Honestly, it's not very clear !

My first impression was "what's that?"

A website that is similar to GPUGRID, a scientific project that uses GPU's ATI / AMD.....But not in fact it is not to run simulations, but given the computing time to convert that into money !!!!

what's that, a new economic order ?

Honestly I do not want to give my GPU time for money even if this money back into this GPUGRID !

There is a donation system for this !

what I want is to run my hardware to perform real simulations are useful.

I am disappointed :-(

Dear Verlyol,

Donate@At home, though also run by us, is a quite distinct project from GPUGrid, and we've made some changes to that website to make it clearer. It's a side experiment for us, which may also have the happy benefit of providing some funding towards the expensive business of running our lab.
On a practical level it has also already provided some very useful experience in deploying OpenCL and AMD applications that will be directly useful in future GPUGrid developments.

By no means should you feel compelled to volunteer for it, if it's not of interest to you. There's a whole spectrum of projects in the distributed computing world, some scientific, some not. It's right that you should contribute your compute cycles to something that interests you.

We are immensely grateful to all of our GPUGrid volunteers without whom we would not be able to undertake our ambitious scientific programme.

MJH

Hi,

You need to do a lot of work because it looks like a scam. Scamming boinc users who don't know bitcoin, and letting bitcoiners beware of boinc.

At the moment, it's a 100% fee pool, and people shouldn't use this. It's something for me to monitor - better would simply to allow bitcoin donation.

Patrick

Donate at Home is a 2day old alpha project. The project, apps and site are very much under development!

The primary aim of D@H is to fund a researcher at GPUGRid and any funds generated will go towards GPUGrid. However D@H should be considered as a completely different project (own server, site, apps, credit...). Obviously funding a researcher depends on the success of D@H, but contributions to D@H will hopefully enable and facilitate everything that researcher does.

Using a GPU to fund a researcher via bitcoin mining should just be seen as a novel donation method. Obviously this is an experiment in itself; the system is untested for this purpose. That's why it's an alpha project.

GPUGrid (which principally does bio-molecular research using GPU's into Cancer, HIV, Neurological conditions and develops research techniques) is a different project to D@H.

Just because you crunch for one does not mean you need to crunch for the other, which is why they are separate projects. NVidia cards are great here, and the ATI cards are great there, but it's entirely the decision of each cruncher whether to contribute to either and to what extent.

I like your idea of being able to donate a bitcoin. It should be considered.

PS. The creation of D@H does not mean that an AMD GPU project will not happen here. If a GCN app can be created that allows for reasonable performance at GPUGrid I expect it to be beta tested and launched. Unfortunately non-GCN AMD cards will never perform well at GPUGrid, but at least with D@H you can make a contribution using your AMD GPU, if you want to.

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

I like your idea of being able to donate a bitcoin. It should be considered.

Being realistic Donations don't work!!!
If you add top 10 donators you get 3764. That could be around 1€/user.

With that money you don't even buy a good server. Not talking to hire a well qualified worker.

what's that, a new economic order ?

Is that supported on a currency that goes from 30 to 5 in one year??? Come on! A good currency need fiat money . The garantee of a central bank, taxes, properties or something to garantee the value, not the goodwill of many people and a electronic number.

That stuff is more close to be a Ponzi Scheme that anything.

I don't think is bad that you do mining. My only complaint is that you should explain more the light side and the dark side of bitcoins. Even if you think it's a good thing you are going to promote something to many people that have not the knoledge and preparation to know what's behind (not easy even for people expert in economics), they could go behing and take financial risks. I can think in many people that went into Stock Exchanges without knowning the real risk of their investment and they lost a lot of money. And there you even have assets to back-up your invest, what do you have in a electronic number?.

The primary aim of D@H is to fund a researcher at GPUGRid and any funds generated will go towards GPUGrid.
...
Using a GPU to fund a researcher via bitcoin mining should just be seen as a novel donation method.

