View Full Version : New ExtremeEdition Conroe 3.3ghz 1333mhz Bus
http://www.xbitlabs.com/news/cpu/display/20060212234350.html
:eek: :eek: :eek: :eek:
hicookie
02-13-2006, 07:17 AM
Gee........:slobber: 3.33G,FSB 1333,4MB L2:slobber:
Tasselhof
02-13-2006, 07:33 AM
More interesting - what MoBo will be good for OC it ?
333Mhz...
Pressure
02-13-2006, 07:36 AM
TDP of 95W.
That's the same as the Pentium D 820 toaster. Food for thoughts...
But to be fair the Conroe is also clocked 533Mhz faster and has twice the available cache.
reatil box cooler with the conroe will be mach 2 GT
lol
Which RAM is able to run 333Mhz? Even the newest DDR2 modules only reach 200Mhz or am I wrong?
eva2000
02-13-2006, 07:55 AM
woah price looks pretty good too!
eat me! :slobber: :slobber: :slobber:
That looks on paper, to be an awsome chip. :D
Donnie27
02-13-2006, 08:58 AM
Which RAM is able to run 333Mhz? Even the newest DDR2 modules only reach 200Mhz or am I wrong?
Yup, wrong. These use two channels of PC-667 DDR2. RAM will out do the Processor's bandwidth even more when DDR2 moves to 800MHz making it fast even to cover the processor even if it was 1600MHz FSB.
agenda2005
02-13-2006, 09:53 AM
Can anyone guess what the defailt Super pi 1Mb will be for this killer? Yonah at 2700MHz and 920MHz bus made 23sec. You will need A64 FX at 3600MHz to acheive that Super pi time. AMD better watch out.
ingentingmendeg
02-13-2006, 09:58 AM
will this cpu be compatible with a P5WD2 Premium?
^don.k's^
02-13-2006, 10:15 AM
will this cpu be compatible with a P5WD2 Premium?
Don't think so, i even doubt that i975X will be...
BTW, that chip would kick some major asses. @4ghz it sure beats everything ever seen...:banana:
rotjunior
02-13-2006, 10:16 AM
will this finally be an extreme chip worthy of the extreme price?
ingentingmendeg
02-13-2006, 10:19 AM
will this finally be an extreme chip worthy of the extreme price?
they all are worth the price.
freecableguy
02-13-2006, 10:30 AM
Conroe XE is high on my list of next main rig processor. This under my cooling will be amazing especially if good SLI support is included with a solid DFI overclocking MB.
d@rkn1ko
02-13-2006, 10:30 AM
impressive !!!!!
ingentingmendeg
02-13-2006, 10:31 AM
Conroe XE is high on my list of next main rig processor. This under my cooling will be amazing especially if good SLI support is included with a solid DFI overclocking MB.
i dont think ima get one. im happy with my 3.73EE ES :)
idiotec
02-13-2006, 10:48 AM
I kept hoping they were going to surprise us and have 1333FSB standard, but oh well, it is nothing some OC'ing can't fix anyways.
Does anyone know if it (all, not just the XE) is expected to have unlocked downward multi's?
caater
02-13-2006, 10:52 AM
TDP of 95W.
That's the same as the Pentium D 820 toaster. Food for thoughts...
But to be fair the Conroe is also clocked 533Mhz faster and has twice the available cache.
i believe the TDP for fx-62 (2.8ghz) was 120-130W.. fx-60 is 110w, fx-57 104w and 2.2-2.4 x2 cpus 110w.
just pointing out :)
Kjaks
02-13-2006, 10:57 AM
TDP of 95W.
That's the same as the Pentium D 820 toaster. Food for thoughts...
But to be fair the Conroe is also clocked 533Mhz faster and has twice the available cache.
AND it's pipeline is shortend by 16 stages ;)
Can anyone guess what the defailt Super pi 1Mb will be for this killer? Yonah at 2700MHz and 920MHz bus made 23sec. You will need A64 FX at 3600MHz to acheive that Super pi time. AMD better watch out.
