Hi Dookey

P

Patriot

[H]ard|DCer of the Month - March 2011/June 2013/De
Joined
Dec 15, 2010
Messages
2,502
Nice spot... I think I will take it. :D
CyLSISB.jpg


Am I doing this right?

http://stats.free-dc.org/stats.php?page=userteamcomb&proj=&team=[H]ard|OCP&sort=today

I am still figuring out this boinc stuff... but I like it.
 
I'm just excited that I even show up on that list :D
 
A lot of GPUGRID points goodness going on there. This is what happens when you have a few semi-decent GPUs: you fire 'em up, start crunching GPUGRID and watch the points roll in. :cool:
 
tjmagneto said:
A lot of GPUGRID points goodness going on there. This is what happens when you have a few semi-decent GPUs: you fire 'em up, start crunching GPUGRID and watch the points roll in. :cool:
Click to expand...

Yes.... a few decent GPUs :D

m37fVbZ.png
 
Looking good! Thanks for joining us. :)

How many GPUs does it take one to amass almost 5M PPD in GPUGrid?

Edit: I was too late in asking! :D The answer is apparently 27 GPUs!
 
metallicafan said:
Looking good! Thanks for joining us. :)

How many GPUs does it take one to amass almost 5M PPD in GPUGrid?
Click to expand...

WYSIWYG...

27 gpu cores from 8 k10s, 9 m2090s and 2 m2070s ...
I haven't decided how much more I want to add.

I have a ton of cpu cores to throw at projects yet... 490c/900t :eek:
 
Patriot said:
WYSIWYG...

27 gpu cores from 8 k10s, 9 m2090s and 2 m2070s ...
I haven't decided how much more I want to add.

I have a ton of cpu cores to throw at projects yet... 490c/900t :eek:
Click to expand...

:eek::eek:

Well welcome aboard indeed!
 
Patriot said:
WYSIWYG...

27 gpu cores from 8 k10s, 9 m2090s and 2 m2070s ...
I haven't decided how much more I want to add.

I have a ton of cpu cores to throw at projects yet... 490c/900t :eek:
Click to expand...

SO WHAT YOU WAITING FOR, GET ER DONE SON :D
 
I knew it was only a matter of time before you figured out those Tesla's. Ramp it up!
 
It's about time you came back and started challenging folks. I missed the banter from when everyone was folding and talking about "mowing" someone else. :D
 
I'm just struggling to stay in the top 10. Keep it up guys. Looks like this is the year of the Kraken. :D
 
DooKey said:
I knew it was only a matter of time before you figured out those Tesla's. Ramp it up!
Click to expand...

I think I have got them figured out... did 6m pts today. :D

Ok guys... Where do I point those CPU cores... they are primarily SB-E with some IB-E mixed in. I have been hearing primegrid is the goto place..
 
Patriot said:
I think I have got them figured out... did 6m pts today. :D

Ok guys... Where do I point those CPU cores... they are primarily SB-E with some IB-E mixed in. I have been hearing primegrid is the goto place..
Click to expand...

That depends.... Do you
A) Want to just push over all BOINC points
B) Want to help push our position in Formula-BOINC (which we are already number 1 in our League) by just completely annihilating the other teams in our league
C) Want to help push our team's standing in the DC-Vault
D) Just stick to a solid running project regardless of science
E) Want to stick to a solid science you care about
F) Want to just focus on the current Challenges Thread
 
Patriot said:
I think I have got them figured out... did 6m pts today. :D

Ok guys... Where do I point those CPU cores... they are primarily SB-E with some IB-E mixed in. I have been hearing primegrid is the goto place..
Click to expand...

Patriot you are going to have some problems with some of the projects not having enough work to keep that many cores busy. So your choices of what to run are going to depend on how much time you want to mess with the rigs. I highly recommend getting familiar with Bam and what it can do, it can be a great time saver especially if you do like brilong and use scripts with it.

I personally like to run all of my cores on one project at a time ( the holly shit effect ) but occasionally split them up when trying to do a push on some of the lesser projects, I have found a few that limit the amount of work allowed to individual machines, CAS@Home limits to 40 WU's per machine running at a time. SZTAKI Desktop Grid has a 1200 WU's daily limit per machine which takes about 2 hrs. to hit on a 4650 4P rig. EDGES has a max of 16 running on a rig at a time. I am guessing there are probably more that I have not noticed yet.

