newbie reporting in
Just get some new toys,
G0 E2140 (Check CPU-Z) & Elpida chipset KVR

3.6G Pi 1M


1.8vdimm 1000MHz dual 32M stable


my rig


offtopic bench


found some issues
1) Blood Iron couldn't post with crucial 10th anny installed
Anyone try 10th+ BI combination before?
2) Blood Iron NB heatsink, can i remove it without voiding the warranty? It's very hot and kissing my HP-1204 air cooler

3) The ram slot hard to close/open with gc installed

4) Not enough JFAN port to use ( 3 only)
I have 4 JFAN ports for my AMD matx mobo TA690G
5) Issit possible to flash LAN Party UT or DK P35 bios into blood iron?
I need CMOS reload function badly

guidance & suggestion are needed, thanks in advance
will try to bench further once I'm free

This post has been edited by dblooi: Apr 26 2008, 04:38 AM

it's stock vcore, i haven't pump vcore yet
1.26v as stated in screenshot

24/7 underclock to 1.2GHz , electric bill is always a big issue to me

May i know your BI using which version of bios?
do you mind to show me some ram oc screenshot on BI? thanks in advance
I've installed a small fan on the stock NB heatsink, the temp is acceptable to me now
anyway thanks for the info mate, really appreciate it

NB temp doesn't change much at load i think,
not sure as i never observe b4.
yaya naked pc , thanks for the pic anyway

ya i'm sure it is G0 revision E2140
it's labeled as SLALS on the IHS of the proc
sorry was using hp camera all the time, pic not clear




Interested? i got 9 more with me
btw, the kingston value ram is awesome
running cooler a lot compare to D9 ,somemore can do 1000MHz with 1.8v


@kmarc
thanks dude, just now i try again with crucial 10th
it works after several tweaks
able to bench with my 10th, happy
here's my latest result for super pi 1M ,
almost sub 16s and pc keep on restart,
couldn't test further anymore, stop here and continue next day


cpu-z validation: http://valid.x86-secret.com/show_oc.php?id=352236

issit a compliment?
my 10th still having problem with blood iron, not stable
will try to push more once i figure out a solution for that

i got it from oversea

---

Added on April 26, 2008, 5:39 pmyeah handpicked

This post has been edited by dblooi: Apr 26 2008, 05:39 PM

thanks for the compliment
for those who had pm me for the G0 E2140
a not-so-good-news for you guys
these G0 E2140s are a bit weird,
not all of them sensitive to vcore ,
you might get a proc can oc to 3.6G with stock vcore,
but after u pump to 1.7G, still unable to push it for a little bit further
i have 9 proc with me now, but i'm not sure which of them behave like this
i will perform a quick test after wednesday
those who intend to break over 4GHz with G0 E2140,
better prepare yourself mentally first

Unlock the 4MB L2 cache is more likely an april fool lame joke to me ( they spread this rumour on april fool)
Allow me to analysis the situation to you

- You have to open the IHS in order to perform the mod on the circuit of the chip. In case they solder the core and IHS together instead of using thermal paste, GG to your proc.

- You have to understand the circuit. Also, all your efforts are wasted if intel lock the L2 cache through microcode instead of modifying the circuit. ( But chances that the cache is locked through the circuit mod are high )

- We don't know what's the actual reason why intel downgrade it to E 2xxx. There is high possibility part of the L2 cache are found faulty during the fabrication process, that's why intel lock the cache and downgrade it from E6xxx to E2xxx. ( Come on, we know intel. They love business! They can sell these chips at better price if they let go them in E6xxx form what)

of course it is possible to unlock it , the chances of success are low and we don't know the L2 cache still can be used or not

This post has been edited by dblooi: Apr 28 2008, 01:36 PM

when you read an article, make sure you read everything
go to last page, read this

I'll translate for you guys

"attention: today is april fool!

there is possibility we can unlock the cache for these G0 E2140 , however we haven't study the internal design of the circuitry or the correct pin in detail, thus we advise people not to imitate this mod."

