狂少's CDT-tweak

[hipro5]

Here's my humble contribution. Purpose was to see what's the difference between CDT and another copy waza method for Spi32M.

So; I've ran 2 32M yesterday @4.8GHz (wished for better clocks but the chip is FSB bugged @about 535FSB ); testing conditions are basically the same (see screenshot for hardware details). Both times, available real mem was about 543k with a difference of let's say 1k which is not significant IMHO.

1) run with CDT = 10m 33s
2) run after reboot with one of my cw .zip file = 10m 29s

Sorry for not having ran an untweaked one, but I'm quite sure it would have been slower than 10m 33s on my P5B.

To be back on topic. As it was said earlier in this thread, if CDT's really 60% of an entire method; so that's logical, it works and after comparison with another cw method, it can be improved.

So what you say is that you haven't seen "THE" gain out of it and it's a bit worse than Copy Waza?....

[aasmaukr]

I'm not going to say anyone is cheating as long as I think the score might be possible with some crazy tweaks. When I get my new HD and P5K tomorrow I'll try running the exact same subtimings and frequencies as yours together with the CDT-IV and some other tweaks and see if I can beat my own personal record @ 3,6G (which is not too good anyways ).

@ OPB: Did you use any other tweaks in that 13.75sec run? And could you tell me what theme you're using? Thank you

[before]

So what you say is that you haven't seen "THE" gain out of it and it's a bit worse than Copy Waza?....

On one hand, that's right buddy, no magic gain, not as efficient as another cw method.
On the other hand, it was said to be 60% of The method so if we believe that... what can we conclude if we don't know the remaining?

TBH, I prefer my method ATM... not as time consuming as 60% of CDT are, and little faster.

I was just curious.

[massman]

Before, are you using a .BAT file to do CW? :o

[M.Beier]

Im ashamed...

RAM EEPROM programmer
Is that freeware, and if not, where can I buy one...? I'd really like that, especially with my new DDR3 =)

.... Sorry for the off topic part above...

Kevin, I'd really like you to explain that neat result... I must admit, I find it obvious as fake... - I'd really like you too explain, cause, you dont admit it being fake - I really have something to learn from you if its valid... - My SPI 1M product is usually between 49800-51000, depending on ram speed and timings...

[Gaayam]

So what you say is that you haven't seen "THE" gain out of it and it's a bit worse than Copy Waza?....

you would be pretty good as a lawyer

[BeardyMan]

For what it's worth Kevin you got my apoligize.
You finally showed something that worked so it's all fine for me.

Those that are still on amd should try the tweak, those that are on intel should reconsider their steps in applying it.
At first i couldn't get it to work either, now i shaved of 0.8secs SPI 1M.

I have a C2D laptop but for some reason superpu won't run on it otherwise i would have proven the tweak on that system as well.

[mrlobber]

Those that are still on amd should try the tweak, those that are on intel should reconsider their steps in applying it.
At first i couldn't get it to work either, now i shaved of 0.8secs SPI 1M.

Nice.

Can you write step by step what you did? (including any pauses between copying etc) If you can, throw in a screenshot showing task manager memory sizes before the Pi run as well.

[BeardyMan]

Can you write step by step what you did? (including any pauses between copying etc) If you can, throw in a screenshot showing task manager memory sizes before the Pi run as well.

Ok i will do that right away

[Zeus]

I don't get it, totally puzzled here.

Freeing up the available memory works fine, i even got up to a point where i had more avaialble memory than system cache (see pic) but it gives me ZERO gain in terms of SuperPi.

Did all what it takes, maxmem set to 600, 512mb fixed pagefile on the same partition i run SuperPi from (D) free'd up the memory with this tweak but got no love.

[BeardyMan]

This is al i did and like said
stock cpu + maxmem at 680 =29.4secs
stock cpu + maxmem and this tweak = 28.8, 28.6 when i push it even more.

[hipro5]

I don't get it, totally puzzled here.

Freeing up the available memory works fine, i even got up to a point where i had more avaialble memory than system cache (see pic) but it gives me ZERO gain in terms of SuperPi.

Did all what it takes, maxmem set to 600, 512mb fixed pagefile on the same partition i run SuperPi from (D) free'd up the memory with this tweak but got no love.

