[H] trending up, EVGA trending down

Doubt it. You never know; we have a lot of folders with ATI cards around, and the increase in production might make it worth it for everyone else with an ATI card that isn't folding because the cost isn't worth the performance. But I doubt it. Even IF Stanford and ATI got their act together.
 
wonders if standford release a proper ati client if we would have enought ppd to cancel the takeover.
No, unfortunately it won't from my estimations, even if we had dozens upon dozens of untapped Ati cards.
 
No, unfortunately it won't from my estimations, even if we had dozens upon dozens of untapped Ati cards.
I reckon if they released a proper client we could get an average of 10-15k ppd per card, (pretty reasonable, I think, including all the lower end 3870s, 4670s, 4550s, etc, as well as the high end 4870s, 5850s, and 5870s), which would mean we would need roughly 200 cards to make up the deficit with EVGA. This is obviously above what any person could reasonably make up in a few months, but the [H]ard forums are huge, and full of people who jumped on the Cypress/Juniper bandwagon. But the main problems would be trying to get all these people to fold, and actually getting a client out.
 
I'm outputting ~11k PPD less now with my 9800GX2 out of the race.

Maybe if my company's first game will be released soon and sells really well I could get some fancy new hardware with the cash in a month or two *hint hint* ;)

What kind of game are you working on? Since most of us are also gamers, i'm sure we would like to know!
 
What kind of game are you working on? Since most of us are also gamers, i'm sure we would like to know!

www.nyanko.ws/ecd/ :)

We got a range of different games planned this year. Since this is our first game things are taking fairly long due to unforeseen things, but we hope to have a playable version ready soon.
 
www.nyanko.ws/ecd/ :)

We got a range of different games planned this year. Since this is our first game things are taking fairly long due to unforeseen things, but we hope to have a playable version ready soon.

Has your company thought about possibly getting into the Facebook game market? That market always seems to be into cutesy type games.
 
I reckon if they released a proper client we could get an average of 10-15k ppd per card, (pretty reasonable, I think, including all the lower end 3870s, 4670s, 4550s, etc, as well as the high end 4870s, 5850s, and 5870s), which would mean we would need roughly 200 cards to make up the deficit with EVGA. This is obviously above what any person could reasonably make up in a few months, but the [H]ard forums are huge, and full of people who jumped on the Cypress/Juniper bandwagon. But the main problems would be trying to get all these people to fold, and actually getting a client out.
I totally agree with you in the event Stanford finally released a good Ati client. We could probably catch up or at least push back the overtake to a year and beyond. OTOH, I expect EVGA would just make up for the increase in our production with relatively little effort and in quick fashion, like they did recently when they saw their overtake increased to the half year mark because of our efforts.
 
I totally agree with you in the event Stanford finally released a good Ati client. We could probably catch up or at least push back the overtake to a year and beyond. OTOH, I expect EVGA would just make up for the increase in our production with relatively little effort and in quick fashion, like they did recently when they saw their overtake increased to the half year mark because of our efforts.

i think their recent surge has more to do with the evga bucks program than anything else, a lot of the more casual folders have just got their fermi upgrades and look to just do the 100k or 250k a month for the next upgrade. Their hardcore folders are hitting the same problems as us, A3 crashing, low points etc.
 
i think their recent surge has more to do with the evga bucks program than anything else....
I respectfully disagree. I've been looking very closely at this on a daily basis and it's apparent to me that their program is likely not the only reason. I'm surprised I'm the only one who sees things otherwise.
 
Has your company thought about possibly getting into the Facebook game market? That market always seems to be into cutesy type games.

It's a highly competitive market which we'd prefer to avoid. FB games also risk getting cut off/sued at any point. I'd rather avoid that too :)
 
Regarding ATI folding...
I have no idea how different the calculations are, and I've heard that F@H sucks on ATI, but for dnetc, my HD5870 is about 6x as fast as my GTX285 was. ~1800mkeys compared to ~300mkeys. I think the CUDA client has actually been out longer and had more tweaking than the STREAM client too. At least theoretically, ATI could stomp Nvidia in F@H too.

I know I've said this several times, but over in the dnetc world, ATI is pounding Nvidia. I had the Fermi guys here run the dnetc benchmark, and the GTX480 overclocked a bit will do about twice what the GTX285 did. This is still less than half of what my HD5870 does. Obviously F@H's computations could be drastically different compared to dnetc, but a good ATI client (or even a level-playing-field OpenCL client) could mean huge numbers.

It's all theory and doesn't do anything for our numbers right now, but it is one possible improvement in the pipeline. OpenMM is working on adding OpenCL support, so hopefully that translates to a good OpenCL F@H client soon.
 
I respectfully disagree. I've been looking very closely at this on a daily basis and it's apparent to me that their program is likely not the only reason. I'm surprised I'm the only one who sees things otherwise.

Feel free to disagree:). Its not the only reason to be sure but it just seemed suspicious that on the 1st of the month their PPD jumped a lot. I also know that they are running a lot of fermi cards over there, and we are hurting due to not having tiger online at the moment, also the lack of A2 WU isn't helping and people on the [H]orde seem to be having endless problems with the a3 -bigadv WU
 
I have lost 15 to 17K PpD in production going from A2 Big Adv to A3 SMP. I lost too many WUs to A3 Big Adv and believe the lowered PpD is warranted for stability.

I am also about to take 30K PpD offline with the shutdown of my GX2's. Been swamped with work (what else is new) and haven't been able to pull them, but will be doing that this weekend. One of them is going to Kendrak, but the heat in my office is unbearable and I don't have a window unit in my current place. I don't want to invest in one till I know what I need in the new house which is only 4 months away.
 
I have lost 15 to 17K PpD in production going from A2 Big Adv to A3 SMP. I lost too many WUs to A3 Big Adv and believe the lowered PpD is warranted for stability.
I am getting sick of all the A3 WUs, to be honest, even the standard ones. A3 is not as accepting of overclocks as older A1 and A2 WUs regardless of OS. I have to reduce the CPU frequency a substantial amount to get these WUs to run well at all. Very annoying since I had little to no issues after so many years. Right now, the only thing that is propping my production is GPU.
 
Feel free to disagree:). Its not the only reason to be sure but it just seemed suspicious that on the 1st of the month their PPD jumped a lot.
What happens on the first of every month?
 
What happens on the first of every month?

They have to fold either 100k or 250k per month in order to get the evga bucks. It looks like they get a surge at the start of every month as everyone gets their points in then tails off towards the end as people qualify and then shut down.
 
They have to fold either 100k or 250k per month in order to get the evga bucks. It looks like they get a surge at the start of every month as everyone gets their points in then tails off towards the end as people qualify and then shut down.
Their current trend line supports your theory. http://folding.extremeoverclocking.com/team_summary.php?s=&t=111065

They started May around 8 - 8.25 mil ppd. They trailed off and ended May around 6.5 mil ppd.

Now at the start of June they're back up around 7.5 - 8 mil ppd.
 
They started May around 8 - 8.25 mil ppd. They trailed off and ended May around 6.5 mil ppd.

Now at the start of June they're back up around 7.5 - 8 mil ppd.
That means about a quarter of their production is due to the EVGA bucks program. But even without these people, they are still soundly beating us. Or is 6.5 now smaller than 4?
 
Back
Top