Hardware Canucks

Hardware Canucks (https://www.hardwarecanucks.com/forum/)
-   HardwareCanucks F@H Team (https://www.hardwarecanucks.com/forum/hardwarecanucks-f-h-team/)
-   -   Quit Folding (https://www.hardwarecanucks.com/forum/hardwarecanucks-f-h-team/24069-quit-folding.html)

Exmortis October 7, 2009 08:23 AM

Quit Folding
 
As of today one can remove my folding sign, I have removed the software from all 4 machines I have working. I am tired of Pande group playing favorite camps, working on disease should be fair and even, but Intel and Nvidia get all the glory. ATI has released its newest card and still Pande group uses brook, how long can one ignore the favors being played. They still have not impelmented any AMD optimizations, but sure have for Intel. Nvidia was late to the party but given centre seat.

Worst part is I have 3 dual core Intels working, 1 AMD quad, which could have been a 5870, but no more.

Sorry all, I will move on, there will be no more updates on my Folding name. maybe SETI plays fair? all for the better anyway, I will save on electricity with machines no longer on 24hrs a day.

stoanee October 7, 2009 08:54 AM

Sorry to hear that, it can be frustrating sometimes. Thanks for helping out and if you change your mind please c'mon back to HWC!

Kilauea October 7, 2009 09:15 AM

It sure does suck, but I'm hoping there will be some changes made in the future (near) so that it works with a similar productivity independently of the brand. In the meantime, I still think that the littleI give is already a step in the right direction, no matter how small it is.

El_Fiendo October 7, 2009 09:59 AM

Folding Forum • View topic - Is PG getting ready for R8xx? [Yes]

He's an AMD rep, and he spends his entire time talking about implementing OpenCL as a way for ATI to get more points. Apparently the main reason ATI takes longer is because it doesn't have enough cache to store all the results it gets in intermediate calculations and thus needs to recalculate a lot more compared to NVIDIA. The fact that an ATI rep comes on and announces ATI is working on ways of improving their product for F@H makes me think you shouldn't be accusing Stanford of bias.

Further info Folding@home: Update on new FAH cores and clients

Remember this:
Both ATI Stream and CUDA were announced around the same time. However ATI couldn't get Stream out to the public until December 2008, a full 2 years after announced and close to 1 and 3/4 years after NVIDIA. What did NVIDIA's first GPGPUs pull for PPD? Well I don't know the numbers they pulled when first released because that was 3 years ago and I'm not sure what console support they had or had to wait for, but I know my 8800GTX pulled between 4k and 4.5k as of a few months ago. What I'm suggesting is, perhaps it'd be best to let ATI mature into their GPGPU architecture and actually develop for it. As far as it sounds to me, its a hitch on ATI's end, not Stanfords.

Dunno man, its up to you in the end. But I think you're giving Stanford the short end of the stick.

4DoorGTZ October 7, 2009 11:39 AM

I'm sorry all I heard was "I'm taking my toys and going home"

If you actually believed in the work done then you'd soldier through, sounds a lot like you were only in it for the points and contests. Perhaps you do believe in the work they're doing, if so hope to see you back when they better utilize whatever hardware you have at the time.

SKYMTL October 7, 2009 11:59 AM

Sorry, I take some serious issues with this thread. The way folding is done on an Nvidia card is totally different from the way AMD does it. There is a limited amount of cache on ATI cards which means results have to be calculated and distributed differently.

Don't blame Stanford. Blame the people over at ATI who never thought much about GPGPU's popularity and then released a half-baked "Stream Compute" interface just to pull their balls out of the fire.

Want to talk about BrookGPU? The language that hasn't moved an inch since something like two years ago? Optimizing Brook to run on Brook+ isn't the easiest thing in the world either considering all of the AMD hardware optimizations including double precision processing and other stuff.

All of these ATI blow-hards that are showing up on every forum these days really makes me wonder where all the intelligence went. It was Nvidia who spearheaded Folding@home on the GPU and it was ATI who said (once again) "me too!" and then started bitching. And you expect the development cycle to be the same? Come on.

Dr_BenD_over October 7, 2009 12:16 PM

I could have sworn I was folding on an Ati GPU long before Nvidia barged in the door.

El_Fiendo October 7, 2009 12:42 PM

Quote:

Originally Posted by Dr_BenD_over (Post 267152)
I could have sworn I was folding on an Ati GPU long before Nvidia barged in the door.

