OCN putting the hammer down

As far as I'm aware, PG has never explicitly condemned or condoned thread spoofing. So all this discussion is pretty theoretical.
 
W.Feather said:
can you please elaborate on how buying ES chips is the same as running the thread spoof? .... In terms of the project, not the teams.
Click to expand...

I didn't say I felt it was. Some teams think that buying ES chips is against Intel's agreement with vendors and they feel you are buying stolen or illegal equipment to run your folding project on. If you look at some of Intel's paperwork for ES chips one could agree with that.

Using the below logic:

Intel never designed these chips to be sold for $350. If they could have done that they would have in a retail environment.

Schro said:
Because your processor is not rated to run at 5ghz as a stable, stock speed. If Intel was able to clock them to that speed reliably,then they would sell and warrant them accordingly. Due to the thermal and electrical stresses of running it above their "safe" speed, your processor is at a much higher risk for failure and/or instability over the course of time. Just because you're stable today, doesn't mean you will continue to be tomorrow when overclocked.
Click to expand...

So all the G34 machines that are overclocked for [H] would fall into the same category. Where you have to flash RAM to get the speeds you need and then flash custom bios information. Look, we all make tweaks to our systems to make them better. The point is that you can't just draw some arbitrary line in the sand that says a 15% overclock on G34s is fine and a 60% on SB is not but a 50% overlock on L5460 is. Stable is stable until its not stable. Every single folder who overclocks has to know that if their rig goes unstable they can make a negative impact on the project.

How about all the overclocked SR-2 rigs that just suddenly stop being stable? Are you going to tell all the [H] members with the L5460 ES chips clocked at 4GHz that they have to undo their overclocks because they could end up hurting the project when their rig goes unstable?

So we accept overclocking to make rigs return more points and do more science as long as they meet the base requirement in the first place? It just doesn't work for me.

If, with any modification, your computer is capable of returning the work unit to PG at whatever time they decide it needs to be returned, who cares? If there were no points attached to the project would any of us really care? I don't think so. I think we would be looking for the lowest TPF to maximize returned work units where we have to spend the least amount of money out of our own pockets. Just because they decided to attach points to the system shouldn't make the end goal any different.

Edited to add: And speaking on stable, what is stable? I have a 4.8 2600k that has never had a problem in the 8 months I have owned it doing SMP folding. Not a single issue, yet I have another one at 4.6 that crashed all the time. I set that to 4.5 and no problems. I believe the same can be said for just about any overclock project. Sometimes you just get bad hardware and that should not impact the overall project.
 
Overclocking takes existing hardware and makes it perform better, however the key is you are not making the hardware lie about what it is. Most of us work very hard to ensure our overclocked hardware is stable and use dummy units for validation. The same applies to ES chips, you can raise any point you want about what the hardware was originally intended to be used for but the fact remains that it is qualified to run the units. In either case, there is no inherent dishonesty.

With thread spoofing, YOU ARE LYING to the client to circumvent a restriction that was intentionally put in place by Stanford. This willful deceit is the key distinction.
 
R-Type said:
That strawman argument against ES chips isn't even relevent. Regardless of where the chips came from, they meet the hardware requirements of the project. Core spoofed systems do not.

One violates the rules while one does not, they are not comparable.
Click to expand...

Sorry I missed this one.

http://www.intel.com/support/processors/sb/CS-030747.htm

That pretty clearly states their position on ES chips. So all I was saying is that one could review that documentation and come up with the fact that you are using stolen computer equipment to run FAH on and that since FAH says you can not use equipment you do not legally own, it is is also in violation of the FAH agreement.

However like the core hack PG has never said using ES equipment is bad. I don't think ES chips are bad. I owned and sold some on this very forum.

So unless you can point to other rules that clearly state that core spoof is against their rules versus the previous argument of it goes against the spirit of the project I am going to have to say your statement is not factual.

The whole point and purpose of this discussion is because we don't[/] have clear guidelines from PG on what is or is not okay when it comes to the core hack. We started here with the "spirit of the project" argument and I think that is where it needs to stay until there is clear direction from PG, if we ever even get it.
 
I think we have said everything that needs to be said, and maybe a bit more on the above topics.

I might suggest moving the topic back to ppd of OCN or the fact that EVGA might have a 2nd team on their over take list (heheh).

Either way, it would be best to leave the above as is.
 
R-Type said:
Overclocking takes existing hardware and makes it perform better, however the key is you are not making the hardware lie about what it is. Most of us work very hard to ensure our overclocked hardware is stable and use dummy units for validation. The same applies to ES chips, you can raise any point you want about what the hardware was originally intended to be used for but the fact remains that it is qualified to run the units. In either case, there is no inherent dishonesty.

