• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

OGR team - how big a CPU pool

JonB

JonB

Platinum Member
By my calculations, and they are based on my own small herd values, I have determined that the AnandTech OGR team is using about 230 GHZ of processors towards OGR.

That translates (using Mika's speed calculator) to roughly 200,000 RC5 blocks.

Anybody (please?) want to try to verify the data? I won't tell you how I did it until I am proven either very smart or very stupid (wouldn't want to prejudice the calculation) 🙂

Why did I bother? Because I wondered how much an impact we (the OGR lovers) could have if we switched back to RC5 for a week or two or three.

I was not really impressed by the 200K block figure. I hope I'm wrong.

 
I think 200K is about right. From what I've seen, a Gnode is a little less than twice as big as a RC5 block. Since we did what, 120 TNodes yesterday, about 200k blocks would be about right.

As far as swtiching to RC5 for a while, I might be willing to switch back. My contribution would be negligible (a little more than a thousand blocks a day), but every little bit counts, right? 😛
 
I'm throwing my work herd back to RC5 (from OGR) for a few weeks. I'm toying with the idea of submitting to the SETI mini-team as a show of respect... Especially since my old account is still getting phantom flusher blocks and it just isn't as satisfying to watch my rc5 stats under that ID...

Yes, I think I will submit to the SETI mini-team!

JHutch
 
Well, now we know a true figure, I guess some of the hardcore RC5 guys will stop complaining. In fact, the fact that we are placed so highly in OGR with such little power is amazing. We could absolutely trash it with a complete use of our power.
 
By my calculations, we would be doing about 200,000 WUs. Since many of us are using K6 series processors, my guess it that the number would actually be lower.
 
NOOOOOOOooooooo.......

We OGR'ers need as much help as we can get. /. is leading, not by much, but they are slowly pulling away.

Let's win OGR!
 
Additional help from the OGR team will not resolve the problem facing our RC5 team.
Only by increasing our active member count can we resolve the problem.
Most people have come to that stark realization.

If someone from another DC project is willing to donate some hardware, this will help, but it is not the final solution.

 
Just to add some more calculations, since it appears my 200,000 block figure is "close enough."

It took 86 individuals to do that much work.

AT needs 800,000 blocks to match DPC or 1,000,000 to regain the daily lead

That means a lot of recruiting. 300 to 500 new members.

though 600 would be nice.
 
JonB, Thank you for doing the calculation and publishing it. I thought of trying to do it, but didn't know which multipliers to use since they are so different. If you assume most of the machines are K6's then you get one result. If you assume most of the machines are Celerons you get a radically different result. I would be interested in your assumptions and methodology. If you feel they would be too boring here, then PM me.
 
Jon-

I think that the 200k/day of RC5 is a good number to play with if for no other reason that it is nice and round.
giantnt4vizmechthunderstorm is 7.58% (OK maybe closer to 10%) of TA's total OGR output. Although I am not opposed to moving my herd around, perhaps it wouldn't hurt to talk efficiency first to improve our total standing in all 3 major contests.
Although SETI has changed with the new clients I don't have a real feel for what is the most efficient platform for doing SETI now.

As for RC5/OGR we should encourage EVERYONE with an AMD K6 to run OGR. They suck at RC5 but almost match the P2/P3/Celeron in OGR.
Pentium MMXs are just the opposite. They need to stay away from OGR and do RC5.
Many output trades have been done in the past with these 2 CPUs. This should continue to be encouraged.
If we catch Slash and stay ahead of them in OGR, I would happily move some of my P2/Celeron power back to RC5, where they are quite effective.

viz
 
I know that our OGR team uses a lot of K6's. I would venture a guess that the percentage of K6's is higher than the percentage of Celeron's. Without a poll or a random sampling I can not say with certainty.

JonB - Thanks for taking the time to create and publish this data. Still looking forward to meeting you.

Regards,

RC
 
Well for me, this is a comparison of my 24/7 (or near it) herd.

K6- 1233 mhz
P2/P3/Cel- 6338 mhz
Athlon- 600 mhz

so my K6s are what- 15% or less of my total output?

Wish they were ALL 24/7....

viz
 
My OGR 24/7 herd:

AMD K7 (Athlon slot A, Duron) - 7.85 GHZ

It would be intersting to hear from LeBlatt. I believe he has a lot of K6s on OGR.

If we continue this thread we may find out that the percentage of K6's is not as high as I had originally thought.
 
I have 5 K6-2s in the 300-450 MHz range (most on upper end of range)

For an average they output 22 Tnodes in 1 month if I'm not mistaken
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top