This is nice, we do have a dialogue started. What do I think about BU? I think it's fine except that the credits awarded are crazy compared to the actual work being done. What do you think about BU?

I think that way too. But it's the simplification of the problem. The "award" for mining is BitCoin itself, it shouldn't be awarded again with BOINC credits. When a miner have some BTC, they can do anything with that BTC they want: donate it for a project, or exchange it to real world currency, and spend it, or donate the real world currency. There shouldn't be BOINC credits involved in donating money (BTC).

The problem with that idea is that each user does NOT donate directly to the projects, it's more like a crowd funding thing. Since each person does not get any recognition for their contribution to the 'campaign' they donate their crunching to, they get credits instead. BU IS a Boinc project, so Boinc Cobblestones are currently the only way to recognize each persons crunching contributions.

As most of us have said before the Cobblestones earned at GpuGrid do NOT comparable to the ones earned at Cosmology, PrimeGrid, Seti, BU or ANY other project, they are best used for comparing results within GpuGrid ONLY. Whether that be user to user, or each users own performance over time. If we could go back in time and get the stats sites to STOP combining them into one big overall number, then this wouldn't be an issue, unfortunately that door has been opened and it may never close again.

I am crunching Cosmology cpu units right now and am getting a flat 480 credits for each unit, with some units taking over 15 hours to complete!! I just checked one of Retvari's units and he is getting 147,600 credits for roughly 7.5 hours of work, HOW is that comparable? The answer is...IT ISN'T!! It NEVER should be, nor should it have ever been. BUT we are all human and want to see how we compare to others. The problem is they just aren't comparable, like comparing me to a football/soccer player, I have bad feet and am over 60 years old, we are NOT comparable!!

It's a bit like comparing my gpu's to both Beyond's and Retvari's gpu's...they are all gpu's but beyond that they really aren't comparable. Beyond has almost all Nvidia 750's, I have some Nvidia 760's, while Retvari has some 980's, 780's, 680's and even a 750! They are NOT comparable to each other beyond the fact that they all crunch!

Beyond has 1,218,146,142 total credits here, Retvari has 4,087,611,712 total credits here, while I have 100,346,915 total credits here. THOSE are comparable!!

Doulble post. My bad.

There shouldn't be BOINC credits involved in donating money (BTC).

There was BOINC credits given when donate was run here. What's the difference?

Cpu crunchers we can initiate to BOINCstats remove the gpu projects!
For us it is discriminatory!
Cpu tasks crunchers job is difficult. It's such a doctorate from the University.
1000 000 rac day in cpu projects requires 15K+Eur + time + one person take care 24/7.
GPU crunching is as elementary school. 1000 000 Day RAC knows every man with two graphics cards. just € 1,000.

Happy Christmas :-) )))))

How about if the stats sites just divide the bitcoin credits by a reasonable amount? I don't know what the constant might be but for instance:
bu crdits/20 = BOINC credits on the stats sites...

How about if the stats sites just divide the bitcoin credits by a reasonable amount? I don't know what the constant might be but for instance:
bu crdits/20 = BOINC credits on the stats sites...

Why should anyone really care about points? If you're doing any kind of DC for points than you have the wrong motives IMHO. If you're going to reduce the BU points in BOINC to 20:1 then maybe you can get them to give WCG full credits instead of 1 BOINC point for every 7 WCG points.

There shouldn't be BOINC credits involved in donating money (BTC).

There was BOINC credits given when donate was run here. What's the difference?

The difference is the impact of these projects.
Donate@home wasn't a big success, as it was inefficient, but it was just as immoral as BU is.
I'm very sorry that I didn't realized this back then. My only excuse for that is the insignificance of that project.
BU haven't had a big impact, until the ASICs came.
When I fully understood what is going on actually, it was too late to take it back.

This thread has 73342 views.

Cpu crunchers we can initiate to BOINCstats remove the gpu projects!
For us it is discriminatory!
Cpu tasks crunchers job is difficult. It's such a doctorate from the University.
1000 000 rac day in cpu projects requires 15K+Eur + time + one person take care 24/7.
GPU crunching is as elementary school. 1000 000 Day RAC knows every man with two graphics cards. just € 1,000.

You are right (partly), but this thread is not about this problem.

Happy Christmas :-) )))))

Same to you!

Why should anyone really care about points? If you're doing any kind of DC for points than you have the wrong motives IMHO.

You are right, but there are such people.
When I realized that the BU project (and donate at home also) actually selling BOINC credits for BTC, I was shocked by the immorality of it.
This shock haven't passed.

