L12 mod

I have done some testing and if i have with my original 1700+ 133 fsb multi under 10x the bandwidth stays the same but with bartons 166 if you run a multi under 10x the score is low. So there is you problem Tension_36th. Trie to run after the mod above 10x multi.

Quote:

---

Originally posted by Tension_36th
danne980

WHAT ARE YOU TALKING ABOUT?

my screenies speak for them selves and there is another screeny posted by someone else at 240FSB he is getting way less than what he should be getting, but i guess he hasn't taken note of it....

at 232FSB am getting roughly around integer bandwidth aorund 3450 mb/sec

after the mod at 250FSB am getting just a lit over 3500mb/sec....
you should be alot closer to 3800mb/sec...and the efficiency goes down....

Now if ya want a pic 232 after and before the mod as an example i'll get it for ya no prob.....but its been proven by some other with their screenies....for instance blinky, his 232FSB and 240FSB are very clear......LOSS OF bandwidth......

guys get 3700m/sec at 240FSB it self, when they don't have to do the "MOD"....

so i really don't understand what you are talking about....

read page 23 again...PLZ...before you draw you own "conclusions"

about multies...yes higher multies do yeild better results but i have compared some them with same multies...go through them again....

and about mentioning voltages?....ahem what difference can it make to the bandwidth....other than stability?..... :D

And yes my system is stable....

So you leave me very confuse with your statements......screenies are no proof? what you want a video or somethin??!!!!!.....but then again you ask for screenies for comparison?


And megatron.....i don't know what on earth you were doing when you were going through the "last pages"....

OF COURSE I HAVE DONE THE MOD....i have metioned it clearly Plz read page 23 again!......

Oh man!, why don't you people have some patience when you read things!

And guys, about being a big shot......if you haven't noticed it yet thats a little joke i and antipop were sharin.......are you both humour impared?

Tension_36th

---

Sorry Tension, i may be a bit bad on expressing myself in english. The point i wanted to make was:

1.I have noticed and read about diffrent multipliers gives diffrent mem bandwidth.
2.I know that total speed also impacts on bandwidth.
3. I belive that there can be a loss in point when oc to the max (mem, cpu, gpu all). So called sweetspots are there many ppl talking about. Thats also why i thaght the Volts should be thesame.

That why i did think there are to many diffrent things that could make the results missleading. So is strange that i think it would be better proof when benchmarking with the same setting before and after.
You wrote like this:

Quote:

---

I i have stated with proof in the above few posts.....

---

And sorry but i dont think that what you showed on page 23 was proof (but it sertainly pointed in that direction).
Proof isnt that to show that something cannot be another way?

But as i (was trying to anyway) wrote in my first post. I DONT think you are wrong. But i think you proof could be better shown.

What are you writing about you have tryed with the same multis and i should go through them again?
I did and what i see is:
first bench is mp11
second is mp8
Third is mp8.5
Forth is mp 8.5
Both of 3-4 is after the mod isnt that right?
Maybe you should read them again?

I wasnt trying to be rude when writing that about "big shot" either, didnt realy think about it at all, as you said it was a joke to begin with and i didnt med to be harsh on you. Didnt try to put you down but when i read it again i can see that it may be could be taken wrong. Sorry about that, not my intention.

What was realy so wrong with my post anyway, isnt it by questioning and reading one shall learn thing throw this kind of sites? And i have read and posted about something i wanted to know. Why do you then write that I/we arent tryingt to learn?
After my post you have been realy rude and i am awaiting some kind of appology, i wrote that i was sorry about that big shot joke and now its you turn. Cant be calling ppl stupid just for not agreeing with whats proof!
Btw you said you thought that I wasnt knowing my stuff about oc, and i can admit that i arnt very experiensed but i know that when changing so much as was changed between thoos screenies there are many things that could be affecting, so maybe you could learn something from me as a noob.

Maybe you should try one thing we got here in Sweden, Valium.

