right now

Are you in the US?
If you can save $10 a week, in 12 months you could afford something along these lines.
i7, 32GB RAM, 1TB SSD $450.
Refurbished system.

---

Grant
Darwin NT

And i keep asking- Why? Why is 5 days reasonable if 3 days is unreasonable?
If it were a 1 day deadline i would agree it's unreasonable to do only 8 hours worth of work. Even 2 days. But 3 days to do only 8 hours of work? 8 hours is just over 11% of 3 days. ie bugger all. It's 1/3 of of the time available if the system is only running for 8 hours a day. Significant, but still less than half.
So why do you consider it a problem?


My projects all have different deadlines. WGC has a 10-day deadline. I think Milky Way has a 7-day deadline. It's been so long since I got any work from Asteroids that I don't remember their deadline. So the completion times can be staggered. If Rosetta had a 5-day deadline, I think it would fit in well with than scheme, and allow all my projects to finish on time..

The models run here have the potential to save or at least improve the lives of billions of people. Why wouldn't want that work back as soon as is possible?
Why do you consider that to be unreasonable?

Yes, both Rosetta and WCG have that potential and already have made major medical advances, and cures. That's why I run them.

I am in Florida on the Gulf Coast at the top of Tampa Bay.

Checked out the link to that refurbished system you sent. That's not a bad price for what looks to be a well-equipped machine.(Although that's what thought about my present computer when I bought it in 2012. It's been running 24/7/365 ever since.)

That system is a possibility. But I doubt I will obtain spousal permission to buy another one for some time to come. I did the beg-forgiveness-rather-than-seek-permission thing when I bought two guitars and one telescope mount. I think I have worn out that tactic.

Although I could probably make this one fail. That would give me a plausible excuse.

S. Gaber

My projects all have different deadlines. WGC has a 10-day deadline. I think Milky Way has a 7-day deadline. It's been so long since I got any work from Asteroids that I don't remember their deadline. So the completion times can be staggered. If Rosetta had a 5-day deadline, I think it would fit in well with than scheme, and allow all my projects to finish on time..

That isn't a reason for objecting to a 3 day deadline.
Once BOINC has sorted out the time it has to process work, and how long it takes to do the various Tasks from the various projects then one project having a short deadline doesn't stop other projects from meeting theirs. It may stop a Task that is presently being processed for a while, while the shorter deadline Task is completed, but it won't stop them from being completed in time to meet their deadline (unless you try to micro manage things).

---

Grant
Darwin NT

That isn't a reason for objecting to a 3 day deadline.

OK. Fine. I 'm not sure I agree, but I defer to your superior knowledge of the subject.

I have a technical hardware question which may also be in your area of expertise.

This computer was a bare-bones kit I assembled with components purchased from Tiger Direct. It is one of three working computers we have in the house. This one is primarily dedicated to BOINC and is very stable. It has a Gigabyte GA-F2A88X-UP4 motherboard with 8 GB of memory, a 1TB hard drive and an AMD 2-core CPU which runs at 3.9 MHz. I added a 75- Watt power supply, two bigger fans and drilled holes in the case for better air circulation. When I got it, I naively believed it would be a much better computer for BOINC, since the previous one was 7 years old.

So I now understand that this belief was erroneous, since the 2-core CPU is not able to run more than two BOINC projects at a time.

My question is, can I upgrade this computer by installing a more up-to-date CPU with more cores and another 8 GB of memory? Will such a CPU be compatible with that motherboard?

Doing that would allow me to justify the expense as a necessary upgrade, thus avoiding the domestic unrest that would ensue if I bought a fourth computer in addition to the other three. Or should I just buy a refurbished computer like the $400 one in that link you sent and suffer the consequences? I'd have to make one of the three computers fail.

My CEO/CFO does not understand that some people have many computers crunching BOINC at the same time. She can understand one BOINC computer.

Thanks for any insight you can provide in this dilemma.

S. Gaber

https://www.gigabyte.com/Ajax/SupportFunction/Getcpulist?Type=Product&Value=4709



You can but you would have to check and possibly update the BIOS on your motherboard and the max you could get would be a 4-core APU.

I personally wouldn't upgrade unless that APU was quite cheap. The most recent APU on that list is from 2015. And it's not an AMD Zen-based APU.