M2 drops this year...don't worry, :p
This looks like it is one good processor though...very good stuff...and the price isn't bad at all for performance like that, !
the basics CPU(not EE) will have HT? i think than extreme edition have HT and the other CPU´s no.
n00b 0f l337
02-13-2006, 01:24 PM
Intel would do good to include HT on dual cores. :(
wittekakker
02-13-2006, 02:40 PM
I wonder WHEN this is going to happen.
M2 drops this year...don't worry, :p
!
I heard it would give only like 10-20% boost at most due to ddr2. That's it
DeltZ
02-13-2006, 03:02 PM
Intel would do good to include HT on dual cores. :(
i thought HT was rendered pointless with few stage pipelines? I thought it just made up for various inefficiencies?
Nasgul
02-13-2006, 03:46 PM
It's been said since November of 2005 that Conroe would have these features:
POST IT. (http://www.xtremesystems.org/forums/showpost.php?p=1195469&postcount=9)
I can say that I ain't surprised.
SoulGG
02-13-2006, 03:54 PM
Definatly shocked at most of the prices...Won't really know how they perform until they've been put to the test. I think i'll just wait for conroe...Yonahs cool too but this one looks like:toast:
agenda2005
02-13-2006, 09:29 PM
It's been said since November of 2005 that Conroe would have these features:
POST IT. (http://www.xtremesystems.org/forums/showpost.php?p=1195469&postcount=9)
I can say that I ain't surprised.
Do you think Conroe EE will have HT? If that's true, then that will be 4 virtual core on that bad boy. I can't wait man.:slapass: :woot:
M2 drops this year...don't worry, :p
This looks like it is one good processor though...very good stuff...and the price isn't bad at all for performance like that, !
M2 or What??? With only 10-15 percent improvement, that will not cut it. Someone who already saw AM2 said S939 owner should keep their system for now as the performance is not worth the wait.
Intel knew what they were doing when they said increasing clock speed alone will no longer cut it for the competitor. AMD is set to release 3GHz Opteron and no one is taking the news with excitement. http://www.xtremesystems.org/forums/showthread.php?t=89262
Conroe and Woodcrest are going to obliterate any lead that Opteron and X2 have now, just wait and see.
Preliminary results coming from Yonah is also a good indication of what is to come. http://www.xtremesystems.org/forums/showthread.php?t=88226
In addition to that Conroe comes with
* 14 stage pipeline
* Improve FPU
* 4 wide issue core
* 4MB inclusive L2 cache (not the crappy exclusive L2 that AMD uses)
* Deeper and wider buffers
* Improved branch prediction
newssss
http://www.digitimes.com/news/a20060213PR218.html
Pressure
02-14-2006, 03:35 AM
newssss
http://www.digitimes.com/news/a20060213PR218.html
I can't see how this relates to the Conroe series of processors at all?
Byron
02-14-2006, 06:10 AM
they all are worth the price.
http://img252.echo.cx/img252/4391/msmiley18xh.gif http://img252.echo.cx/img252/4391/msmiley18xh.gif http://img252.echo.cx/img252/4391/msmiley18xh.gif
liked the way you said it http://img293.echo.cx/img293/4732/chatsmiliescom8921zb.png
While I really don't feel like getting pulled into the beaten-to-death AMD vs. Intel debate...
M2 (or AM2, if you will) will have the said improvements, along with the other improvements that are being added (DDR2 isn't the only thing). Note also that 65nm will be deployed relatively soon after M2's release as a platform, at the time that most people start adopting it, which will lead to increased thermal / energy consumption / performance properties, just as it has for Intel's lineup.
Regardless of whose CPUs are actually better, microprocessor developement in general is being spurred on by the rivalry and it's keeping costs down for the consumers, so I'm happy either way.
Salvador
02-14-2006, 10:26 AM
While I really don't feel like getting pulled into the beaten-to-death AMD vs. Intel debate...