@danne
Y have you got valium in Sweden?
Cold weather? long winter nights?
I thought all you northern folk were internet freaks

We have it so that we can be easy when ppl say we are stupid on Forum sites.
But about internet freaks i dont know, i think there are a big number of swedes on internet, almost everywhere one goes there are a bunch of swedes. But it might just be me looking extra hard for them?

Think that long cold winters was the best with Sweden btw, snowmobiling is great.

here are my results with the L12x mod and without:

No L12 mod-
225Mhz x10 CL2.5
14014 3DMarks2001se, Sandra MemBench 3423/3175

L12 mod-
225Mhz x10 CL2.5
13937 3DMarks2001se, Sandra MemBench 3395/3145

Conclusion: Without L12 mod its a little faster.
BUT
with L12 mod I can run upto 240Mhz much faster overall.
225 Mhz FSB was my stable limit before the L12 mod.

:) :) :)

After I received an ear bashing from Tension, about not reading things correctly; it seems too hypocritical to be taken in seriousness. I pointed out that he had misread my post but I just let it slide at the time. I just brought it up since, if u can make silly mistakes like that on a simple subject then you could well make mistakes on a more complex one.

Setting aside all the unnecessary abuse Tension is putting about. Ok if u tested and your results are proved to be correct by independent tests then I agree with your results. However are you claiming that the L12 mod is not worth the effort since you get overall less bandwidth or simply a slightly diminished return on the results extrapolated if the same fsb were possible without the mod?

Also earlier you dismissed the L3 8-bit mod which with the Tbreds would allow the CPU to be set up without using "user defined", if I got the gist of it, Mr. PC Ice said that this could also allow a certain advantage when trying to maximise your fsb.

All this leaves me wondering if you are a new member here because you upset a lot of people at your old forums...

I was under the impression that the CPU interface was the 1t command rate setting, and possibly it was not activated whatever the bios setting on multipliers lower than 10, but that is just conjecture.

I am a little shocked at the responce from Tension since I completely forgot about this thread and was not email notified. I was otherwise distracted because I just recieved a r9800 pro 256Mb xt retail (Sapphire) for £88 + del:
http://forums.hexus.net/showthread.php?t=15162

I dont mind you finding things out and proving them. You make everything you write too long and boring because of all the insults. If u are still wanted at that forum then I suggest u go back because, you do not try to answer my questions and try to insult me all the time. Its funny I only meet people like u on the internet probably because u feel safe insulting me from so far away. You are a computer engineer? I am also I can put them together and have installed windows. Wow I am impressed.

Yes I have heard of signal to noise ratio it came up in my 3rd year project in electronic engineering in the context of fibre optic data transmission.:toast:

EDITed by IFMU
Good lord... megatron this post is above and beyond outright flaming. You have plenty of posts here so it shows you have been here for enough time to know our forum rules.
Consider this an official warning, you may want to reread the rules to refresh your memory.
IFMU

Quote:

---

Originally posted by Tension_36th

LOL how can i know if the mods giving a negative impact on bandwidth with doing it?.....your obvious mistake number one.....

No: 2 you say that "got a link to that proof you were talking about?".....em what was that on page 23?, and on oage 24???.....are you going still ask this a gain?.....

---

Read what I said again if u don't understand. People of inferior interlect will struggle to understand a simple single sentance. I said u did the mod but is it still applied?

No2 I am not wasting my time searching for your proof amounst all your other drivel so I asked for a link.

In the end I don't give a rats ass what u found out it still don't change a thing cause we get better results with the mod even a monkey can see that.

Note I didnt call u a monkey or say you had inferior intellect. I am not abusive.

And u didnt comment on the fact I got a r9800pro xt 256Mb for £88 + del, enough to make u sick isnt it?

Note I don't care what kind of mediocre responce u give because I just unsubscribed from this good thread turned sour all due to Tension 36th

Best wishes u insignificant person.:)

guys plz clear this thread up so it wont get locked. theres plenty of info here that many of us actually want to know. :)

good work tension.
keep it up.

