Trying to obtain 20k PPD

D

dcds1

Gawd
2FA
Joined
Sep 12, 2006
Messages
537
I have been reading through the forums over at F@H and I see where people are saying they are getting 20k with quad core Phenom II. I am running a 1075T and I am only avg around 15k PPD. How are people getting 20k with a quad core?
I am running at 260 MHz fsb, 2600 Mhz HT, memory at 693.3 MHz and core multiplier of 13 for processor speed of 3380 MHz. The reason for the 13 multi instead of the 15 is because 3380 was the highest the processor would go at 1.25 core voltage and be stable.
Any insight to this would be helpful. Thank you.
 
Two things I can think of - they are able to get a higher OC than you or they are running linux instead of windows.
 
Not sure how the quads are getting 20k, never heard of that amount for consumer and outside of a thuban

The people getting 20k on thubans are at 4.0 or higher, and typically on Linux
 
Ok this is someones reply on FoldingForum.org. (I didn't want to link cause I am not sure if that is allowed.)
QUOTE "Okay I went stock on my Phenom II 965 on cpu, memory, and cpu/nb and I am avg 16kppd. I avg'd out your 8150's numbers for the first 12 hours and came up with about the same avg, 16kppd. I am telling myself that it is probably the difference in WU's but that is a huge difference (or non-difference depending on how you look at it). I also take your 1090T's 11kppd in the consideration. something is not right for my 965 gaining 5kppd than your 1090T did. Is there something I am not considering?
I can actually get my 1075T up to 4.2 GHz but then I have to run 1.45 vcore. I chose the 1.25 vcore so the processor would run a little cooler and then just maxed out my cpu at that voltage. But if it take raising the vcore and frequency just to get higher PPD then it really isn't worth it to me. But I just cannot figure out how a quad core at stock is higher PPD than I am.
 
Wait a minute, I was under the impression that the x6 I just picked up will do about 10 - 15k PPD. You're telling me x4's are doing the same? How so? The x6 should do more than the x4's. Kind of got me excited as the CPU hasn't come in yet and I do intend to run Linux on it. I think my i7 870 only does ~15kPPD on a good WU.
 
He is bullshitting, a stock 965 won't get 16k ppd, especially not with some of the new units, last I knew
 
I'm not on the up and up with smp work units but there can be a pretty big point differential between bigadv work units. If you guys aren't running the same project that's probably it.
 
Christ, my i7 @4.0Ghz is only getting ~21K, so theres no way thats true.

Are you running a gpu?
 
We are talking smp units a 965 won't run anything else


Kardonxt said:
I'm not on the up and up with smp work units but there can be a pretty big point differential between bigadv work units. If you guys aren't running the same project that's probably it.
Click to expand...
 
You used to be able to get up to 20k PPD on a Phenom. Over time however they've made the WUs harder and on older gear lower output.

I used to get 20k PPD on my i7-950 at stock now I only get 8-10k PPD typically in Windows.
 
It is not me with the quad core getting 16k. I am running a 1075t at 3380 Mhz avg around 15k PPD. That is what I am trying to figure out what I am doing wrong because I have come across post at FoldingForum.org of people saying they are achieving 16k up to 20k on quad cores.
 
I have a 4.4Ghz 2500K , let me look at my HFM stats...

It seems on average its ~ 18K-20K but there are some workunits that tend to do much higher ppd, such as this one

This is from my 4.4Ghz 2500K:

Project ID: 7027
Core: GRO-A4
Credit: 600
Frames: 100


Name: CT-NewMain Slot 00
Path: crystaltones.teraknor.net-36333
Number of Frames Observed: 100

Min. Time / Frame : 00:01:30 - 51,260.8 PPD
Avg. Time / Frame : 00:01:31 - 50,418.2 PPD


FWIW Here is the 4.8Ghz 3930K on the same WU in windows:

Project ID: 7027
Core: GRO-A4
Credit: 600
Frames: 100


Name: SnB-E Slot 00
Path: localhost-36330
Number of Frames Observed: 88

Min. Time / Frame : 00:01:01 - 91,865.8 PPD
 
As I said, I call bull until they provide actual proof of their claim, since I have yet to see it reproduced, even with my 1090t @4.0 I rarely hit 22k ppd

dcds1 said:
It is not me with the quad core getting 16k. I am running a 1075t at 3380 Mhz avg around 15k PPD. That is what I am trying to figure out what I am doing wrong because I have come across post at FoldingForum.org of people saying they are achieving 16k up to 20k on quad cores.
Click to expand...
 
Yes the 7026 and 7027 are freak smp WU's my Gulfies @ 4.375Ghz to 4.5Ghz will get between 66,000 and 85,000 PPD on them but it is very rare for a person to get 1, I may get one out of every 200+ WU's. It is possible that the person calming 20,000 PPD on a 965 is only quoting the PPD for one of those WU's (some people do that) it makes there machine look better. I think my i7 2700K at 4.8Ghz is getting around 22,000 PPD when looking at a 500+ completed WU avg.
 
A 2600k at 4.4 gets 36k ppd in nix, all smp for more reference
 
Have not folded on it in a while but it was closer to a 19k avg
 
You must log in or register to reply here.
Back
Top