L625A348 E6400 ... 4Mb L2 ???

**phile** · 10-09-2006, 06:03 PM

For clarity's sake, the E6300 and E6400 chips ARE conroe cores with half the 4MB L2 disabled. CPUZ is the only application that tags them as Allendale, and they are mistaken. The genuine Allendale cores have a native 2MB L2, and certain conroe features disabled, like VT. I wish to God that the CPUZ people would stop perpetuating this error.

-phil

**Fred_Pohl** · 10-09-2006, 06:09 PM

Originally Posted by Evilsizer

wow if that is the case then kiwi i want one too! pretty please! yes for superpi times. the extra cache make a difference but in gaming and other benches overall increase from 2mb more cache is 8-10%.

use intels cpu id program, i dont know what it is called. i have only seen it used a few times in the past.

0-10% actually with about a 5% average increase from an extra 2MB of L2.

**sarachem** · 10-09-2006, 11:24 PM

Originally Posted by TaPaKaH

But the results were doudtful because the CPU was faster than most of conroes @ that freq

allendales, you mean...
i think that its possible... E6500 already =)
intresting to compare your exlusive processor with real conroe on lowered multiplier

**cpuz** · 10-09-2006, 11:58 PM

In order to get the REAL L2 size, run either cachemem or latency.

ID programs (BIOS included) use cache flags, and sometimes it can be buggy.

**theteamaqua** · 10-10-2006, 12:02 AM

Originally Posted by cpuz

In order to get the REAL L2 size, run either cachemem or latency.

ID programs (BIOS included) use cache flags, and sometimes it can be buggy.

yeah just runs some bench and ocmpare it with E6600 say both at 3GHz, and see how that goes, both at 8x multi so that the only difference is the cpu

**cpuz** · 10-10-2006, 01:05 AM

I uploaded cachemem for you :
http://www.cpuid.com/download/Cachemem.zip

**[cTx]Philosophy** · 10-10-2006, 02:08 AM

interesting for sure..

**camouflage** · 10-10-2006, 02:28 AM

Maybe the amount of alcohol addicted INTEL factory workers increased ..............

**thunderstruck!** · 10-10-2006, 03:22 AM

Originally Posted by phile

For clarity's sake, the E6300 and E6400 chips ARE conroe cores with half the 4MB L2 disabled. CPUZ is the only application that tags them as Allendale, and they are mistaken. The genuine Allendale cores have a native 2MB L2, and certain conroe features disabled, like VT. I wish to God that the CPUZ people would stop perpetuating this error.

-phil

And you know this how? What's your source?

**phile** · 10-10-2006, 03:36 AM

Plenty of evidence.

- E6300/E6400 have the full 4MB L2 cache on the die, with half disabled
- CPUZ is the only application that tags the 2MB conroes as Allendale
- The E4300 is a genuine Allendale core CPU with a native 2MB L2 cache, and is not yet released

And, then there's this:

"You'll find plenty of sources that will tell you the code name for these 2MB Core 2 Duo processors is "Allendale," but Intel says otherwise. These CPUs are still code-named "Conroe," which makes sense since they're the same physical chips with half of their L2 cache disabled."

http://techreport.com/reviews/2006q3...f/index.x?pg=1

-phil

**TaPaKaH** · 10-10-2006, 04:44 AM

As asked - CacheMem done

http://img244.imageshack.us/img244/2562/untitledhd0.png

It clearly says - 4Mb L2

**cpuz** · 10-10-2006, 05:39 AM

TaPaKah you made it well but all of this is fake.

This is your validation :
http://valid.x86-secret.com/show_oc?id=128596

This is a fake because the code name Allendale only results from the L2 size detection (namely 2MB). If there were really 4MB, the validation would have displayed Conroe (Phile : this may be wrong, you're right, but in this case this helped finding out the fake

)

Also we found out that you already tried your cheat here :
http://valid.x86-secret.com/show_oc?id=122548
(the submission IP is the same as yours)

Some other fakes attempts by TaPaKah (these were well detected as fakes)
http://valid.x86-secret.com/show_oc?id=120207
http://valid.x86-secret.com/show_oc?id=120541
http://valid.x86-secret.com/show_oc?id=120546

**TaPaKaH** · 10-10-2006, 05:44 AM

Also we found out that you already tried your cheat here :
http://valid.x86-secret.com/show_oc?id=122548
(the submission IP is the same as yours)

That`s not mine !!!

http://valid.x86-secret.com/show_oc?id=120541
http://valid.x86-secret.com/show_oc?id=120546

These aren`t properly validated `cuz of my OS (already said that in the previous topics ) !!!

