'Nehalem' at 2.9GHz and X58

[taltamir]

Originally posted by: aigomorla

Originally posted by: taltamir
what worries me is... why is it only 2.9ghz on an engineering sample?

cuz the stock speed is 2.9?

They usually dont overclock until after NDA comes off.

the original phenom was shown at what...3.4 ghz? and then it came out as 2.2ghz and slowly inched up to 2.5ghz

intel already has set 3.0ghz + as their standard for high end... a sample below it worries me in that it might have issues like phenom did...

Or maybe intel intends to start us with low speeds initially, now that they supposedly eliminated overclocking, it makes sense to start it with a very low multi and jack it up later.

 

[Idontcare]

Originally posted by: taltamir
the original phenom was shown at what...3.4 ghz? and then it came out as 2.2ghz and slowly inched up to 2.5ghz

It was publicly demo'ed at 3GHz IIRC.

The good news is that just about everyone has been removed of their duties who was involved in the farcical PR blitz about Barcelona being the second coming and Clovertown's undoing.

The bad news is I don't think we ever really got the scoop on why K10 clockspeed came in so equisitely low and never markedly increased.

Surely this was a surprise to even AMD as they would not have demo'ed a 3GHz K10 and roadmapped 2.8GHz FX models were they not convinced their K10 design and 65nm process technology were capable of generating the SKU's.

So how/where did the wheels fall off the wagon? Will the public ever know?

 

[bharatwaja]

Originally posted by: taltamir

the original phenom was shown at what...3.4 ghz? and then it came out as 2.2ghz and slowly inched up to 2.5ghz

intel already has set 3.0ghz + as their standard for high end... a sample below it worries me in that it might have issues like phenom did...

Or maybe intel intends to start us with low speeds initially, now that they supposedly eliminated overclocking, it makes sense to start it with a very low multi and jack it up later.

I think intel wants to do a reverse of what AMD did.... like make us think that way and then BAM! AMD gets hit by a truckload of 4GHz Neha's to start with........

but hey, who cares about clock speed? its intel's own mess.... The architecture could be so advanced that the chip makes up for by leaps and bounds, clock for clock i mean....

And from the word out there, it already seems so....

 

[Nemesis 1]

Originally posted by: taltamir
what worries me is... why is it only 2.9ghz on an engineering sample?

Don't worry about it. I am sure intels Nehalem will work great with already proven Intel 45nm.

You know whats going to be really strange about Nehalem . Its the DDR3x3 on the extreme 1366,
Ya know how everyone who is overclocking is pushing high FSB . ON the Extreme setup with unlocked up cpu. The good performance is going to come from lowering the DDR3 mgz. In favor of lower timings. SO the Extreme unlocked Nehalem is going to have some very interesting advantages over the performance models. Inso far as having the ability to O/C and still maintain tight low timing. If Nehalem can come close to using this kind of bandwidth at DDR3x3 800 tight timings I would be extremely surprized . So low tight timings on the extremes will be the performance setup.

 

[aigomorla]

dude Neha will rape.

I'll show you guys when i have the chance.

Im just fearing for AMD. Becuase its gonna be that big of a gap.

OMG i found a case i want. Thank you.

Now to find it.

 

[dmens]

Originally posted by: aigomorla
dude Neha will rape.

I'll show you guys when i have the chance.

you should pm me what frequency you are at and what stepping you are using. so i know if it is major pwnage or just regular pwnage.

 

[Phynaz]

Originally posted by: dmens

Originally posted by: aigomorla
dude Neha will rape.

I'll show you guys when i have the chance.

you should pm me what frequency you are at and what stepping you are using. so i know if it is major pwnage or just regular pwnage.

This implies a large perfomance boost between steppings.

Generaly not something I would expect.

 

[myocardia]

Originally posted by: Idontcare
It was publicly demo'ed at 3GHz IIRC.

The good news is that just about everyone has been removed of their duties who was involved in the farcical PR blitz about Barcelona being the second coming and Clovertown's undoing.

The bad news is I don't think we ever really got the scoop on why K10 clockspeed came in so equisitely low and never markedly increased.

Surely this was a surprise to even AMD as they would not have demo'ed a 3GHz K10 and roadmapped 2.8GHz FX models were they not convinced their K10 design and 65nm process technology were capable of generating the SKU's.

So how/where did the wheels fall off the wagon? Will the public ever know?

Are you talking about that video on youtube, where they were demonstrating the new Spyder platform, and it's related software? If so, that wasn't a 3 Ghz chip at all; the video showed the AMD tech overclocking that 2.2 or 2.3 Ghz ES chip. IIRC, according to AMD, that video was only to showcase their new overclocking software that they were so proud of, that could only be used with the Spyder platform.

edit: And personally, I think the video was just some more of Hector's BS.

 

[Viditor]

Originally posted by: myocardia

Originally posted by: Idontcare
It was publicly demo'ed at 3GHz IIRC.

The good news is that just about everyone has been removed of their duties who was involved in the farcical PR blitz about Barcelona being the second coming and Clovertown's undoing.

The bad news is I don't think we ever really got the scoop on why K10 clockspeed came in so equisitely low and never markedly increased.

Surely this was a surprise to even AMD as they would not have demo'ed a 3GHz K10 and roadmapped 2.8GHz FX models were they not convinced their K10 design and 65nm process technology were capable of generating the SKU's.

So how/where did the wheels fall off the wagon? Will the public ever know?