So what you say is that you haven't seen "THE" gain out of it and it's a bit worse than Copy Waza?....

you would be pretty good as a lawyer

[Johnny Bravo]

So what you say is that you haven't seen "THE" gain out of it and it's a bit worse than Copy Waza?....

and there are other who have seen "THE" gain out of it and it's a bit better than Copy Waza hipro.....

...so mixed results rather than an all out failure so far

[Zeus]

So what you say is that you haven't seen "THE" gain out of it and it's a bit worse than Copy Waza?....

Nope, cannot say that cause i haven't tried copywaza yet but i will later on.

[eva2000]

and there are other who have seen "THE" gain out of it and it's a bit better than Copy Waza hipro.....

...so mixed results rather than an all out failure so far

that can be said for original copy waza too from my tests LOL :o

[Johnny Bravo]

that can be said for original copy waza too from my tests LOL :o

Now there IS something - I wonder if you randomly choose 30/40 people on this forum and asked them to explain how to do the "copy wazza trick" how many different ways you hear? 5, 10, 15?!

Everyone seems to have their take on what needs to be done - even if you do it the same the error margin isn't exactly small as I'm sure you or anyone else who's tried this is aware.

In all honesty we'd be better off agreeing on what exactly IS the correct copy wazza process, then making a comment on what exactly OPB is trying to do here as well. I fear however neither of these will get resolved anytime soon

*edit*

Eva you ever get a chance to see this:

http://3oh6.com/forum_posted/ocx/jas...pywazza-1.html

it's quite good - watch it (maybe a few times) and tell me who it stacks up against with you've done things in the past, may be of some use?

[massman]

What wrong with your explanation, Johnny? (http://www.ocxtreme.org/forumenus/sh...38&postcount=3)

[eva2000]

*edit*

Eva you ever get a chance to see this:

http://3oh6.com/forum_posted/ocx/jas...pywazza-1.html

it's quite good - watch it (maybe a few times) and tell me who it stacks up against with you've done things in the past, may be of some use?

yeah i have read that and followed the advice a few times not much difference from untweaked/no maxmem runs variations between -/+ <1-3 seconds inconsistently

[harleybro]

First I must say I applaud the work that OPB has done explaining this on OCX. Next it would be great to see more posts in this thread with results. I planned to run stock vs. CW (if anyone wants me to use specific numbers for this or a specific method feel free to say so) vs. CDT last night but had a problem with my OS install. If I can't resolve it this evening then I am sure I will have time to on saturday. I know in the past the large gains had been disputed at the same time I must admit the descriptions of Super Pi (how it works, and what part of the system is used for each section of calculations) I have seen from OPB have exceeded any I have seen anywhere else. Finally I would like to think I am a very neutral party in this dispute.

[Charles Wirth]

CES is coming up fast and I can arrange a bench off.

If that doesnt work we can take it over to fight club and use their cage.

[eva2000]

CES is coming up fast and I can arrange a bench off.

If that doesnt work we can take it over to fight club and use their cage.

hmmm that actually would be interesting to see.. the bench off i mean

[s7e9h3n]

CES is coming up fast and I can arrange a bench off.

If that doesnt work we can take it over to fight club and use their cage.

OPB's not going to CES.....

[BeardyMan]

Tweak seems to have a opposite effect on intel, can't figure out why.

I have installed xp instead of Vista on my laptop

and stock i get 23.2secs
with tweaks applied like on the amd system i lose time instead of gaining !

23.5

don't know why though, will try to figure it out later today .

[Gautam]

if you can understand this, I don't see why I can't, if I am a cheater with high definition..lol, would not I can't cheat with more skillful or even cheat it closer and cheat it more legit in the range of Vapor's stupid formula? so I can let the score inside of the legit range and I got WR as well
so that's why I said the whole thing is being set up wake up people

oh I remember about somethinglike when barry Bonds or Jordan...(of course I am NOT comparing I am like them, I am not a smart guy , far more than that) achieve some amazing milestone, they always were doubt if they take any drug..oh ya, now I understand... and is there anyone can even touch them, nope.

Just wanna say one thing Kevin...note my usage of the words "readily explained." Not once did I use the word "cheat"...you're the one that brought that up. We're ready for any sort of an explanation. Even reading you clearly state that you're using your EEPROM programmer would be some progress. ATM, as I said, we aren't getting anything but flames out of you.

Get the cage ready Charles, and I'll get my popcorn ready. Give the guys a couple of months to hit the gym first to make it enjoyable for the rest of us.

[aasmaukr]

What's an EEPROM programmer anyways? Some kind of software?