2p & 4p only receive 8101

Yes. 8101 is a bigadv16 WU, so it will only get assigned to boxen with 16 or more cores.

 

welcome to the new age of big adv lol

It's not that bad really. Sucks, but meh, what ya gonna do lol

 

All I *ever* get are 6900's. Even on my 4P E5-4650 based box! It goes through 6900's at 4m30s/tpf and a few people, myself certainly included, have been impatiently awaiting the assignment of something properly *BIG* for it to crunch. All luck of the draw I guess.

 

My only bigadv capable rig in this new world has only got 8101 since its release

 

I'll be painted as a blasphemer here, but has any one considered using the OCN core hack to make the 16+ machine to appear as it had 12 cores, then fold on all 16+? I follow that core hacking to make the machine appear to have more threads than it has can be seen as a EULA violation. How about the other way?

 

Small hijack: My 2P 6128 (obviously) can't complete the 8101's in time. I've switched to SMP. Is there anything else I could / should do?

 

Word.

 

Delete this if you want, but I did try in desperation, and my 1 attempt failed. seems the higher true core count went through anyway.

Nathen, simple fix. turn off HT, 12 core run 15sec slower than 24 on 6901

Easy 6128 fix is just fold smp 12 until theyre gone.

 

Set the available memory in the client to less than 2800 = 69xx.

 

hmm... less than 2800M.... Hadn't thought of that... yea, that would do it. I think going less than 2024 will get you 6901s and 2s....

and no breaking the EULA to do it either.

 

That doesn't work either, though it does give you the larger regular smp units, which performed better for my setup by 6k ppd.

 

@Jester:

You don't need a core hack. Just use the right flag and your rig will get nothing but 69xx work units.

Stop the client, then replace the -bigadv flag with -bigbeta, in the client.cfg file, (V6.34), and save it. Then restart your client.

I use this in Tanker (my 4P) which has 32GB of RAM. Works fine.

I'm a bit ticked with PG for "playing" the "secret flag" game with us. United Devices was famous for that, years ago, and it became very unpopular with the volunteers.

I don't fold for the points, but I despise being "played" by a person or group, that I'm volunteering my time and money to help.

 

That's not how [H] rolls.

 

Actually, i think he was agreeing with Sbinh that we don't do core hacking

 

Lol, I think he was agreeing with 402 as well

 

The -bigbeta flag is confirmed as working, my L5640 just pull a 6903 for a 104k PPD total, thats 51k more than 3 hours ago

 

I may just have to fire up my hexen boxes to fold the 6903/4's. My D8 is getting garbage.

 

There you go!

 

It could do with the new version of the kraken as DLB doesn't engage properly with the A5 core but i'll worry about that in June

 

My hexes have never stoped getting 6901 / 6904 WU's. I still have not figured out why Stanford did it the way they did. I do not believe they thought it all the way through.

 

-bigbeta not -bigadv is all I have to say

 

Thanks for the tip, I remembered I had the FAH VM still loaded. Fired it up and got a 6901. YES!

 

I'm betting they never truely tested on actual 16 core systems. Cause as we've seen, even true 24 core systems just finish in time, and any burp, kills the units

 

As a reminder you will get no help if you get a beta unit and things go south...
As to why they got pulled to the beta flag I don't know...they either drug them there to die or to change the points down.

 

Pande Group would be wise to realize they will have a TON of furious folders to deal with, if they keep playing around with "secret" (and important) flags, like they've done here.

If only the Beta group and some select teams are going to know all the flags for the folding client, then OK - let the Beta group and the select teams fold ALL the work units.

Once you start playing "secret flag" games, FAH moves from science to playing favorites, and I'm not interested in playing favorites with PG.

By a rough estimate, this particular "secret" flag, has cost me personally 2 1/2 Million points, at least. Which I don't care about, because I'm not a points hound except when I'm racing or, like this month, have a particular point goal in mind.

But the idea that SOME teams are told the "secret" flag, while others are left in the dark, month after month, *REALLY* chafes my hide.

 

It's not really a secret flag, it was made public on the FF http://foldingforum.org/viewtopic.php?f=66&t=18031

Personally I fill it should have been left private but that was their decision.

I tend to agree with Patriot's take on why they got moved to beta, or maybe more to appease the "69xx are getting to many points" crowd.

 

It would be my guess that it was done to appease the ("69xx are getting to many points" crowd.) The only group of people that this move affected was / is the MP bigadv folders. All of the 12 core 8 core and 4 core bigadv folders can and are getting the 6901 / 6904 WU's using the regular bigadv flag. (and have been all along) Stanford kind of slapped those of us with MP rigs in the face with this one. And like I said before I can not figure out why Stanford did it.

I actually believe the allowing the 6901 through 6904 available to beta team members was an oversight by Stanford. Before they changed them from beta to general public release the beta team members only got 8101's with the beta flag which tells me the beta weighing and assighnment was removed from the 8101's but they forgot to remove it from the 6901 through 6904. (That is just a Guess though)

 

I'd guess rather than any nefarious reasons that the assignment logic is either flawed or too limited to handle every case correctly. There have been similar issues in the past - for example at one point if you were using the beta flag and there was no beta work then you got no work at all. In this case, nothing got moved back to beta (since 12 cores and under still get them without any beta flags), it's just that the fallback assignment for bigbeta when no beta work is available appears to be the "old" bigadv server rather than the "new" one (with 8101s).

 

It's interesting that some of you have been getting 69xx while I and my team mate who also has a 4p have been getting 8101s for a few weeks. For sure, there's some hidden logic here.

Perhaps, it makes sense because 8101s are harder to complete in time for some of you.

 

Posting up a flag only in the beta group is not a public announcement - not by a long shot.

 

Since that forum is open for viewing by the public it does make it semi-public. The only difference being that non-beta team members can't post in it. It's there to be read by all.

That forum should have remained private just like the DAB is private. Do we need to know what’s being said there...probably not.

 

So all the other announcements about bigadv go in the bigadv forum area, but a MAJOR announcement like this, gets buried in the beta forum, only.

If the flag didn't make much difference, I wouldn't care, but it makes a HUGE difference in points, and even decides what systems can fold bigadv, wu's, and what systems can't.

I don't see that as an acceptable way of handling flag changes. Not at all. This year's Chimp Challenge wasn't all that close - but it could have been. That one semi-public flag, could have decided the whole race!

 

This isn't a major announcement, or a minor one at that. All this is what someone tried and found that works. He is sharing that info with the folding community.

What would the team that won the CC have done...not used the flag. It would not have changed a thing. Ever team in the CC would have used it and the outcome would have been the same.

 

While that may be true I do not think it was intentional. Prior to the public release of 8101 all you got with a MP rig was 8101 if you were running the beta flag. After they were removed from the beta state there were no more bigadv WU on the beta server, I believe the beta flag just defaulted to next available bigadv WU's which was the 6901's through the 6904's which are on the server that supplies the 12 core bigadv WU's

I have personally been a beta tester for years and that has been the standard practice it just defaults to the next available beta or public server that has the WU's, since there are no 16 core beta WU's it is defaulting to the 12 core server.

There was never a secret message from Stanford saying anything about the beta flag getting the 690x WU's it just happened. I am upset enough about the slap in the face that Stanford has given to MP folders by there attempt to force the MP folders to do only 8101's while allowing lesser machines to fold the greater PPD valued 6901's through 04's (That was and is quite a insult to me and I am still upset about it).

If I had believed the flag thing was intentional I would have been using some pretty strong and colourful language to express myself.

 

I can not publicly answer that being a beta team member, but Post #24 has your answer