---

My question is, can I upgrade this computer by installing a more up-to-date CPU with more cores and another 8 GB of memory?

Even if you could get one for $40 or less i wouldn't bother, because even then, it'd be money better put towards a new/refurbished system.

The most powerful CPU supported by that motherboard appears to be the A10-7890K (4.1GHz base 4.3GHz boost) or an Athlon x4 880K (4GHz base 4.2 GHz boost) which have 4 cores- only 2 more than your present CPU, And pretty much the same clock speeds.
Any mid range current Ryzen 5 or Core i5 system would have 6 times the performance, and use less power. As well as having more cores & threads.

eg AMD A10-7890K vs AMD Ryzen 5 5600X

---

Grant
Darwin NT

I think I have found a solution, or at least a work-around, for my problem.

In the Options, Computing Preferences tabs, I specified to store 0.02 days of work and an additional 0.02 days of work, and to switch between tasks every 60 minutes.

The result is that there are always two tasks running and sometimes three. One will be a Rosetta task, one will be a WCG task and there may also be a Milky Way task. Two of these tasks will be running, the third waits to run. Sometimes there are two Milky Way tasks. In this scheme, usually one of the projects will be nearing completion. So far, projects have been completed before their deadlines. I am a happy cruncher.

This arrangement allows me to continue participating in the projects I prefer. It seems, to me at least, to be efficient use of computer resources. As long as this result continues, and as long as this machine holds out running 24/7/365, it will postpone the day when I can justify purchase of a new computer.

I thank Grant and Falconet for their advice and mentorship.

Steve Gaber
Oldsmar, FL

As long as this result continues, and as long as this machine holds out running 24/7/365, it will postpone the day when I can justify purchase of a new computer.

If you do what i do -put a bit aside each week- then the longer it lasts, the better the system you can afford when it does finally die.

I thank Grant and Falconet for their advice and mentorship.

Glad we were able to be of some help.

---

Grant
Darwin NT

If you do what i do -put a bit aside each week- then the longer it lasts, the better the system you can afford when it does finally die.

I have enough in my bank account to pay cash for a computer.

The obstacle is getting approval past my CEO/CFO, who sees that I already have three computers and doesn't think I need any more.

And she may have a point. In addition to the three computers, I have three guitars, two large telescopes with equatorial mounts, a mid-high end stereo system, two Nikon DSLRs, four Nikon and one Pentax film cameras plus 9 lenses for those. One can only acquire so much stuff. :>)

Steven Gaber

If you do what i do -put a bit aside each week- then the longer it lasts, the better the system you can afford when it does finally die.

I have enough in my bank account to pay cash for a computer.

If you can afford it, i'd advise waiting till around mid-next year before the present system dies for whatever reason.
Intel's next generation of CPUs actually looks like it might be not just competitive against AMDs current offerings, but possibly even better. Then very late this year or early next year AMD are meant to be releasing their next Ryzen series which is rumoured to be significantly better than their current one (which is excellent) and take the lead back from Intel again (if it proves the rumours about Intel's soon to be released offerings are correct).
And hopefully by next year the supply chain problems with be over (or at least significantly reduced) & pricing for hardware (along with availability) should be much improved. So if you can afford a new system, and are able to wait, then some time from mid-next year would be when to go for that new system.

---

Grant
Darwin NT

Intel's next generation of CPUs actually looks like it might be not just competitive against AMDs current offerings, but possibly even better.

We can hope so, but they have some sort of high speed/low speed combination of processors.
That might work for ordinary use, but how about BOINC projects? I don't feel at all comfortable about it until I see some tests.

---

Hello,
I'm using the PC where Rosetta computations are installed only few hours per week. Most of the jobs downloaded can't be finished.
Long-running tasks are out my reach. I went to the forum to find how to limit the size of my donwloded jobs :-p
Sincerrely yours,
Rémi

Hello,
I'm using the PC where Rosetta computations are installed only few hours per week. Most of the jobs downloaded can't be finished.
Long-running tasks are out my reach. I went to the forum to find how to limit the size of my donwloded jobs :-p
Sincerrely yours,
Rémi

Go to this [ur=https://boinc.bakerlab.org/rosetta/prefs.php?subset=projectl]page [/url] and choose the runtime you want for each task.
The default is 8 hours of CPU time. Choose the most adequate for your system.

---