8101 Point Adjustment

[kromberg]

I think you guys really need to rethink the points awarded on these WU. Processing time is about the same as the 6903 WU, but points are a bit less than 6901 WUs. I know of several people with 4P rigs already blacklisting and refusing to process these WUs. Just a heads up.

[ChasR]

p6903 and p6904 were anomalies and PG has said as much.

[7im]

And instead of thanking PG for all the extra points they get from P6903s, they violate the best practices rules by circumventing the Assignment Servers. Well done.

[Meh_Lay_Lay]

You can blacklist and refuse WUs @@? What's the point lol, someone's got to fold them, and your contributions to science would be the same.

[Macaholic]

I think you guys really need to rethink the points awarded on these WU. Processing time is about the same as the 6903 WU, but points are a bit less than 6901 WUs. I know of several people with 4P rigs already blacklisting and refusing to process these WUs. Just a heads up.

They have, and will continue to do so. Formal announcement was made here.

[Grandpa_01]

Yeah just fold on a single processor 8 or 12 core rig and you will not get them you will just get those worthless overvalued 6903 / 6904 which by the way are only overvalued if you are folding them on a MP rig.

Hint Hint

And which person from Stanford said they were anomalies. I do recall 1 post that said may have been or something like that. So the word may constitutes is. Interesting.

[Macaholic]

And which person from Stanford said they were anomalies. I do recall 1 post that said may have been or something like that. So the word may constitutes is. Interesting.

Start at the beginning of the eight page p8101 thread in the beta team forum of which you are currently an active participant. The answer lies there.

[kromberg]

Keep in mind these people are not the causal folder with a 6c/12t machine using corehack to get bigadv WU. These guys have 4 processor 48 core machines. The announcement said the changes should be unnoticeable. A 33% to 50% drop in point production is far from unnoticeable.

It is easy to blacklist a WU, there are many client wrappers out there that will do it.

Sure, the 8101 WU have to get folded one way or another. The Pande Group is certainly not giving any incentive for people with big MP rigs to do. Let the corehackers do them I guess.

[Grandpa_01]

And which person from Stanford said they were anomalies. I do recall 1 post that said may have been or something like that. So the word may constitutes is. Interesting.

Start at the beginning of the eight page p8101 thread in the beta team forum of which you are currently an active participant. The answer lies there.

Perhaps you should go back and re read that thread. There is no such comment in there from any Stanford member, Some people just amaze me how the wish to manipulate a comment to what they want to here.

[Adak]

Easy does it, Kromberg. The people you're chatting with in this thread are some of the most prolific folders in the project. Macaholic folds for several teams (including my team), and has 3 or 4 4P server rigs. His build log was the inspiration for myself and a few others, to build our own 4P rigs.

ChasR is a veritable encyclopedia of folding knowledge, as well as having hundreds of millions of points, most done when a wu was worth a LOT less.

So please, step down from your soapbox. We're all in this together.

When you see a project's wu's earning WAY high points, isn't it expected that others will be WAY less, at some point? That the road just doesn't go downhill all the time, there are uphills as well? That's expected isn't it?

Let's try and focus on the science, and not on our ego's. If a project's points are low, we should fold them, with the same mindset that we fold the high point wu's. Anyone with a new 4P folder, will be moving up the ranks, quite sharply. You can only go to page *one* - that's it - there is no more. Whether we're on page 1 or page 101, we need to keep focused on what we're here to do - fold! Pande Group makes 'em, and we fold em - high, low, or in-between - it doesn't matter. Think of the points as the beachball by the pool - just there for fun, not for swimming.

There's nothing wrong with making your request, or commenting on it, but can we keep our focus on the science? Working *with* the guidelines that the Pande Group give us, and not working on ways to circumvent those guidelines?

And now I'll get down from MY soapbox!

[Macaholic]

Keep in mind these people are not the causal folder with a 6c/12t machine using corehack to get bigadv WU. These guys have 4 processor 48 core machines. The announcement said the changes should be unnoticeable. A 33% to 50% drop in point production is far from unnoticeable.

It is easy to blacklist a WU, there are many client wrappers out there that will do it.

Sure, the 8101 WU have to get folded one way or another. The Pande Group is certainly not giving any incentive for people with big MP rigs to do. Let the corehackers do them I guess.

Yes, this has all been known for quite some time and it really all boils down to this;

Considering the great variety of donor opinions on this matter, it is no surprise that we agree with some donors and disagree with others.  Moreover, with points, there will never be any system which makes everyone happy, but our goal is to try our best to balance the project as a whole, taking donor input seriously, and making hard calls when we feel it is necessary.  This was definitely a hard call, but hopefully in time donors who disgaree now will come to understand the issues raised by the other donors and appreciate their point of view.

[7im]

Keep in mind these people are not the causal folder with a 6c/12t machine using corehack to get bigadv WU. These guys have 4 processor 48 core machines. The announcement said the changes should be unnoticeable. A 33% to 50% drop in point production is far from unnoticeable.

It is easy to blacklist a WU, there are many client wrappers out there that will do it.

Sure, the 8101 WU have to get folded one way or another. The Pande Group is certainly not giving any incentive for people with big MP rigs to do. Let the corehackers do them I guess.

Again, it's only a change because 6903s/4s were abnormally high in points. kasson decided not to reduce the points on those, but PG still could if this becomes too much of an issue.

8101s points were set correctly. They are not going to change. As Gramps often says, only PG knows the correct scientific value. You have to trust they set the points correctly, considering it's been benchmarked more than once.

[Patriot]

You have to trust they set the points correctly, considering it's been benchmarked more than once.

That must be why they have adjusted the points on individual bigadv wu 4-5 times after public release.
Both UP and down.

The fact that they adjusted the 6901 up rather recently tells me that the 8101 needs to be higher... approximately the work of a 6903 and gets less value than an 6901.

[ChasR]

... which by the way are only overvalued if you are folding them on a MP rig.

If that were true there would be no reason to use a core spoof to get p6903 and p6904, would there? In fact they provide a 157% increase in production over the average smp WU on a 2600K.

[7im]

You have to trust they set the points correctly, considering it's been benchmarked more than once.

That must be why they have adjusted the points on individual bigadv wu 4-5 times after public release.
Both UP and down.

The fact that they adjusted the 6901 up rather recently tells me that the 8101 needs to be higher... approximately the work of a 6903 and gets less value than an 6901.

Or it tells you 6903s need to be much lower, as kasson has said.