Right, until NVIDIA applied 'GPGPU's and CUDA to folding. ATI had a GPGPU program at the same time, but have done little with it. The guys at Stanford aren't experts in ATI hardware and coding and definitely shouldn't be targeted as not doing enough. Remember, they have CPU and NVIDIA coding to do as well. So why isn't ATI, the experts of their own hardware, stepping up and making as big a fuss out of folding as NVIDIA did? I'm sure that if ATI put their hearts to it they'd be able to code for something as relatively simple as F@H.

They talk about coding for OpenCL, which has only just come into the spotlight, so that eventually everything will fold well and can be done with one client. Though again, this takes time. They didn't develop ATI Stream or Brook, both of which have been out for awhile. So hopefully ATI develops stuff for OpenCL to make up for it.

somecanadianguy October 7, 2009 12:45 PM

Quote:

Originally Posted by El_Fiendo (Post 267061)
Folding Forum • View topic - Is PG getting ready for R8xx? [Yes]

He's an AMD rep, and he spends his entire time talking about implementing OpenCL as a way for ATI to get more points. Apparently the main reason ATI takes longer is because it doesn't have enough cache to store all the results it gets in intermediate calculations and thus needs to recalculate a lot more compared to NVIDIA. The fact that an ATI rep comes on and announces ATI is working on ways of improving their product for F@H makes me think you shouldn't be accusing Stanford of bias.

Further info Folding@home: Update on new FAH cores and clients

Remember this:
Both ATI Stream and CUDA were announced around the same time. However ATI couldn't get Stream out to the public until December 2008, a full 2 years after announced and close to 1 and 3/4 years after NVIDIA. What did NVIDIA's first GPGPUs pull for PPD? Well I don't know the numbers they pulled when first released because that was 3 years ago and I'm not sure what console support they had or had to wait for, but I know my 8800GTX pulled between 4k and 4.5k as of a few months ago. What I'm suggesting is, perhaps it'd be best to let ATI mature into their GPGPU architecture and actually develop for it. As far as it sounds to me, its a hitch on ATI's end, not Stanfords.

Dunno man, its up to you in the end. But I think you're giving Stanford the short end of the stick.

Quote:

Originally Posted by 4DoorGTZ (Post 267119)
If you actually believed in the work done then you'd soldier through, sounds a lot like you were only in it for the points and contests. Perhaps you do believe in the work they're doing, if so hope to see you back when they better utilize whatever hardware you have at the time.

Quote:

Originally Posted by SKYMTL (Post 267140)
Sorry, I take some serious issues with this thread. The way folding is done on an Nvidia card is totally different from the way AMD does it. There is a limited amount of cache on ATI cards which means results have to be calculated and distributed differently.

Don't blame Stanford. Blame the people over at ATI who never thought much about GPGPU's popularity and then released a half-baked "Stream Compute" interface just to pull their balls out of the fire.

Want to talk about BrookGPU? The language that hasn't moved an inch since something like two years ago? Optimizing Brook to run on Brook+ isn't the easiest thing in the world either considering all of the AMD hardware optimizations including double precision processing and other stuff.

All of these ATI blow-hards that are showing up on every forum these days really makes me wonder where all the intelligence went. It was Nvidia who spearheaded Folding@home on the GPU and it was ATI who said (once again) "me too!" and then started bitching. And you expect the development cycle to be the same? Come on.

+1000000

If ATI users aren't happy with the ppd they're getting buy an Nvidia card to fold with until ATI gets their :censored: together.
Or just sell your ATI card alltogether and go with Nvidia because their cards play all the games and fold like crazy too.
I'm sure eventually ATI will get things sorted out... it just may take a while.

Don't let stuff like this turn you away from folding.
It's for a good cause... shouldn't matter how many ppd you are getting really.
The competition is fun, but that's really not what it's all about.
Just do the best you can with your hardware or get new hardware.

Soultribunal October 7, 2009 12:58 PM

We all choose to do our own things for whatever reasons.
In the end though to be honest I don't care how we get to the conclusion so long as we get there.
If you choose to walk away, thats fine.
For me I am fighting tooth and nail to get back into it lol.

Any research done, any way we help the cause is worthwhile.

When one person donates a can of goods to the foodbank its nothing.
When 1million do it, people aren't hungry anymore.
Does anyone care if a lot of people donate the same stuff? Nope, because the ends are more important.

Do I care that nVidia is better than ATI atm? Nope.
All I care about is waiting for the day when the Dr's say "Oh thats just Cancer, here, take these and see me in two weeks".

My Mom is a double survivor.
And I have brushed close with it for 5 years.
So, yeah.

ST


All times are GMT -7. The time now is 09:13 AM.