Why should anyone really care about points? If you're doing any kind of DC for points than you have the wrong motives IMHO.

You are right, but there are such people.
When I realized that the BU project (and donate at home also) actually selling BOINC credits for BTC, I was shocked by the immorality of it.
This shock haven't passed.

I respect your point of view but I disagree. BOINC points have no intrinsic value and therefore can't be sold IMHO. If I choose to run my miners at my own expense to raise donations for projects I deem worthy then that's my choice. The "points" are just an acknowledgement of my contribution and nothing else just like the badges that some projects award. If any of it stirs up a little friendly competition and raises money how is that immoral?

Why should anyone really care about points? If you're doing any kind of DC for points than you have the wrong motives IMHO.

You are right, but there are such people.
When I realized that the BU project (and donate at home also) actually selling BOINC credits for BTC, I was shocked by the immorality of it.
This shock haven't passed.

So by that logic we are buying Scientific credibility by crunching too? They are not "selling" Boinc credits for bitcoins, they are awarding credits for crunching, just like every other project does. Here at GpuGrid we crunch and the project does what they do with the results and we users get credits, THAT'S the Boinc way. You provide your resources to a project or your choice and you get credits, that are in fact worthless, in return. Retvari you personally have SEVEN pc's here, all of them Intel i-7 cpu's, and you also have some of the latest and greatest gpu's in them all crunching for credits here at GpuGrid. Are you "selling" yourself to GpuGrid for those little icon thingys saying your work was used in a "Scientific Publication"...I don't THINK that's the way you see it.

Then WHY would you think that crunching for BU is "selling" yourself for bitcoins? When someone crunches for MalariaControl are they "selling" themselves for the lives of those saved from Malaria? NOT in my opinion they aren't! We all contribute what we can, how we can, and in return we get "cobblestones" that are in fact just numbers on a page. One Trillion "cobblestones" and five dollars can get you a cup of coffee at Starbucks, BUT without the five bucks you CANNOT get any coffee at all! You can ALSO take ONE CREDIT and five bucks and get that SAME cup of coffee though!!

The point is that each of us contributes to the Boinc project of our choice, some of us more than others, but we don't do it so WE get anything in return, we do it so OTHERS can advance that project further than they could without us! In our own way we ARE helping advance things we often have no clue about, nor have the time to learn HOW they do it! At bu we are doing the same thing, except they are generating money instead of 'mapping the cosmos', or 'solving some math equation' or 'folding proteins' or whatever other projects we chose to contribute too. At BU the Admins then take that money and DONATE it to different "campaigns" that have been asked, or have asked, for funding help. Several Boinc projects and even a stats site or two wouldn't be here today without bu's help. Seti itself is now accepting help from bu, it turns out that no matter how many times they ask their crunchers just won't contribute enough to keep them running, much like other projects that have funding problems.

GpuGrid could even contact bu and ask for help for this or that purpose, the Admins there will be happy to talk about the possibilities. BUT bu is also picky in that they don't just accept anyone off the street that needs money either. They vet every request, with some being accepted and some being rejected. The 'buying more and better pc's for mikey' request would most likely be rejected as it doesn't benefit the 'greater good', but keeping the Boinc Projects Seti and MilkyWay around was accepted as they DO!

I feel the main points are being missed. I'll summarize what I believe the main points are:

"New Resource Utility - Bitcoin Mining"
- BOINC used to do: Donate spare CPU/GPU cycles ==> Enhance science/publications ==> User rewarded with credits
- BOINC can now do: Donate spare CPU/GPU/ASIC cycles ==> Enhance campaigns with Bitcoin ==> User rewarded with credits
... In both cases, the user is "selling out" their resources.
... But in the Bitcoin Mining case, you literally can make a proportional comparison between amount of resources allocated, and amount of money generated.

"New Stats Resource - ASICS"
- BOINC projects used to do: Use benchmarks to determine how many FLOPS a CPU/GPU can effectively do, then use that in an equation to determine credits ==> The CPU/GPU resources can be used at any project that implements an application for them, and cross-project credits should be semi-comparable, even capable of generating competition.
- BOINC projects can now do: Determine a credit value for ASICs that actually is based upon a comparison against FLOPS, but is very very high in credit value ==> ASICs can only be used for a single purpose (are not generalized computation devices; can only be used at certain projects that use that exact same purpose), and thus cross-project credits aren't comparable. Also, their high credit values negate an ability to generate credit competition.

