Here are a few recommendations for Einstein.
 
The BRP6 - Parkes PMPS XT has the latest most optimized CUDA and OpenCL apps. The newest version has narrowed the gap between AMD and NVIDIA performance. AMD still takes the lead though. For the purpose of points competition, the BRP6 application will give the highest daily production.
 
In the Einstein preferences, there is an option called GPU utilization factor of BRP apps. You can use this to set the concurrency.
 
Utilization Factor - Concurrent Tasks #
1.0 - 1, 0.5 - 2, 0.33 - 3, 0.25 - 4, etc.
 
For NVIDIA cards, in my testing a concurrency of one task or 1.0 factor is optimal. You may be able to see a small boost by running additional concurrent tasks but I am not sure the gain is enough to justify the extra heat generated with these cards. The CUDA application is less CPU intensive per task than the OpenCL application.
 
AMD 7970: A factor of 0.25 is optimal. Tahiti cards in general should run fine with 0.33 or 0.25 utilization factor.
AMD R9-290x: I found that running more than one task will give many invalid tasks. For this reason, I would suggest a factor of 1.0. The same may apply to other Hawaii cards but I have not confirmed that.
 
For HT enabled CPUs, I have found a small performance increase by either disabling HT or using Process Lasso to prevent the application from running on non-physical cores. In other words, running one task per physical CPU core. In addition, I would suggest leaving at least one physical core open for the GPU tasks. Due to the OpenCL application using more CPU resources, you may want to leave multiple cores open depending on what the utilization factor is set to.
 
If you are running an AMD card in Linux, driver 14.6 beta is an optimal choice. I did some testing in Windows 10 but did not see much difference between versions in this OS.

Done!
780 is out performing the 970. 780 on track to finish 2 in about 100 minutes and the 970 looking more like 140 minutes. 970 running pretty cool @ 51c and the 780 @ 62c. CPU set @ 75% on both.

Barbarossa
Bugger, i can´t seem to queue up Einstein tasks

Same here...thought I was missing something in my firewall rules. I was able to queue up by increasing my cache (cpu and gpu) and then suspening the network in BOINC manager.

If anyone has some free cpu resources, we could use help in Malaria Control to hold our current position.

Hi everyone, turned on malaria and einstein

Hey tanner2!
 
I'll be dumping a lot from all 3 projects tonight. Does it make any sense to start collecting Yoyo, or not worry about it because it's 7 days?

Yes. its woth it, my 2 i7  and S3 are also already running yoyo.
Or you can still continue running malaria.
Me thinks that  mighty dragon there on our back is starting to spit fire soon

way to go Devlin!
I think the true bunkering is still going on. A lot of Einstein+ Malaria bunkers are gonna get emptied midnight UTC.
 

Doesn´t make a lot of sense around position 14, but for the top teams.
It´s to hide where you gonna start your next attack, and doesn´t mean your´re neglecting a project. Don´t forget each position in each category counts, esp if you wanna win the whole pentathlon or a discipline.
For us its more looking back and defending. Or you can also call it running for our lives/ positions.

Tomorrow is the 9th, the day one may drop Einsteins.
Many may have bunkered Malaria or WCG together With them.
Einstein didn't have too many GPU WU's so the biggest GPU heads around may have been a bit wing clipped here. Depending on what they have allowed to be downloaded.
If they narrowed theis WU's to GPU only, they may have dried out.
Einstein CPU work has been rather tough to slower cpu's.
I'll release some stuff tomorrow, however, I had to let the air out of the Balloon early due to dead lines. My mistake there. But hey, work was delivered.
 
Also, this Challenge, this is a tactical bunker Challenge. Yes, I know some about the consequenses about bunkering, but I see this Challenge different. Actually, I see this Challenge as the only regular Challenge where bunkering is allowed.
Bunkering is normally a no no in my book, but this is a Challenge where bunkering is used tactical, and huge amounts of work is delivered.
The Projects should be thrilled, and I have never heard any Project admins complain. They rather gasp over the workflow delivered and they have theis servers troughly stress tested.
The Project admins should know about it in advance, and declined when SETI made request about them participating in this particular Challenge.
 
Looking forward to tomorrow, my Storage might be just a fart in Space but atleast it will be a .....yes, just that
 
 
 
 
 
 

Yes, the Einstein stretch of the contest has now started.

dropped what i had for einstein all that my lil 560 could produce over the last few days :) 

Malaria and Einstein dropped.
Keeping at Einstein but will let malaria dry out to switch later.

Evolution @ Home tasks are very long and don't know when to stop. I have 1 at 108%, 1 at 106% and several in the 90's So we'll see.
 
Also I have an ecm_es that's still @ 0.00% after quite a while.
 
So far this project isn't conforming to my common sense approach to crunching.

Ahem, I guess this may sound a little odd, but I happened to find a bunker
While much away these days, I have to remote in to watch over my "Children". One rig was overlooked when releasing the horde.
The bunker contained all still valid Einstein and Malaria.
Speaking of the Horde, it could be interesting to see if we got separated:
 
 
 

Orange_1050
Ahem, I guess this may sound a little odd, but I happened to find a bunker
While much away these days, I have to remote in to watch over my "Children". One rig was overlooked when releasing the horde.
The bunker contained all still valid Einstein and Malaria.
Speaking of the Horde, it could be interesting to see if we got separated:
 
 
 

We have a few late comers going to bring a couple GPU's to Einstein, but I think most of our GPU support was on FAH & GIMPS. Getting active members on board for the challenge is the real challenge. We were lucky that 2 of our most popular projects were selected. Otherwise we probably wouldn't be doing this well right now.