M2 (or AM2, if you will) will have the said improvements, along with the other improvements that are being added (DDR2 isn't the only thing). Note also that 65nm will be deployed relatively soon after M2's release as a platform, at the time that most people start adopting it, which will lead to increased thermal / energy consumption / performance properties, just as it has for Intel's lineup.
Regardless of whose CPUs are actually better, microprocessor developement in general is being spurred on by the rivalry and it's keeping costs down for the consumers, so I'm happy either way.
AM2 won't do much better than 939 performancewise. 10-15% at the most. AMD's 0.65nm process won't be 'out in shops' at earliest during 1H/Q1 2007, only a half year+ before Intel is slated to release it's 0.45nm process(Q4 2007) for mass production.
And 'bout AMD's 2006 desktop-plans .. -> Booooooring. A new X2-5200 and a FX-62 @ DDR2 ? :clap:
AM2 won't do much better than 939 performancewise. 10-15% at the most. AMD's 0.65nm process won't be 'out in shops' at earliest during 1H/Q1 2007, only a half year+ before Intel is slated to release it's 0.45nm process(Q4 2007) for mass production.
And 'bout AMD's 2006 desktop-plans .. -> Booooooring. A new X2-5200 and a FX-62 @ DDR2 ? :clap:
They'll be out Q4 this year or Q1 next, yeah. Also:
http://www.amdboard.com/65nm_120605.html
Keep that in mind while considering their shift to 65nm. The 10-15% performance boost figure is at a per-clock comparison to 939 systems, but M2 will also bring clock speed and other small improvements to AMD's lineup.
Either way...M2 + shift to 65nm = Intel won't be dominating anyone, regardless of what you think. Conroe and Woodcrest will be good, no doubt, Intel is putting lots of effort into their developement and I doubt people will be let down. Just don't overestimate Intel or underestimate AMD. You make it sound like Intel will be rolling out industry-changing products while AMD is going to be releasing a new Sempron.
I would also like to point out that AMD doesn't have to re-design everything to escape design limitations like Intel has, or slap new names on their products. I find quite a bit of irony that it is now Intel, and not AMD, that is pointing out that clock speed isn't the true measure of performance, when not too long ago it was the other way around (and in some cases, still is).
Anyway...this isn't really the place for this discussion I don't think...I'm sorry I ever posted X.X
coldpower27
02-14-2006, 12:36 PM
Let's see what thing will shape like in Q3 2006:
Intel Q3 2006
Core EE Exxxx 3.00GHZ or 3.33GHZ/1333FSB/4MB 999US TDP 95W
Pentium EE 965 3.73GHZ/1066FSB/2x2MB 999US TDP 130W
Core E6700 2.66GHZ/1066FSB/4MB 530US TDP 65W
Core E6600 2.40GHZ/1066FSB/4MB 316US TDP 65W
Pentium D 960 3.6GHZ/800FSB/2x2MB 316US TDP 130W
Core E6400 2.13GHZ/1066FSB/2MB 241US
Pentium D 950 3.4GHZ/800FSB/2x2MB 241US TDP 130W
Core E6300 1.83GHZ/1066FSB/2MB 209US
Pentium D 940 3.2GHZ/800FSB/2x2MB 209US TDP 130W
Pentium D 920 2.8GHZ/800FSB/2x2MB 178US TDP 95W
AMD Q3 2006 (Personal Guess, very few leaks)
Athlon FX 62 2.8GHZ/2x1MB AM2 1031US
Athlon FX 60 2.6GHZ/2x1MB AM2 827US
Athlon 64x2 5000+ 2.6GHZ/2x512KB AM2
Athlon 64x2 4800+ 2.4GHZ/2x1MB AM2
Athlon 64x2 4600+ 2.4GHZ/2x512KB AM2
Athlon 64x2 4200+ 2.2GHZ/2x512KB AM2
Athlon 64x2 3800+ 2.0GHZ/2x512KB AM2
n00b 0f l337
02-14-2006, 12:50 PM
960 looking amazing.... May we see 6ghz+ dual core? ;)
^don.k's^
02-14-2006, 12:59 PM
960 looking amazing.... May we see 6ghz+ dual core? ;)
That wouldn't be any new, already there...:D
mr_knowitall15
02-14-2006, 01:04 PM
yep. Fuggers been there, done that. I AM kind of surprised that we havent seen any 8Ghz yet from Fugger or Team Japan with a cedar mill. hell, the japanese have hit 7.5xxGhz with a 670, youd think with an ES 661 or 671(if they exist yet) this can be beaten.