While that reasoning is all well and good from a volunteer's perspective, I would be surprised if the proceeds from D@H would be treated as "donations" from an accounting/reporting/compliance perspective.

I don't know about Spain, but in the US a project such as D@H run by a non-profit university would be considered an "Auxiliary Activity" (essentially a commercial venture whose profits are earmarked for support of the non-profit organization). Auxiliary activities run by non-profits in the US are subject to a completely different set of rules & regulations; in some cases they are subject to taxation.

Maybe things are different in Spain. If not, I sure hope those in charge have done their homework on compliance.

Of course we have spoken with the University before starting it and it works similarly to the donations with paypal from the admin perspective.

gdf

Finally I found some information - in the GPUGRID cafe! :D

We seriously thought this was a scam/fraud project during first days at the team...finally there seems to have been more information and WHOIS directing to Gianni, so at least to somebody involved linking to GPUGRID. Not really brilliant on information.

But what really bugs me: you DO know how to make high paying projects! Now that ATIs can be used for generating money, they get real credits. What does that tell about the appreciation for the science work done?

Basically I think it's a smart idea, but this is looking dubious in so many ways, I really find that hard to believe, sorry.

---

Of course we have spoken with the University before starting it and it works similarly to the donations with paypal from the admin perspective.

gdf

Thanks for your reply, GDF. It was not my intention to imply any shortcoming in your project admin process, but after re-reading my post today I can see how it might have come off that way. My apologies for any offense taken.

I did want to raise the compliance issue on the off chance it wasn't on your radar, because running afoul of the "burocracia" can mean trouble (as I'm sure you know). I worked part-time in my university's internal audit department while in grad school; I saw instances where the needs of the bean-counters were overlooked by PIs, and the consequences were...severe.

Even though I believe D@H is a step in the wrong direction, GPUGRID will continue to have my full support. :)

Have you considered the impact that donate@home is having on the environment? You ask us to crunch nonsense tasks on our GPUs and waste a lot of electricity in the process. This pollutes the earths atmosphere with a lot of unnecessary CO2. All of this just to generate a few meager euros?? Have you calculated how many kWh of electricity need to be spent just to generate 1 BitCoin? And how much that electricity costs us?

I think this is a really bad idea showing reckless disregard for the earths environment. I am really disappointed and will stop running GPUGRID at least until this donate@home site has disappeared.

I agree with you that bitcoin computing is absolute waste of energy and earth ressources. But nobody must join Donate@home ;) Donate on the traditional way when you want to donate for gpugrip, and compute like you computed until today ;)

---

---------- 24/7 Crunching since 2011 -----------
DSKAG Austria: http://www.dskag.at

I am really disappointed and will stop running GPUGRID at least until this donate@home site has disappeared.

I also decided to completely stop my participation @ GPUGRID as donate @ home is active.

For your information i am highly involved in the project folding @ home as a Beta-test, and I can tell you that a project like donate @ home will not become reality one day ....... it is simply incorrect and not in the spirit of gridcomputing !!

I agree with you that bitcoin computing is absolute waste of energy and earth ressources. But nobody must join Donate@home ;) Donate on the traditional way when you want to donate for gpugrip, and compute like you computed until today ;)

The question is not whether to join donate@home or just keep on GPUGRID simulations ?

This is a choice of the GPUGRID team which is completely outside of the spirit of gridcomputing !!

I also agree regarding the impact on the environment for the creation of 1 Bitcoin !!

I think this decision is very bad for the reputation of the project.
That is why as long as donate @ home is active i stop my participation to GPUGRID !

People should try to formulate their own decisions about each project they crunch for based on information and merit, and try to respect other peoples individual choices.