Some of the older more established projects will have plenty of work, right now we are lacking in the medical field as far as DC Vault goes our lowest scores are the medical field are Malariacontrol.net and Rosetta@Home. The PPD in the Bio/medical field tends to be lower than some of the others but it is irrelevant because everything is = withing the given projects. In the Math field there are several projects that have both higher PPD values and plenty of work NFS, NumbersField, Primgrid to name a few.

Then there are the lesser projects which tend to be a bit frustrating at times due to lack of work, run parley and just general problems. Bam can be very useful for these project because you are going to need to run several of them at time on any given machine. And most of the time you will need to run an established project with them at the same time to fully utilize you rigs.

Now get er done. :D
 
^^^ This, and no matter what you support, make sure to have a well established project that has tons of work set as a backup project just in case. Just pick one and give it resource share of 0.

SETI is a project that we have low standing in. The science is not one that I care much for or have much faith in. However, they use both GPU and CPU. The other downside is that it can take forever to get work validated and get your erm .. uhh.. I guess SETI still awards what are called points. But don't expect a lot. I know you weren't a fan of Einstein's points...SETI is worse. There is currently a challenge going for SETI BETA by the way. And another one following the current one that last a week each.

Climate Prediction has stats export issues right now and also lacks in work. So, I would not recommend them.

I would like to see how well Edges can handle work unit distribution. It is on my list of projects to watch and analyze for future project additions into the DC-Vault.
 
Alright... step one is create 2 work groups in bam... to keep from screwing up the teslas by pushing work to them... I have done that a few times already. I pushed a project that I had locally set to no more work and well... nothing happened.
 
You need to set your computing preferences in bam also, you can set everything in bam that you can set in the boinc manager. Just make sure it updates then set everything in boinc manager to zeros. Then synchronize with bam. At least that's the way I do it
 
: pokes dookey:

Don't just roll over.... cmon man RAMP.

Ks6Jaf8.png
 
Collatz is down so I'm losing a third of my production right there.
 
What have you rolled the cards over too? They should be able to hit PrimeGrid pretty well....

Edit: Spoke too soon...looks like that is where your points are at...
 
Gilthanis said:
What have you rolled the cards over too? They should be able to hit PrimeGrid pretty well....
Click to expand...

I am going to try primegrid today... I keep running gpugrid dry on work... lol.
 
Patriot should set up a huge DIMES box...I bet he could triple the teams total in a few weeks....lol
 
I need another gpu project for my 290's. Unfortunately it appears only collatz will work on them.
 
How is that performing? I thought Milkway was pretty DP heavy, and isn't the 290 limited in that area?
 
Milkyway is DP mandatory and most cards are limited in DP. It just depends on what you are comparing. Something is better than nothing right?
 
EXT64 said:
How is that performing? I thought Milkway was pretty DP heavy, and isn't the 290 limited in that area?
Click to expand...

I'll let you know after it runs 24hrs. Seems to be doing wu pretty damn fast.
 
Even though it's definitely DP heavy, even a lowly 780 Ti and even a 750 Ti seem to finish the work units in reasonably fast times. I'd expect a maximum of 14 minutes with an average of around 8 in there.
 
Kai, have you done any testing on how many work units to run at a time on the 750Ti's? I don't believe I have messed with it, but would be interested to know what they handle. Perhaps we should start a thread for discussing how many work units we successfully run at a time with each card before it craps out or their is no gain...
 
I haven't yet, but I can give it a go though. I'd imagine two might be the max on return.
 
I'm comparing to the 7970/280X. Will certainly be interested to see how it does. My 7970s run 26 seconds and 51 seconds for the two WUs they get. I thought I read (might vary by card or be wrong) that for MW multi WUs really just hides the new WU startup period, while both are running there is no benefit. That being said when they align right that can be quite a gain.
 
The shit talking in this thread sucks. I am outa here :D
 
EXT64 said:
I'm comparing to the 7970/280X. Will certainly be interested to see how it does. My 7970s run 26 seconds and 51 seconds for the two WUs they get. I thought I read (might vary by card or be wrong) that for MW multi WUs really just hides the new WU startup period, while both are running there is no benefit. That being said when they align right that can be quite a gain.
Click to expand...
Wow, nice. To give another data point my 7870 does the WUs in about 2:15 and 4:30 so the 7970 clearly has way more DP power. Anyone have any Nvidia data points or are they junk for MW?

sc0tty8 said:
The shit talking in this thread sucks. I am outa here :D
Click to expand...

You guys all suck. :D
 
Patriot is still a cocky little twerp I see, I love it. Keep it up. :D
 
You must log in or register to reply here.
Back
Top