How do Conroe Mhz compare to Yonah? Yonah @ 3.3Ghz would be a total beast :D
metro.cl
02-14-2006, 02:24 PM
remember that intel is coldbug free that is awsome and a must have for us im sick of my opty with coldbug (when playing with dryice) -30c is ok but colder is no go
idiotec
02-14-2006, 03:59 PM
How do Conroe Mhz compare to Yonah? Yonah @ 3.3Ghz would be a total beast
I am wondering the same thing. Yonah has 12 stage pipes where as Conroe has 14 stage pipes, correct? So that gives some advantage to Yonah. But, Yonah has 166Mhz FSB compared to Conroe at 266Mhz (non EE). So yeah, I am curious how the shorter pipes/lower FSB compares to slightly longer pipes/higher FSB.
vapb400
02-14-2006, 04:53 PM
remember that intel is coldbug free that is awsome and a must have for us im sick of my opty with coldbug (when playing with dryice) -30c is ok but colder is no go
the current intel chips are coldbug free. Hopefully the future chips wont be, but we just dont know.
Vapor
02-14-2006, 05:15 PM
I am wondering the same thing. Yonah has 12 stage pipes where as Conroe has 14 stage pipes, correct? So that gives some advantage to Yonah. But, Yonah has 166Mhz FSB compared to Conroe at 266Mhz (non EE). So yeah, I am curious how the shorter pipes/lower FSB compares to slightly longer pipes/higher FSB.Don't forget about notably higher MHz and different architecture in general. :fact:
idiotec
02-14-2006, 05:36 PM
Don't forget about notably higher MHz and different architecture in general. :fact:
Well yeah, there are the higher clock speeds, but the question is how they will compare clock for clock. Since we have some benchmarks for how Yonah performs, understanding how they will compare clock for clock, would give a good idea of how Conroe will perform, given the higher clocks. And the fact that they have some architectural differences is the exact reason why we can't necessarily assume there will be a 1:1 clock ratio between the chips. Whether that ratio will be higher or lower in favor of Conroe is the question.
Sentential
02-14-2006, 05:41 PM
Unless something miraclous happened I seriously doubt we will see 3.3ghz XEs this year. I hope Im wrong tho :rocker:
AkXb70
02-14-2006, 07:34 PM
I am wondering the same thing. Yonah has 12 stage pipes where as Conroe has 14 stage pipes, correct? So that gives some advantage to Yonah. But, Yonah has 166Mhz FSB compared to Conroe at 266Mhz (non EE). So yeah, I am curious how the shorter pipes/lower FSB compares to slightly longer pipes/higher FSB.
actually, dothan and yonah have a 14 stage pipeline (whether or not its the same pipeline in conroe, i cant say)
kl0012
02-14-2006, 08:09 PM
Whether that ratio will be higher or lower in favor of Conroe is the question.
Actualy Merom should be much faster than Yonah. Yonah was developed with the support to the low energy consumption and it has many weak points (for example decoder and FPU), which suppose to be fixed in Merom. On some information Merom will have the following improvements:
*Improved decoder. Theinq.org hints to Macro-op fusion which meen that if you have a multiply followed by an add, Macro- op fusion can turn that into a multiply and Accumulate.