You know tension, i have not planned (havnt got my cpu yet) to do any L12 modd just because of what YOU have written :)
(There goes the teori of me beeing angry just because bad performance showen)

Thank you for the help.
The first post i wrote was because i wanted to see more screen shoots also that you have given so thank one more time.

But please no more flaming on me. it was a bad post by me, but if you know what i meant you wouldnt have taken it hard. But i still think you where over the top on page 24. I may dissagree with someone but no one in a forum can make me that mad as you seemed to be.
Never claimed that your job wasnt appriciated btw.

Tension, are you sure you remembered to set CPU Interface to "Aggressive" or "Disable" it or whatever it is for your board?

Initially, I thought I had memory issues too. Then after running through the bios options, I realized that since I had reset the CMOS, I forgot to set CPU Interface to "Agressive." Doh! My scores were exactly the same after I switched that.

Quote:

---

This is the first I heard of this in 24 pages. I didn't quite understand wicked cause he said he does 250FSB but hasn't quite tested the limit. Sig is conflicting info too.

---

megatron, my sig is conflicting because I am not running BH-5 and my memory is not capable of reaching 250fsb. I was forced to run with a divider to push the max fsb.

What I can say is at the same 220fsb on the same multi, I experienced no bandwidth loss.

Easy up the guy doesnt even read this any longer he says, no point in flaming him anymore.

The bandwidth efficiency shown here doesn't look too sad (97% efficiency):
d22_215fsb_2 2 2 11 cpu interface enabled_11.5multi
This is after the 166 L12 mod. Perhaps this bandwidth is lost or gained with the correct use of flash method, i.e. use flashmenu 1.2 to flash or reinstall the nforce memory controller driver. I won't do a before and after because that would be too nerdy for my tastes.:stick: :banana3: :owned:

Has anyone employed the L12 mod on a new Xp-Mobile CPU with the NF7-S v.2?? Inquiring minds...

C

I plan too. just havent got around to it yet.

Quote:

---

Originally posted by charlie
Has anyone employed the L12 mod on a new Xp-Mobile CPU with the NF7-S v.2?? Inquiring minds...

C

---

Anyone???

I'd like to know also. If others have seen success with this I will do this today on my 2600+ xp-m. Should be here in a few hours. My mb is the nf7-s v2 also.

-NJJ

hey guys im fixin to do the L12 mod to my mb and i have a few quik questions? what do you use to paint the mb? i want to paint the back of the mb instead of painting the cpu's pins ill just flip the board over and conect the pins on the back of the mb you can see the same patern you see when looking at the socket from the top or the cpu itself. i just dont know what to use for condutiveness? do i have to use a conductive pen or can i use a led pensile? im too poor to buy a cunductive pen this week :(

thx perc,

just make a little U shape out of 0.5cm of copper telphone wire and stick it in the L12 pin holes under the CPU. If you drop it down one of the holes accidently then you can always use your hand vaccum cleaner to suck it out again.

dude perc, just get an old IDE or floppy cable and cut out one strand, and cut it open and take one strand of wire, then using needle nose pliers put it in the socket where u need to and it works great, its easily removable too

urrrrmmm :( ok il will try it that way. i know last time i spent 30 mins and never did get the little wire in those damm holes :mad: :mad:

*edit* and to think those hole are freakin big compared to the holes on the new 64/FX series mb's!!! i cant even imagine that one :eek:


thx perc,

use a pair of tweasers and a flash light :p:

Chalk up another L12 mod. I have a 2500+ mobile that was stuck at 200 FSB on an NF7-S v2. Did the 133 -> 166 L12 mod and I'm now running stable at 220x11.5. Time to move the FSB up and see where it maxes out. The same components were running 227x10.5 with another 2500+ earlier, so I know the board and memory have more headroom. I'll add more later.

:toast: Glad to hear it!:thumbsup:

so it looks like the L12 is really helping with the mobile barton.

because the infinity does just fine with the mobiles