• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Update on GPU3

M

Mr. Pedantic

[H]ard|Gawd
Joined
Sep 19, 2009
Messages
1,707
Here is an update and some more details on the third generation GPU core. As I stated before, this core is built on OpenMM, which brings the science further along and allows us to make easier updates. OpenMM supports OpenCL in beta form in its 1.0 release (scheduled for late January 2010), but it is important to stress that the OpenCL support is very early and so we will not be relying on OpenCL in the first releases of the GPU3 core.

This means that the core will only roll out for NVIDIA first. ATI has depreciated Brook, but does not have a fully-functioning OpenCL implementation, so we are stuck in between support on the ATI side. Once the OpenCL implementation matures (on both ATI and NVIDIA), we will be able to finish and optimize our OpenCL code (we can't reliably optimize code until the implementations are more finalized), and then the OpenCL portions will go into QA.

So, if you are interested in OpenCL (especially in terms of ATI support), this will not be in the initial release, but once the environment has matured, we will push forward. By the way, once OpenCL support for multi-core boxes has matured, we will also see about porting the GPU3 code as a new SMP2 style core as well (i.e. threads-based SMP support). If the performance is strong, it's appealing to think that we can go back to having a single dominant code base for much of FAH's calculations.
Click to expand...

Source

Discuss.

Oh, and I have an announcement to make. We are making renovations to our home, including the room where my system stands. The renovation in that room will take about a week, and we've decided that it's much more convenient for everything to stay where it is covered with plastic and/or cloth sheeting to protect from dust. This includes my system, which will be glad-wrapped and wrapped in a plastic bag. Obviously because of heat dissipation issues, this means that the system will be out. The best estimate we have is between 5-7 days, though knowing the nature of building schedules I wouldn't be surprised if it were up to a fortnight. The best estimate of when this starts is on Monday.
 
well thats lame(quoted text)..whats the point in even releasing the client at all until the opencl implementation is done correctly.. because otherwise its just going to be another gpu2 client with a list of promises and nothing ever happening.. oh well screw it im done waiting for it anyways.. going back to boinc with my gpu now..
 
But this isn't a release yet is it? So all this is just fluff, meani g they aren't going to release a half-ass client correct?

And this could also be good news because how I read it this means were closer to have a simple to install client, like a new standard client. That's what I get from that anyways
Posted via [H] Mobile Device
 
So does that mean we wont see SMP2 for a long time?
 
nitrobass24 said:
So does that mean we wont see SMP2 for a long time?
Click to expand...
This update has nothing to do with the development of SMP2 despite the reference Vijay made. In fear of violating my NDA, I can't comment much on the current status of SMP2 but I can say it will likely be released well before GPU3. :cool:
 
Vijay has since published some clarifications:

Just an update on my blog post. I want to make it clear that we're seeing problems in both of our ATI and NVIDIA GPU OpenCL implementations for OpenMM. Maybe it's on our side (I'm bet some bugs are on our side), but we suspect there could possibly be some issues with both the NVIDIA and ATI OpenCL implementations (and maybe even Apple's). We are working with everyone closely to either fix our bugs or get our code compliant with the OpenCL implementations.
Also, it looks like I should have been more careful in my wording of my previous post. I was referring to the issues above when talking about whether the ATI implementation was "fully functional" and some groups have gotten this out of my intended context. As far as I understand, the ATI OpenCL implementation is fully functional.

Finally, it is important to note that our Brook code is still supported, just not actively developed so Folding@Home/OpenMM is concentrating on shifting to OpenCL, so no more active development using Brook.
Click to expand...
 
So... Translation?
Posted via [H] Mobile Device
 
"You will likely never see better support for ATI GPUs under F@H"

I suppose I had better run off and bury all these ATI cards in the desert like E.T. cartridges.
 
*idly waits for SMP2* :(

Can't wait to drop this Linux VM :D
 
After reading the stats for F100 or whatever it is, (512 SHADERS!!!) I'm Thinking of Selling my Dual 295's and getting some Dual Fermi's. Should pull less power and hopefully net the same or more points

Or just get some Fermi's in addition to my 295s...
 
Vaulter98c said:
After reading the stats for F100 or whatever it is, (512 SHADERS!!!) I'm Thinking of Selling my Dual 295's and getting some Dual Fermi's. Should pull less power and hopefully net the same or more points

Or just get some Fermi's in addition to my 295s...
Click to expand...


i wouldnt plan to sell them any time soon until there are some real numbers out.. for all we know it will draw more power then the 295's and maybe net the same PPD.. no one knows.. any numbers you see are based purely on what nvidia says which we all know its a load of crap.. the biggest thing that will tell the performance of the card is the gpu/shader clocks.. and nvidia refuses to release that information because they still dont have a working consumer level gpu..
 
I think the full chip will draw almost right up to the limit of the PCI-e specification - that is, about as much as a 5970 at load. And NVidia may have to make even more concessions with the dual-chip card than they did with the 295 to fit that into the PCI-e spec as well, and to make sure it doesn't spontaneously combust while gaming. But I read somewhere (I've forgotten where) it's going to fold good. Real good.
 
Mr. Pedantic said:
But I read somewhere (I've forgotten where) it's going to fold good. Real good.
Click to expand...

I read it too.

Rumors, and I say rumors were in the neighborhood of 30k ppd! :eek::eek:

But then you add up power usage and the cost of this thing....
You KNOW $599.99 will be the cost of these bad boys.
Not cheap!
 
sirmonkey1985 said:
and nvidia refuses to release that information because they still dont have a working consumer level gpu..
Click to expand...
They most certainly do have working consumer-level cards. They were demonstrating them at CES.
Mr. Pedantic said:
I think the full chip will draw almost right up to the limit of the PCI-e specification - that is, about as much as a 5970 at load.
Click to expand...
There is no limit to the amount of power a card can draw in the PCI-E specification. Maintaining PCI-E 1.x support and using two external 8-pin power connectors, a video card can use up to 375W of power, but there is nothing stopping nVidia from either dropping PCI-E 1.x support to give the cards an extra 75W of power, or from adding more PCI-E power connectors.

By the way, a 5970 does not use 300W of power at load.
 
You must log in or register to reply here.
Back
Top