With thread spoofing, YOU ARE LYING to the client to circumvent a restriction that was intentionally put in place by Stanford. This willful deceit is the key distinction.
Click to expand...

Again, I agree with you. This is why I said it was a personal choice. I choose to report my equipment as it stands. My argument is not that people should use the core hack.

My argument is that if PG cared for actual performance they would not limit it to some arbitrary number of cores when clearly, newer technology with fewer cores is able to outpace more cores of older technology.

Please do not confuse my position with what I actually do. I do not use the core hack, that doesn't mean I want to prevent others from gaining the knowledge of how to do that. Also realize that the more hostile this gets the less likely I am to participate. [H] members are comfortable with what they do to get the most points they can, and OCN members feel the same way. I am not here to change your opinions. I am trying to point out that the overall system is flawed due to the arbitrary core count, and that all teams do things that other teams feel violate "the spirit of the project".

Kendrak said:
I think we have said everything that needs to be said, and maybe a bit more on the above topics.

I might suggest moving the topic back to ppd of OCN or the fact that EVGA might have a 2nd team on their over take list (heheh).

Either way, it would be best to leave the above as is.
Click to expand...

I don't know, I have the unfortunate position of liking both communities. When most of this thread was basically OCN bashing on how people were getting the points I felt I needed to let people know that not everyone is such a bad person. It really sucks to be called this:

Officially Corrupt Noobs

By another community you enjoy partaking in. I think its a little insensitive to be quite honest and I feel I have a right to defend myself being called Corrupt and a Noob.
 
Buckwheet said:
Sorry I missed this one.

http://www.intel.com/support/processors/sb/CS-030747.htm

That pretty clearly states their position on ES chips. So all I was saying is that one could review that documentation and come up with the fact that you are using stolen computer equipment to run FAH on and that since FAH says you can not use equipment you do not legally own, it is is also in violation of the FAH agreement.

However like the core hack PG has never said using ES equipment is bad. I don't think ES chips are bad. I owned and sold some on this very forum.

So unless you can point to other rules that clearly state that core spoof is against their rules versus the previous argument of it goes against the spirit of the project I am going to have to say your statement is not factual.

The whole point and purpose of this discussion is because we don't[/] have clear guidelines from PG on what is or is not okay when it comes to the core hack. We started here with the "spirit of the project" argument and I think that is where it needs to stay until there is clear direction from PG, if we ever even get it.
Click to expand...

While they may have not officially stated the core spoof was not acceptable, I thought it was pretty universally understood that the PG revision to the thread requirements and deadlines was in direct response to the rampant proliferation of thread spoofed 1155 systems. Do you disagree with this?
 
Hey guys good news... I finished a Uni WU on my Amazon EC2 t1.micro instance... 69 points! Just wanted to gather a data point on how fast the instance was.
 
R-Type said:
While they may have not officially stated the core spoof was not acceptable, I thought it was pretty universally understood that the PG revision to the thread requirements and deadlines was in direct response to the rampant proliferation of thread spoofed 1155 systems. Do you disagree with this?
Click to expand...

I do. I think they are looking at all systems involved not just 1155 to determine what they want for overall project performance. The only message I got from them was from time to time the need to evaluate current projects and deadlines compared to widely available desktop hardware and make adjustments. But lets be honest, if SB-E at 5ghz will meet the deadlines people will core hack.

So PG just needs to say its not allowed, period, and be done with it. Or they need to just start awarding people based on system performance regardless of cores.
 
Buckwheet said:
Again, I agree with you. This is why I said it was a personal choice. I choose to report my equipment as it stands. My argument is not that people should use the core hack.

My argument is that if PG cared for actual performance they would not limit it to some arbitrary number of cores when clearly, newer technology with fewer cores is able to outpace more cores of older technology.

Please do not confuse my position with what I actually do. I do not use the core hack, that doesn't mean I want to prevent others from gaining the knowledge of how to do that. Also realize that the more hostile this gets the less likely I am to participate. [H] members are comfortable with what they do to get the most points they can, and OCN members feel the same way. I am not here to change your opinions. I am trying to point out that the overall system is flawed due to the arbitrary core count, and that all teams do things that other teams feel violate "the spirit of the project".



I don't know, I have the unfortunate position of liking both communities. When most of this thread was basically OCN bashing on how people were getting the points I felt I needed to let people know that not everyone is such a bad person. It really sucks to be called this:

Officially Corrupt Noobs

By another community you enjoy partaking in. I think its a little insensitive to be quite honest and I feel I have a right to defend myself being called Corrupt and a Noob.
Click to expand...
I admit my acronym is inflammatory and apologize for that, it was a comment born of frustration with this ongoing issue. I hope you will agree that we have been largely rational and civilzed with our arguments against the thread spoofing issue. This isn't something we are going to solve here but we as a team are not likely to stop taking issue with it. I enjoy having a diverse membership here on the forums and hope you will continue to contribute.

