CONROE speeds leaked?

[apache_lives]

http://translate.google.com/translate?hl=en&sl=fr&u=http://www.x86-secret.com/?option=newsd&nid=909&prev=/search?q=x86+secret+conroe&hl=en&lr=&c2coff=1&rls=GGLD,GGLD:2005-27,GGLD:en

Translated thanks to google, X86 Secret (if true?) is saying the next gen Conroe will initially be available at 2.93ghz with FSB 1066 (266) and have an optional 4mb L2 per core.

Has anyone else got any more news about it? and If this is a fake or un-true - tell us.

 

[endyen]

IDF is next week, so we will find out then.
Since Conroe is P-M based I dont see it, but we will see.

 

[apache_lives]

Yeah its aparently Pentium M based and will be at 65nm and will have pentium 4 / netburst technologies included so id say its correct since they currently do overclock Pentium M's beyond 3.4ghz.

IDF will reveal conroe but what exactly? Hopefully it wont just be a sample or picture or something thats meaningless - hopefully specs and technology.

 

[endyen]

We are talking dual core here. I had heard bothe that it will, and that it wont have HT. I'm pretty sure it will have SSE2 and SSE3, though they will be less effective, due to slower clocks.
I would be very supprised if the started at ckock speeds over about 2.5ghz, or used an fsb higher than 133.
The main idea for intel is to get temps down. Then they will begin to ramp speeds again.
At least at 2.5, they will be able to compete with Amd, and at 65 nanos, production costs should be low, if yields are good.

 

[apache_lives]

i believe it will start at 2.93ghz and FSB 1066 (266) - WHY? because they push current pentium M's (90nm) to 3.4ghz with FSB 1072 and they are the fastest things available - http://www.xtremesystems.org/forums/showthread.php?t=36835

They will not come out at 2.5ghz nor with a FSB of 133 (533) - THIS IS NOT A MOBILE CHIP - the current pentium m's are clocked slower and without SSE2 and 3 TO SAVE POWER - WHY WOULD A DESKTOP CPU NOT HAVE SSE2 AND 3? A PENTIUM M AT 2.5GHZ TODAY COMPARES TO A 4000+ A64 - WHY WOULD THEY RELEASE A CHIP A YEAR LATER TO COMPARE TO TODAYS PERFORMANCE? AND LASTLY IT WILL BE AT 65NM AND WILL ALLOW EVEN MORE CLOCK SPEED POTENTIAL.

As for HT (or intels version call CSI) if it was going to be with conroe WHY WOULD CONROE STILL FIT INTO THE SOCKET 775? LATER ON MAYBE BUT NOT YET - Intel said by 2009 they will have it on just about all processors.

 

[endyen]

It's really too bad that 166 quad pumped is the number of the beast. Seriously, mad OCers, with overvolting, extreme cooling, and more dead chips than live, have barely been able to get the dothans past 200mhz fsb, and you think intel will intro thier dual core version at 266? Seems a little far fetched, but hey, you can dream.
I've always found it was safer to under expect from intel. Then you are not usually too dissapointed.
Dont count too much on the die shrink, the last one only bought intel a lot of headaches.
Things like FDSOI will do a lot more than the die shrink, for giving speed. Even at that, and with a nice core like dothan, expect a few major revisions, before it reaches it's potential.
It would be great to see intel moving ahead, so I will cross my fingers.

BTW, intel's HT is hyper threading. It probably wouldn't do much for conroe, if it has the type of IPC the Ms have.

<P ID="edit"><FONT SIZE=-1><EM>Edited by endyen on 08/20/05 03:19 AM.</EM></FONT></P>

 

[apache_lives]

FSB doesnt limit overclocking - the core clock limits the overall clock speed, and as a well known fact - the pentium m's are an unlocked processor to allow SpeedStep power saving through lower clock speeds and also helps overclocking by upping the FSB and lowering the clock speed to reasonable numbers.

I overclocked a P1 233mmx from 233/66 to a nice 266/133 (2x multiplier, double the FSB) - thanks to a nice board supporting FSB133 (Gigabyte GA-5AA, ALi Aladdin V) and a decent batch of processors - FSB is not a limiting factor processor wise.

HT - i thought you meant hyper transport (cause there was word about intels answer to hypertransport - CSI/Common System Interface), anyhow - why would they want HT (hyperthreading) when dual core is available? HT was an evolutionary step in the market for the future of multi-threaded apps - a cheap step.