My conclusions:
- This issue is affecting the community, and should be addressed.
- I do not think that ASIC credits should be arbitrarily lowered to try to "fit" the GPU/CPU model of credits.
- Rather, I think we need an easy way to, when viewing stats, optionally "strip out" the "single-purpose ASIC credits".
- To do this, we need a way to segregate them; a new shiny "bucket" to put them in.
- David Anderson has created a proposal (which makes sense to me), and he has requested feedback on the BOINC Projects Mailing List.
- The links are:
http://boinc.berkeley.edu/trac/wiki/CreditGeneralized
http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_projects
- Please make sure feedback makes it to him.

Regards,
Jacob

I'm going to Hell in a Hand Basket, so sue me, I'll talk to the Devil when I get there about Increasing the Inadequate Watered Down Credits we get there now ... :P

---

STE\/E

- Rather, I think we need an easy way to, when viewing stats, optionally "strip out" the "single-purpose ASIC credits".

Bad idea IMHO. Credits are credits no matter how many or how they are generated. The GPU trumped the CPU. The GPU credits weren't stripped out. Now asics have trumped the GPU. If you "strip out" the asic credits now what happens when the next tech advance surpasses asic capabilities? Does this whole discussion start again?

- Rather, I think we need an easy way to, when viewing stats, optionally "strip out" the "single-purpose ASIC credits".

Bad idea IMHO. Credits are credits no matter how many or how they are generated. The GPU trumped the CPU. The GPU credits weren't stripped out. Now asics have trumped the GPU. If you "strip out" the asic credits now what happens when the next tech advance surpasses asic capabilities? Does this whole discussion start again?

Please do not confuse the reason for the request to re-categorize the credits.

The reason is NOT: Because ASICs work harder/faster/better.
The reason IS: Because ASICs computations cannot be generalized in a way to be applicable to other projects, like CPUs/GPUs are.

There is a huge difference in those reasons, and you keep erroneously mixing them up.
You do know what the acronym ASIC means, right?

The reason IS: Because ASICs computations cannot be generalized in a way to be applicable to other projects, like CPUs/GPUs are.

So accoding to you and for that reason alone the ASIC ponts should be stripped out of BOINC awards? Pretty narrow minded IMHO

There is a huge difference in those reasons, and you keep erroneously mixing them up.

Your point of view on the reasons does not extrapolate to an error on my point of view.

You do know what the acronym ASIC means, right?

You do know what the acronym EBCAK means, right?

ASIC credits should not be "stripped out". They should be "categorized" such that a user visiting a stats site can choose to not include them, since they came from a resource that is not generalized.

I was seriously asking if you knew what ASIC meant. Your insulting reply was uncalled for.

ASIC credits should not be "stripped out". They should be "categorized" since they came from a resource that is not generalized.

Again, your opinion.

I was seriously asking if you knew what ASIC meant. Your insulting reply was uncalled for.

IMO your question was condescending and got the reply it deserved. I'm done. Have a Merry Christmas and a Happy New Year.

Sorry for offending you. Some people don't know that an ASIC (Application-specific integrated circuit), by definition, is not capable of being generalized for use by all projects.

And I am well aware that my proposals are "opinions". They are, what I believe to be the best approach to alleviate the concerns people have. In a sense, everybody wins -- If you visit a stats site and want to see ASIC included, that's your choice! If you visit a stats site and want to exclude credits from project-specific resources like ASIC, that's your choice!

Happy holidays to you.

Sorry for offending you. Some people don't know that an ASIC (Application-specific integrated circuit), by definition, is not capable of being generalized for use by all projects.

Happy holidays to you.

My question then becomes 'what happens when project xyz comes online, using asic's, and is all about 'the Science'? Suppose Seti, for example, figures out how to optimize their Astro-Pulse units so as asic device could be used? Would that change the equation?

In the beginning ONLY cpu's could be used in Boinc and gpu's were for 'the gamers', then someone sat down and figure out how to utilize the gpu's specific calculating power to crunch Boinc workunits. Someone WILL figure out how to utilize an asic device, maybe a whole new app, just like when the gpu's started out, and then what? Do we STILL strip out the asic stats?

To me the genie left the bottle when cpu stats and gpu stats were combined, to now say well asic's are 'different' so they must be stripped out is kinda late. What about all those cpu ONLY crunchers, aren't their stats being blown away by those that use gpu's at the same projects...YES they are! How are asic devices any different? Boinc Cobblestones are about 'numbers', nothing more and nothing less. That fact that a gpu can kinda sorta be related to the way a cpu works is immaterial, they both produce relevant data to the project. An asic device can and does too, to penalize ALL future projects because of one newcomer is short sighted and alot like closing the barn door after the barn has already burnt down to the ground around it, IMHO.