Kendrak said:
I think we have said everything that needs to be said, and maybe a bit more on the above topics.

I might suggest moving the topic back to ppd of OCN or the fact that EVGA might have a 2nd team on their over take list (heheh).

Either way, it would be best to leave the above as is.
Click to expand...

Fair enough ;)
 
Buckwheet said:
I do. I think they are looking at all systems involved not just 1155 to determine what they want for overall project performance. The only message I got from them was from time to time the need to evaluate current projects and deadlines compared to widely available desktop hardware and make adjustments. But lets be honest, if SB-E at 5ghz will meet the deadlines people will core hack.

So PG just needs to say its not allowed, period, and be done with it. Or they need to just start awarding people based on system performance regardless of cores.
Click to expand...

Ah, we agree here. There is no rational reason an overclocked SB-E shouldn't be able to complete 6903's as they currently stand if Stanford lowered the core count requirement. However, if Bigadv is truly supposed to be only for high end multiprocessor systems then I expect new WU's to surface where they would fall into the same dangerous zone that 1155 rigs do with deadlines.

We have had many discussions that core count is an increasingly unscientific method of determining performance and a benchmark based method of WU distribution would solve both the procedural question at hand as well as the risk that you are too close to the deadline for a work unit.

Hopefully this is something the DAB can pursue this year.
 
Last edited:
Shrinking the deadline and uppling the core count is a "patch".

Most on the DAB agree that a mini bench system that gives you WU based on system preformance is the best way to go (and deadlines to match).

At current, building this into the client is not possible because of other projects and lack of resources. But the mini bench is the long term fix if it can be put into place.

PG DOES NOT want people just barely making the deadlines as a way of normal practice. The "extra time" is there incase the system needs to do other things while it has that paticular WU, but normaly the system gets the WU done faster. It could be seen as a "buffer" so instances of no points would be uncommon. However people have pushed it to the limits. This slows down research and the issuance of new WU. (new WU are based on the previous batch and cannot be generated untill the previous ones come back.)
 
KMac said:
They did. See #4.
Click to expand...

Best practice is just that. If you do not enforce it people will break the rules.

R-Type said:
Ah, we agree here. There is no rational reason an overclocked SB-E shouldn't be able to complete 6903's as they currently stand if Stanford lowered the core count requirement. However, if Bigadv is truly supposed to be only for high end multiprocessor systems then I expect new WU's to surface where they would fall into the same dangerous zone that 1155 rigs do with deadlines.

We have had many discussions that core count is an increasingly unscientific method of determining performance and a benchmark based method of WU distribution would solve both the procedural question at hand as well as the risk that you are too close to the deadline for a work unit.

Hopefully this is something the DAB can pursue this year.
Click to expand...

Agreed on all accounts, and I do agree with being civil. Out of common frustration progress can be created. Sounds like we have done just that.

We agree that we are doing scientific calculations with a imprecise measurement of calculation capacity. I hope PG addresses this soon.
 
jebo_4jc said:
Great point
Click to expand...

Sorry I had to find the thread:

http://foldingforum.org/viewtopic.php?f=55&t=15882&p=158383#p158383

It was listed in their source on the FAQ. So basically they "discourage" it but are not actively declaring it cheating because you are not modifying the client.

This is what needs to change. You can't say the speed limit is 60mph but if you modify your speedometer to say 60 when you are doing 80 you won't get a ticket. Either its wrong or its not. They need to fix it.
 
I joined H just to share my thoughts here, I am an OCN folder.
First of all, I hold the opinion that corehack is fine, if PG wanted a 2600k rig to NOT be able to fold the 6903/4's, they'd have set deadlines accordingly.
But the focus of my post should be this - unlike OCN as a whole, I wholeheartedly disapprove of the HPCS program being used for (non-Anonymous) Folding.
I would also, on a slight tangent, like to thank H for infecting us with the 4P bug.
I have much respect for 33 and their efforts, contributions, and innovations in regards to FAH.
However, I do get the slightest hint that you guys look down at us for our opinions on certain Folding matters. This may be just me, but still, we can agree to disagree, and I look forward to having a friendship thrive between 33 and 37726.
Sorry if I went a bit off-topic here.
 
csm725 said:
I joined H just to share my thoughts here, I am an OCN folder. <snip>

... and I look forward to having a friendship thrive between 33 and 37726.
Click to expand...

