Ah, I see the problem. You don't understand what the problem is. BOINC doesn't do any CUDA computation itself, it just needs to be able to deal with applications that do. And it doesn't manage that properly.

You don't need to understand what's going on to be optimistic about something. And I don't need to argue with you. You're optimistic, I'm realistic. That annoys you, for some reason.

We'll have CUDA support in time. You just need some patience. World Community Grid have no particular wish to waste time and money being on the bleeding edge, when they can step back and reap the benefits later.

And, given that you took all my previous comments out of context and misinterpreted with them, I will not discuss this further with you. Period.

Otis11

I share your concern about the board remaining peaceful, and a place to discuss issues related to grid computing and the various projects that run on WCG. That's why, despite the numerous false informations and personal attacks sent by Didactylos in nearly all threads about GPUs for months, I just stopped posting about this subject as did most other posters, exactly as Didactylos wanted.

Someone started this thread and Didactylos went at it again. I added complementary information (Autodock is the software that runs the FAAH calculations, so it's at least a bit relevant) and got another disrespectful comment by Didactylos asserting his right to be "vociferous".

Sorry, but this guy will not rule the whole board with his vociferations, and it's about time his lies get exposed. It's not about GPU being good or bad, it's about defending your point with valid arguments and facts, rather than the kind of lies and smear he uses against other posters.

So, sorry for it, but bear with me while I'm dealing with the little bully ...

------------------------------------------

Didactylos wrote:
"Ah, I see the problem. You don't understand what the problem is. BOINC doesn't do any CUDA computation itself, it just needs to be able to deal with applications that do."

Nice try, but where did I mention or even suggest that BOINC was doing the calculations ?
Absolutely nowhere.
To the contrary, my first post in this thread was about the GPU enabled version of Autodock. As you know, Autodock (and not BOINC, maybe you confuse them) is the software that does the actual calculation for FAAH and the Dengue drug project. So, since both BOINC and Autodock are now GPU enabled, it means there is a strong possibility to use them on a grid network such as WCG to speedup those projects ...

What you did here (invent an argument your opponent did not make, and then prove it false) is a known tactical fallacy called "strawman argument" and it's use reflects poorly on the ethical and moral values of the person using it. However, it can sometimes be effective if your opponent does not know about it. No such luck here !
http://en.wikipedia.org/wiki/Straw_man

Didactylos wrote:
"And it doesn't manage that properly."

Once again, where are the proofs, where do you get that from, beside your conviction that it should not work properly ?
Is that some form of truthiness ?

Didactylos wrote:
"You don't need to understand what's going on to be optimistic about something."

Just like you obviously don't need to understand what's going on to voluntarily spread lies and disinformation based on rigid misconceptions. Or maybe you do, and that's one less excuse for what you've been doing for months.

Didactylos wrote:
"And I don't need to argue with you."

You'd better not, as you obviously can't back any of your claims with facts.

Didactylos wrote:
"You're optimistic, I'm realistic. That annoys you, for some reason."

If you were realistic in any way, you would have facts to back your outrageous claims.
I have little patience for people who voluntarily and maliciously plant baseless disinformation in boards as an attempt to mislead others or prevent discussions on a particular subject they dislike by playing little bully. In the future, if you keep making those false allegations, be prepared for some backdraft.

Didactylos wrote:
"And, given that you took all my previous comments out of context and misinterpreted with them, I will not discuss this further with you. Period."

The good old, but tired "out of context citation" defence ! Who do you think you're fooling ?

You have no evidence to support any of the claims you made. It's all a lot of hot air and "vociferations" without substance. I feel your pain, sometimes, it's better to simply refuse answering questions.

To finish on a more positive note, Didactylos also wrote:
"We'll have CUDA support in time. You just need some patience. World Community Grid have no particular wish to waste time and money being on the bleeding edge, when they can step back and reap the benefits later."

It sure will happen, the question is when. If it takes too long, WCG, despite it's goal of furthering public domain research will fall behind, as people join other projects that make a better use of their machine. Since people who have GPUs are among those who have the most recent / fastest machines, losing them (or not gaining new ones) is an important loss for WCG. The question won't go away by preventing people from talking about it in the boards. If there is still no GPU enabled projects in a few months many people will feel like it's as inadequate as a project that would still be unable to use dual/quad core (anticipating your answer, I know WCG handles that, thank you).

Enabling at least one WCG project to use GPUs is hardly a waste of time and money.

That's why many people in these boards keep posting new threads about GPU use. We all would like to see at least some update, some feedback showing WCG is not asleep at the switch regarding this important development.

Now that there is a GPU enabled version of Autodock, FAAH being the longest running project on WCG with still a lot more long phases to run, it seems to be a perfect candidate for such a speedup.

Some people seem not to know , how many hours you have been working here Didactylos

Just reading thru here and I can understand the passions of our Canadian friend and also the point that Didactylos is trying to make.
It seems to me that yes the GPU will do"much" more work or faster per se but there are still some bugs in it.
I see this daily at the Xs forum in the folding at home section.
The guys there keep running into "problem" work units but when the work units are "correct" the GPU's show their huge speed advantage over the cpu's.
I was lucky to get a i7-965 setup before release and tested it here:
http://www.xtremesystems.org/forums/showthread.php?t=206572
Even though I work on WCG I did tests on Folding at home also.
What I saw was that a $50.00 video card, a EVGA 9600GSO that I bought as a refurb from EVGA on eBay,would put out as much work in a day as the Nehalem cpu would at 3733mhz.
There are downsides to using video cards for this work.
From what I've seen they don't hold up long term as a cpu will.
Now some will dispute that but I can only comment based on what I've seen myself.A great many cards have died while being used to run GPU work units on Folding at home.
I have no bias in this discussion. I only want what is best long term for the project and to get the work done as quickly as possible as lives are at stake here and some seem to forget that point.
I think, not know, that what is missed here is the mindset of the developers of the project.
Setting up projects costs money and the budget may not allow for implementing a new GPU based program while they see that it still has bugs. They may well be thinking from the perspective of money expended vs potential gain and have decided to sit back and wait until this new technology is perfected.
Again, that is what I feel not what I know to be fact.
My feeling is that we the "worker bees" on the project recognise the position the developers are in and support them until such time as they decide that the GPU issues are solved and then implement them.

It is a matter of patience on our part my friends and also understanding that "Rome wasn't built in a day"
GPU usage will come to WCG when it's time is right and until then we will do our best with what is available and right now that is cpu's.
Step back for a minute and think of the advances in cpu's over the last few years.
In 2005 my top machine was a dual socket,single core, Netburst based Xeon machine that did app 7500 points a day.
Today my top machine( per core)is a Nehalem quad that does 26,000 a day. That's one hell of an increase IMO.
If we look at the glass as 1/2 full vs 1/2 empty we will have a much more positive approach with this work we have chosen to do.

A last point but an important one I think.
When we come here to discuss issues it should be the same as two friends talking in one's home.
Courtesy and respect for the others point of view should be the priority in the discussion.
Ego's and the desire to"always be right" should be tossed in the garbage as they only serve to separate us when our desire should be to bring us together.
One also has to remember when discussing with someone whose position is different than our own that this other person is here donating their time and money from motives that are just as good as our own.
Yes, we have differences, but what we have in common is oh so much more important:
The desire to make a healthier world for our children.
Thanks for reading.

Three years at the Xs forum reading the posts and seeing many vid cards die from the heavy load when used in this fashion.
Feel free to read through the folding at home section there and you'll see what I mean.