I can see past the creation of BitCoins (method) to funding a researcher (goal), and understand the consequences of the alternative - not being able to have a researcher. I can also see some of the limitations, concerns, novelty and potential of the project. While I don't want to overemphasize other benefits, as it's primarily a funding project, I see the development of an OpenCL project and apps, and researchers gaining experience with ATI/AMD. I've also gained some more AMD GPU insight as have other crunchers who shared their experiences in the forums, and this is one week in.

I'm not sure what 'spirit' you are referring to and how you would rate some other projects 'spirit'. I value GPUGrid highly, and some other projects highly, but certainly not all projects. What's the difference between using electric to find the 5Mth digit of Pi, a prime number or a hash? It's down to choice for the cruncher. If you have a mathematical interest, and want to try to solve some puzzle it's your choice. If someone else wants to use their home GPU to contribute to funding a Bio-molecular research team that develops and expands research methodologies using GPUs, it's their choice. GPU crunchers/enthusiasts tend to push the frontiers more than CPU crunchers because it's in their nature. Killing innovation isn't.

I'm keeping a very open mind about what is a one week old experimental project. How much funding can be generated is still unknown, but it's definitely limited. You can't work out the relative electrical cost against creation of BitCoins/funds without actually trying it, and trying to develop apps. That's partially why it's experimental and Alpha. So how much merit this method has for generating funds is still unquantified. Instantly dismissing it is reactionary and prejudgemental, which I can understand as I was the first. However I didn't stamp my feet and throw the toys out of the pram.

If people are interested enough and want to engage in sensible discussion, ask questions and get informed replies about D@H, go to D@H and ask there.

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

I see this issue as follows:

1. There are several ATi owners (and lesser nVidia card owners like me was for more than a year), who'd like to contribute to GPUGrid, but they can't.
2. Some of these potential contributors complaining for years for an ATi client, they get tired of waiting in vain, therefore they get frustrated.
3. The people at GPUGrid get tired of making an ATi client for their project, get frustrated from the complaints, but luckily they came across donate@home, and think "Wow! finally we can offer a way to our potential contributors to become actual contributors!" therefore they set up this donate@home website for them.
4. Some of the potential and some of the actual contributors gets angry, because from their point of view donate@home is not green enough, it's a scam, it's dubious, it's against the idea of grid computing, while nobody forcing them to join.

I would like to encourage these potential contributors that instead of waiting in vain and wasting the researchers' time with complaining, they should take some actions, like selling their cards, and donating the money they received for their cards directly to GPUGrid (it's very green, it's not a scam, it's straightforward, it's supporting the idea of grid computing); or adding a little amount to it, and buying a CC2.0 nVidia card to become an actual contributor. I did the latter. It works.

You can't work out the relative electrical cost against creation of BitCoins/funds without actually trying it, and trying to develop apps.

With all due respect, anyone who is familiar with bitcoin - or has done a modicum of research on the matter - knows this statement to be false.

I would like to encourage these potential contributors that instead of waiting in vain and wasting the researchers' time with complaining, they should take some actions, like selling their cards, and donating the money they received for their cards directly to GPUGrid (it's very green, it's not a scam, it's straightforward, it's supporting the idea of grid computing); or adding a little amount to it, and buying a CC2.0 nVidia card to become an actual contributor. I did the latter. It works.

Retvari, I sure hope you are paying less than €0.0578/kWh for your electricity. If not, you will quite likely be paying more for the electricity to run your CC2.0 nVidia card than the value of the BTC it can reasonably be expected to produce. I make this statement assuming today's BTC price of €3.48, current mining difficulty of 1379647, 50 coins generated per block, and mining with a GTX580 drawing 244W.

To put it another way, if you are in the Czech Republic paying €0.1541/kWh (the rate listed at http://www.energy.eu), bitcoin prices would have to increase from today's €3.48 to around €8.83 just for your expected BTC contribution to match your electricity costs . If you bought anything below a GTX580 or GTX570, the numbers work against you even more harshly.