I agree.

We all have the same goal: slay the dragon, and kill the beast.

Fold on!
 
csm725 said:
I joined H just to share my thoughts here, I am an OCN folder
However, I do get the slightest hint that you guys look down at us for our opinions on certain Folding matters. .
Click to expand...
Nah.

What you're detecting is our strong opinions. We do happen to think were the best team around, but we definitely don't look down on you. Maybe some other teams, but that's another story.
 
Speaking of hurt on evga lol, [H] is now 1.1 years out. You know what that means? That means we need to pour on the pain lol
 
My guess is that there are probably not too many of the top teams that are not using HPCS to do some folding. I don't think this skews the points very much if at all. Some may be doing it more effectively than others and if OCN is more effective that's great. They have a history of pulling together when opportunity presents itself.
Since the use of HPCS seems to be an equal opportunity access to the resource, and since neither HP or PG seem to have any problems with use for that purpose, it should not be viewed as foul play. As long as you fold in a way that will not corrupt the WUs, or keep them from completing on time, it should not be a problem to use it if one chooses to do so.
Fold on, brothers. We await your arrival.
 
I got approval for using HPCS .. but when activate server, it asks for CC .. it's scary ...
they said they would give users "plenty of time" before the beta is ended. Hell no .. I ain't give them my CC .. Don't wanna got heart attack when seeing the BIG bill from HP ... :D :D ...

Rather stick with all I have been runing at my house ... :D
 
I don't see the issue with putting in a CC. Imagine how many credit card disputes would happen if they just charged everyone out of nowhere. This isn't some fly by night company.

I am amazed at how fast my points have gone up since I started folding on the servers. My first point was logged January 1, 2005 for [H]. And I remember how long it took to get points back then, I think I had 14 computers at my disposal at the time. HPCS is just amazing. And why not for a good cause find new ways to fold?
 
sbinh said:
I got approval for using HPCS .. but when activate server, it asks for CC .. it's scary ...
they said they would give users "plenty of time" before the beta is ended. Hell no .. I ain't give them my CC .. Don't wanna got heart attack when seeing the BIG bill from HP ... :D :D ...

Rather stick with all I have been runing at my house ... :D
Click to expand...
I'll take your spot in the beta.
 
jebo, PM me, ill shoot you my info for it, i got in but am not going to attach a CC
 
Hey, I was trying to be subtle in my previous post, but H is definitely a better team than OCN and EVGA right now. There, I said it. OCN is pushing in the wrong direction and isn't advancing 4P as much as we should, and EVGA's Folding is unreliable, and most of it is done from GPUs.
By the way, guys, you CAN used a pre-paid credit card for HPCS with like $5 on it, so even if they charge you something without prior notification (which I don't think is very likely), they have no way to take it. ;)
 
A couple of my credit cards allow me to make virtual numbers and then close them when I want. So I made a virtual card and added it, then closed the account. They won't be able to bill it.
 
I gave it a try last night but ran into issue. HPCS would NOT let me create 8-cores instance(s). It only accepts 4-cores and limits 5 instances per server.

Jebo..I'll send you PM with info ....
 
sbinh said:
I gave it a try last night but ran into issue. HPCS would NOT let me create 8-cores instance(s). It only accepts 4-cores and limits 5 instances per server.

Jebo..I'll send you PM with info ....
Click to expand...

Thats what led to alot of the discussion on pages 1 through 3, people are applying a core spoofing hack to make the 4 thread instances appear as 8 threads to the folding client.

Just let it run SMP.


csm725 said:
Hey, I was trying to be subtle in my previous post, but H is definitely a better team than OCN and EVGA right now. There, I said it. OCN is pushing in the wrong direction and isn't advancing 4P as much as we should, and EVGA's Folding is unreliable, and most of it is done from GPUs.
Click to expand...
There's always room in the asylum for more. ;)
 
sbinh said:
I gave it a try last night but ran into issue. HPCS would NOT let me create 8-cores instance(s). It only accepts 4-cores and limits 5 instances per server.

Jebo..I'll send you PM with info ....
Click to expand...

It's broken...I've been able to create them, but it was hit and miss. Of the ones that got created, none were able to get a public IP.

This was all on AZ2 though....I already had working instances on AZ1. I haven't tried to terminate and recreate them on AZ1, since I figured I might not be able to recreate the ones I already have up and working.
 
Doesn't work on both AZ1 and AZ2 ... anyway, pass it to Jebo... it's his problem now... :D
 
L2F@H csm :D

BTW Guys, don't sign up, they charge like $250 on your card 6 hours after you sign up.
 
You must log in or register to reply here.
Back
Top