Pentium M's are (if you havnt noticed) optimised for MOBILE use, a little tweaking and a few netburst technologies added here and there will make a dam quick chip.

 

[endyen]

FSB doesnt limit overclocking

It has by my experience. The faster the fsb, the more heat it conducts to the chip. You probably could have gotten better perf, and OC, just by using 4X and an fsb of 75.
(JP2 1-2, JP1 2-3, JP14 2-3 and JP16 open, according to my manual) I should also say that I liked the bus mastering, it gave the old quantums a little extra kick.

 

[Unolocogringo]

No go the faster FSB is the faster overclock.
Had an old P5A-B that would run 120fsb with no problems. A Pentium 1 with a 120 fsb was amazing for its time.

I aint signing nothing!!!

 

[apache_lives]

ok that just told me alot:
1: the Pentium MMX's and all socket7 cpu's have available multipliers of 1.5x(->3.5x), 2x, 2.5x and 3x, if you set it as 4x it would skip over to one on the list here - 4x and above aswell as below 1.5x are not accepted.

2: the 350nm process used to make the Pentium MMX's allowed a max overclock of 292mhz (and my sample does it) and ABSOLUTLY NO FURTHER - even with an overvolt from 2.8v to 3.5v on an Copper Based AthlonXP ("2000+ max") cooler it wont reach the desktop - even with 100x3/300mhz/3.5v with four diffrent MMX cpu's it wont happen.

3: IDE Bus mastering - it would kick in for any IDE device not just the quantums.

As for heat - maybe, more likly more heat for the north bridge, and when the FSB is faster it does allow more data to be transported so yet - it would go up a little but were talking about a cpu designed for it not just a copy of a laptop chip.


We are talking about the posible specs and making sence of the details leaked about the future P4 sucessor CONROE - NOT HOW GAY YOUR COMPUTER SKILLS - PLEASE - LEAVE THIS FORUM IF YOU HAVE NOTHING BETTER TO DO

 

[apache_lives]

glad you agree - 20mhz on the FSB does more then 20mhz more clock speed, but 20mhz on both is even better.

Good example of FSB - Pentium 4 northwood the C class with FSB800 vs the older B and A class - the C's kicked some sweet ass for the same (or less) clock speed and as for heat - not much diffrence.

 

[mozzartusm]

We are talking about the posible specs and making sence of the details leaked about the future P4 sucessor CONROE - NOT HOW GAY YOUR COMPUTER SKILLS - PLEASE - LEAVE THIS FORUM IF YOU HAVE NOTHING BETTER TO DO

Are you serious? In the words of Ned Flanders "KNOW THY PLACE" This is one of our most knowledgeable senior members you said this to. Your time here will be very short if you continue to make comments such as this. This isnt a threat, its a promise!


ASUS P5WD2 Premium
Intel 3.73 EE @ 5.6Ghz
XMS2 DDR2 @ 1180Mhz

<A HREF="http://valid.x86-secret.com/records.php?PHPSESSID=792e8f49d5d9b8a4d1ad6f40ca029756" target="_new">#2 CPUZ</A>
SuperPI 25secs

 

[endyen]

PLEASE - LEAVE THIS FORUM IF YOU HAVE NOTHING BETTER TO DO

Just trying to be pleasant to an idiot inside boy. Dont like it piss off.
We will find out on the 23rd who the fking dreamer is.

 

[mozzartusm]

Some people just dont get it. Hes in critical condition at this point.

ASUS P5WD2 Premium
Intel 3.73 EE @ 5.6Ghz
XMS2 DDR2 @ 1180Mhz

<A HREF="http://valid.x86-secret.com/records.php?PHPSESSID=792e8f49d5d9b8a4d1ad6f40ca029756" target="_new">#2 CPUZ</A>
SuperPI 25secs

 

[endyen]

Thanks mozz. I dont know what happened, he's usually a bit of a zelot, but not usually a jerk.

 

[Ned_Flanders]

Haha, you quoted me :lol:

OnT: What Mozz said will be backed up by a lot of members, so watch it!

_____________________
<A HREF="http://snipurl.com/fxwr" target="_new">Welcome to the House of Horrors, welcome to the House of a 1000 Corpses</A>