*4-issue core instead of 3-issue in Yonah.
*Better FPU which can handle two 64-bit fpu-operations per cycle (128-bit SSE2/SEE3/SSE4).
AkXb70
02-14-2006, 08:38 PM
Dothan does not have a 14 stage pipeline!!! It's 10!!!
it is 14...trust me. check EOCF...theres a thread or two with the info in it
also, it cant be 10 because as per confirmed by intel...its longer than tualatin, shorter than northwood...this means anywhere between 12 and 19 stages (as tualatin is 11 and northwood is 20)
EDIT: guess that post didnt last....
Vapor
02-14-2006, 09:17 PM
No, it didn't last....I mixed up data, L2 latency is 10cycles in Dothan and 14 in Yonah.
Pardon me. :(
idiotec
02-14-2006, 09:23 PM
actually, dothan and yonah have a 14 stage pipeline (whether or not its the same pipeline in conroe, i cant say)
Oh :hitself: I thought for sure it was 12... :p:
d@rkn1ko
02-14-2006, 09:48 PM
it's 12 for PM and 14 for yonah I think :toast:
Well, anyway if it performs similar or even a bit worse at the same speed then @ 3.3Ghz it will totally own :D
wittekakker
02-15-2006, 05:35 AM
I wonder WHEN this is going to happen.
Intel suspects first chips near the end of this year.
It's a 14 stages pipeline and it can do 4 instructions/second, where AMD and older Intel chips can only 3i/s. This is something that has been used for quit a while, but now Intel has the balls to take it to 4i/s. This provides a higher IPC, but I thought there were some problems with that too (complex chips or something like that).
Now not only L2 cache is shared, there is aslo a bridge between the L1 cache from each core. An intellectual part of the chip will then say how the L1 cache will be used.
It will perform 30% faster then current P-M.
This is news from few months ago, IDF Fall 2005.
Also they have added SSE4, wich has 'Significant Video Enhancements' in it. This is an display interface, the improved version of VGA, DVI, HDMI.
More info we will get on next IDF on 7-9 March.
http://www.tweakers.net/ext/i.dsp/1124828126.jpg
Kjaks
02-15-2006, 05:59 AM
I really wonder what cache latency Conroe will have.. 10 cycles? 14 cycles? What do you suggest? Are there any facts on the subject?
Edit: ah, I didn't see that post about the cache latency up there.
Sumanji
02-15-2006, 02:17 PM
I heard it would give only like 10-20% boost at most due to ddr2. That's it
20% is still quite a lot though imo...
But Conroe EE will destroy all with that demonic clockspeed :eek:
Suman
AkXb70
02-15-2006, 03:54 PM
it's 12 for PM and 14 for yonah I think :toast:
yonah is the exact same pipeline as dothan...so 14 for both ;)
Pressure
02-16-2006, 06:30 AM
yonah is the exact same pipeline as dothan...so 14 for both ;)
Intel have not disclosed the exact number of interger pipelines but they have said that it is longer than the Pentium III but shorter than Conroe/Merom.
The Pentium is a 10-stage integer pipeline and the Conroe/Merom uses a 14-stage integer pipeline. If it uses a 12-stage integer pipeline it matches that of the AMD Athlon64.
AkXb70
02-16-2006, 06:21 PM
Intel have not disclosed the exact number of interger pipelines but they have said that it is longer than the Pentium III but shorter than Conroe/Merom.
The Pentium is a 10-stage integer pipeline and the Conroe/Merom uses a 14-stage integer pipeline. If it uses a 12-stage integer pipeline it matches that of the AMD Athlon64.
P6 (coppermine/tualatin) is 11 stages at last count (6 in order + 3 out of order + 2 in order)
and dothan / yonah is 14, as i said, theres info in the last few pgs of the conroe thread on EOCF about this.
Brett Vankirk will also have information
vBulletin® v3.7.0, Copyright ©2000-2008, Jelsoft Enterprises Ltd.