Obviously , the author already told everyone this is just a april fool joke.
i wonder, why there are some people keep insist on this ?

new super pi 1M result, 15.766s
looks like sub 15s for E2140 is possible


cpu-z validation: http://valid.x86-secret.com/show_oc.php?id=354370

nope, i'm using PC Cooler HP-1204


Currently it reaches the limit , but I'm not sure it is limit of processor or cooling
tonight I'll setup chilled water cooling to find it out ,
Will back to you guys after that

almost sub 15s for super pi 1M
prediction was right, the heat is bottlenecking the oc previously
use chilled water cooling, able to push a bit more only,
this baby is so hard to serve
best try so far, 15.281s


cpu-z validation: http://valid.x86-secret.com/show_oc.php?id=355388

the idle temp is 13C for 1.5xV at 3.8GHz,
for stock vcore and stock speed, idle temp is -3C to 0C at coolest moment

mind to show the link where you get the picture?
Go to the last page, read the last line

"Happy April Fool"

btw, there are some forumers already argue with me on this.
they tried to pursue me to believe that this is true,
they sent me few links, all are this april fool jokes.
ok, funny things happened,
when i replied them, why don't you get one from me and test yourself?
these are replies that i get from them,
"i'll buy from oversea instead of getting from you!"
"Your proc sucks, that amount of money i better get E7200! lolz"

I do not know their motives, but i know that this is not ethical for doing so.
Why they crop the picture only and post in the BSEL mod thread?
They can provide the whole link what?
They don't dare to show you the link, i'll do

Go through this link carefully ( i assume that you can read chinese)
http://diy.21tx.com/2008/04/01/10060_2.html

FYI, the mod instruction that you uploaded is available in page 3.
Also, make sure you go to the last page, read the last sentence



simple translation from me ( again and again )

"p/s: today is 1st of april ( April fool)

It is possible to unlock the cache, but we haven't studied the circuitry and location of pin in detail, we advise people not to imitate this.

However, the 333MHz mod is real, you guys can try it on mobo that doesn't support it. Make sure that CPU has oc headroom that capable to work under this frequency."

Now i wonder, what's is the purpose when one already know it's fake,
but he still crop the picture from the link ( why don't dare to post the link? lolz) and say dblooi has this processor but he dowan to try it because he thinks that this is a joke.

honestly, i dunno why.

Lastly, i'll repeat something that i mentioned before for one last time
1)There's no way you can unlock multiplier and L2 cache by performing BSEL mod (You can control the FSB through mobo, but can you control multiplier ( increase it beyond limit) and cache size through motherboard? BSEL mod is basically changing something where you can do through the motherboard, eg increase voltage and fsb) unless you open the IHS and perform mod on circuitry on the PCB( Intel are not stupid, that's the reason why they put the IHS from socket 478 era onwards). To do so, three requirements must fulfill

i) The core and IHS are not soldered
ii) You understand the circuitry on PCB
iii) Intel do not lock the cache through microcode

2) Intel do not lock the cache for fun. They can sell it at better price if they remaining these proc as E6550. One solid reason is the caches are found faulty during fabrication process.

This post has been edited by dblooi: May 17 2008, 09:06 PM

until now you still can't stop dreaming for unlocking the cache?
what make you so optimistic ? the greed of men?
it is possible but it can never be done by BSEL mod!!!!!!!!
if you know the difficulty and risk of circuit mod probably you'll never think about it anymore
Anyway, i'm sure you already know the news is fake,
mind to explain why you still poisoning others with the fake news?

yeah i know many guys unsatisfied with the price and tried attacking me in various ways,
i didn't jack the price for huge profit, the price is expensive bcoz supplier side offering it at higher price.
It's handpicked 3.4G G0 E2140 , specially selected for its oc ability.
and they travel long way from oversea to me. Expensive courier fee charged
I will neither sell it cheap as RM 220 nor expensive as RM 440. Stop dreaming and attacking me.
i haven't start the sales and I'll start it tonight.

oh ya, please stop spreading the fake news for unknown reasons.
I'm very tired of explaining it again and again and again and again and again.
Thank you

This post has been edited by dblooi: May 18 2008, 11:35 PM

Tried DI oc on G0 E2140 yesterday,
best pi result so far, 14.875s,


CPU-Z Validation: http://valid.x86-secret.com/show_oc.php?id=367764

boot up on 3.6GHz, then top up by using setfsb
check the temperature, 7C under DI cooling
This G0 is heat monster , at 1.8v temperature is 60C with DI cooling

best shot, 500 x8 on E2140
http://valid.x86-secret.com/show_oc.php?id=367765


Thanks to J.J for benching with me in this dry ice session

here you go, 14.610s

CPU-Z validation: http://valid.x86-secret.com/show_oc.php?id=368476
Just check the 14.75s result, that guy uses blackops + DDR3 pair with E2140
what an imbalance combination

I tried 32M as well, something wrong again,
result slower a lot when compare to others
probably my tweaks cannot apply for 32M


CPU-Z validation: http://valid.x86-secret.com/show_oc.php?id=368478

random shot for this dry ice session

he pump extra vcore, that proc can stable at 4.1GHz with 1.35v