If it's not too late, I urge you to return your new nVidia card and get a nice AMD card - a 5970 if you can find one. That, or wait for a Kepler model if they are shown to be decent miners.
_____________________________________

The fact that a GPUGRID volunteer as sophisticated as Retvari Zoltan has somehow concluded that it makes sense to purchase a brand new nVidia card for use on D@H should be considered a red flag for D@H admins. It makes absolutely no financial sense for the overwhelming majority of people paying their own electric bill to mine bitcoins with ANY currently available nVidia card.

If anyone wants a second opinion, I invite them to post on one of the many bitcoin forums and ask the folks there what they think about using nVidia cards for bitcoin mining.

Nvidia cards which are quite fast here lacks some shuffling instructions which AMD cards have. They take 3 instructions to the same. The net result is that for the calculations done at donateathome a 6970 is twice as fast as a GTX580. This will probably change with kepler, with nvidia being at the same level.
We posted a set of technical links for these sort of comparisons (donateathome.org/project.php).
gdf

To SMTB:

You misunderstood me, maybe I was not clear.
My suggestion was for the potential volunteers that they should change their ATi (AMD) card, if they do not like the idea of donate@home, because at the moment the only way to directly contribute to GPUGrid is by crunching for the project with top nVidia cards (or by donating it).
I hope it'll change for the better, namely the top AMD card owners could crunch for the project too, but it still will require the latest GCN architecture models i.e. many ATi (AMD) owners have to sell their current cards and get a new one.
But, I think - against any complaints - the current situation is better than before donate@home, and donate@home was made with the best intentions of the GPUGrid staff.

To SMTB:

You misunderstood me, maybe I was not clear.

My apologies, Retvari. I admit to being a little worked up about nVidia on D@A...I guess I read something into your post that wasn't there. :)

My motivation is not to try and torpedo the D@H project, but to draw attention to the folly of using nVidia to support it. I plan on continuing to voice my objections until the leadership here makes it crystal clear to everyone that nVidia is not currently recommended - or if they drop nVidia support entirely.

(or until my posting privileges are revoked lol)

On several occasions I mentioned that the performance of NVidia is ~1/3rd that of ATI for D@H, so did GDF, and others. I also suggested that people use their high end NVidia's here. On the D@H site there is a new 'Project' page with a link to performance of cards. It basically shows all NVidia cards as being incapable of gaining enough bitcoins to pay for the card+running cost.
The tables on this page suggest that several AMD cards are capable of paying for themselves (purchase cost + running cost), eventually, if you are mining for yourself. So for example a GTX5850 at todays supposed price ($365.68) would pay for itself in 411days. Obviously if you buy it cheaper, say second hand, tune it well, and have cheaper electric then it would pay for itself much faster. But the reverse is also true. As for the accuracy of the table - it's outdated and there are errors:
Radeon GPU Approx M-hash/sec Approx ฿/day Approx $/day Watts $/day less Power GPU Price Break-even (days)
6970 323 ฿0.25 $1.39 220 $0.60 $89.99 149.98
If that really was the case then we would all have HD 6970's, but that's a PSU price! Prices are from Amazon! Good idea but doesn't work for older cards.
My 5850 cost ~£100 and OC's to 775 fine (dual fan). Reducing the GDDR5 helped stability. Using some data on that table and some of my own it would be more like 150 to 200days for it to break even, in theory. Of course anything generated is nonchalantly being donated, but I fully take your efficiency point. As mining is hit and miss, pooling is the only real option for casual miners. I've already made the obvious suggestion regarding this - make a table, show the results. Better still make an online calculator so people can enter the GPU cost, electric cost and just be told how efficient or inefficient a card actually is. Would be useful to be able to enter actual Watts consumed too! These tend to be around 70% of the TDP for most GPU projects, but actual values might vary.

Just to add that while people should generally keep their high end NVidia's here, it doesn't mean that you can't run the odd task now and again at D@H, to help development. You can and it's your choice.

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help