**mackanz** · 10-10-2006, 05:44 AM

Ouch!

120207 is a nice one =P
Nice to have a cpu where you can switch between 2 and 4mb of cache

**phile** · 10-10-2006, 05:49 AM

What are the chances a sole E6400 left manufacturing without having had 2MB of the L2 disabled? Until I know different, I must believe that such an error would have happened with an entire batch of chips, and not just a single one.

Anyone have detailed knowledge of the manufacturing process, especially the step where half the L2 is disabled? Is this done manually, by a technician? Is this step automated?

-phil

**mrlobber** · 10-10-2006, 06:25 AM

Originally Posted by cpuz

TaPaKah you made it well but all of this is fake.

This is your validation :
http://valid.x86-secret.com/show_oc?id=128596

This is a fake because the code name Allendale only results from the L2 size detection (namely 2MB). If there were really 4MB, the validation would have displayed Conroe (Phile : this may be wrong, you're right, but in this case this helped finding out the fake

)

Also we found out that you already tried your cheat here :
http://valid.x86-secret.com/show_oc?id=122548
(the submission IP is the same as yours)

Some other fakes attempts by TaPaKah (these were well detected as fakes)
http://valid.x86-secret.com/show_oc?id=120207
http://valid.x86-secret.com/show_oc?id=120541
http://valid.x86-secret.com/show_oc?id=120546

Ok, I don't get it then, what's your final verdict?

Does TaPaKaH have a 4Mb cache or a 2Mb cache CPU, in your opinion?

And what is he exactly faking then?

**Quentin** · 10-10-2006, 06:40 AM

Faker of the year

**The Stilt** · 10-10-2006, 06:45 AM

The evidence that Franck showed is water proof so the case is pretty much closed, don´t you all agree?

-> TaPaKaH

**jmax_oc** · 10-10-2006, 07:04 AM

He faked for a lot of things

I said it 1 month ago

**kiwi** · 10-10-2006, 08:34 AM

I sold him this cpu, so cpu itself is not fake

**erwinz** · 10-10-2006, 09:10 AM

Originally Posted by kiwi

I sold him this cpu, so cpu itself is not fake

I second this.. I think he is not a fake.. We have seen that Intel have an ES E6400 with 4mb cache.. why would this be any different..

He is just a lucky guy..

**cpulloverclock** · 10-10-2006, 09:28 AM

ok but in the doubt, its wrs are removed until he'll give us better proofs

**thunderstruck!** · 10-10-2006, 09:45 AM

Sorry to hear this. I still think TaPaKaH didn't fake any of these scores. He even ran cachemem and it showed 4mb. Tell him what he has to do to prove he's not a cheater and we'll see if it's true. Why would he make this thread??

**ANP !!!** · 10-10-2006, 09:46 AM

Ur one Lucky D*G.. LOL Sorry..

Good for u. Have fun !!!!

**cpuz** · 10-10-2006, 09:54 AM

If the CPU really was a E6400 with 4MB it would have reported something like this :
http://valid.x86-secret.com/show_oc?id=100737

This a E6400 ES, and these CPUs actually had a 4MB L2. And notice that the code name is reported as Conroe. This is logical, as cpuz relies on L2 size to report either Conroe or Allendale.

On this validation now :
http://valid.x86-secret.com/show_oc?id=128596
4MB L2 and Allendale. If the CPU was real, it would have said Conroe, and period. Of course the author of the validation could not have known that.

IMO cpuz was cheated and the L2 size was changed with a debugger.

Concerning the cachemem screen, it could have been made on a E6600/E6700 and pasted over a fake cpuz screen. It's not very complicated to do.

Now, if I'm wrong I'll be the 1st to apologize. But at that moment I'm sure the validation shows a fake.

Thread: L625A348 E6400 ... 4Mb L2 ???

Thread Tools

Search Thread

Rate This Thread

Display

Bookmarks

Bookmarks

Posting Permissions