Are you talking about that video on youtube, where they were demonstrating the new Spyder platform, and it's related software? If so, that wasn't a 3 Ghz chip at all; the video showed the AMD tech overclocking that 2.2 or 2.3 Ghz ES chip. IIRC, according to AMD, that video was only to showcase their new overclocking software that they were so proud of, that could only be used with the Spyder platform.

edit: And personally, I think the video was just some more of Hector's BS.

No, they actually demoed a working 3 GHz Phenom for preview at one of the big shows (I can't remember which off the top of my head).
I have to agree with the consensus that the 3GHz demo was a huge mistake...as was most of the marketing going on at that time.

 

[BrownTown]

Originally posted by: taltamir
what worries me is... why is it only 2.9ghz on an engineering sample?

you've got this backwards, ES chips almost always come out at LOWER clock speeds then the highest binned retail parts. The first ES Conroe chips came out at 2.0Ghz. Also, this is a proven 45nm process, there aren't changing that any, so the final speed shouldn't change much either way.

 

[Nemesis 1]

IF Intel nehalem is = to Merom at launch in so far as function. Nehalems pipes have been lengthened to 16 I believe. That in theory should allow for higher clocks.

Also if you compare 65nm to 45nm you will see power and performance scale about = showing the 45nm improvements. Intel said Nehalem would fully take advantage its 45nm tech. RIght now at this time I am inclined to except intels word. So if what were seeing in performance increases = the same in O/C as were seeing between merom/Penryn . Iam looking forward to some very high clocking unclocked Nehalems.

Nehalems are exciting. We all pretty much believe already their going to kick ass. compute wise. But the Most exciting thing to me is finally well here alot more about sandy. I got to know what those israelites are up to . After seeing how they took and old core and made intel see the light. I cann't wait! This well be something special.

 

[Markfw]

Yup. I now have seen neha... It kicks a yorkie's ass.

Now lets just wait until the show begins......

 

[Idontcare]

Originally posted by: BrownTown

Originally posted by: taltamir
what worries me is... why is it only 2.9ghz on an engineering sample?

you've got this backwards, ES chips almost always come out at LOWER clock speeds then the highest binned retail parts. The first ES Conroe chips came out at 2.0Ghz. Also, this is a proven 45nm process, there aren't changing that any, so the final speed shouldn't change much either way.

Yeah, isn't the primary function of an ES chip to be a debug chip for the rest of the hardware that the chip is going to be plugging into when everything goes retail? (mobos, ram, chipsets, etc)

I was under the impression that ES chips can actually contain pretty fatal speedpath issues that are worked around with BIOS's and on-die fuses just so long as the functionality is maintained.

ES chips are definitely NOT retail nor OEM for some good reason.

 

[Cookie Monster]

When are we expecting to see the LGA1160 for us more budget oriented folks? Next year or so?

 

[Drsignguy]

Originally posted by: Markfw900
Yup. I now have seen neha... It kicks a yorkie's ass.

Now lets just wait until the show begins......

, yeah, I still love my yorkie, even though it got it's butt kicked.

 

[aigomorla]

Originally posted by: Nemesis 1

Also if you compare 65nm to 45nm you will see power and performance scale about = showing the 45nm improvements. Intel said Nehalem would fully take advantage its 45nm tech. RIght now at this time I am inclined to except intels word.

nemisis, you'll be very pleasently happy when you see it unleashed.

trust me.

 

[Idontcare]

Originally posted by: Cookie Monster
When are we expecting to see the LGA1160 for us more budget oriented folks? Next year or so?

Merely see it...or also to find it in stock and at a reasonably acceptable budget-oriented price?

My bets are on no one really seeing it at a reasonable price-point until 32nm shrink (westmere) rolls out of the fabs. (smaller die-size enabling a needed reduction on production cost over existing 45nm penryn chips at similiar pricepoints but higher gross margins for now)

 

[Cookie Monster]

Originally posted by: Idontcare
Merely see it...or also to find it in stock and at a reasonably acceptable budget-oriented price?

My bets are on no one really seeing it at a reasonable price-point until 32nm shrink (westmere) rolls out of the fabs. (smaller die-size enabling a needed reduction on production cost over existing 45nm penryn chips at similiar pricepoints but higher gross margins for now)

The latter of course . I was kind of afraid of this, but guess this means the LGA775s aren't really going anywhere even when nehalem ships. How many years you think before we see the cheaper nehalem derivatives trickle down to us mainstreamers?

 

[Idontcare]

Originally posted by: Cookie Monster
How many years you think before we see the cheaper nehalem derivatives trickle down to us mainstreamers?

Just one year (early 2010) from initial release of Bloomfield, but those 12-18 months will seem like forever as they always do feel as a consumer in this industry.

 

[aigomorla]

Okey guys.

I dont know if you missed this.

http://www.techpowerup.com/652...ensive_Evaluation.html

It has cinebench and Sysandra.

Now you guys will see why i am saying a budget line dell will spank an overclocked AMD.


No fanboyism here. Im looking at straight and raw numbers. Wanna argue with me on sysandra or cinebench, thats fine, but i take those 2 as good benchmark programs.

 

[Idontcare]

Originally posted by: aigomorla
Okey guys.

I dont know if you missed this.

http://www.techpowerup.com/652...ensive_Evaluation.html

It has cinebench and Sysandra.

Now you guys will see why i am saying a budget line dell will spank an overclocked AMD.


No fanboyism here. Im looking at straight and raw numbers. Wanna argue with me on sysandra or cinebench, thats fine, but i take those 2 as good benchmark programs.

OMG those numbers! daddy like!

 

[AmberClad]

Me? I like the projected price - $284 for the lowliest Bloomfield model. Course, that's the price per unit with 1000 unit lots. Not sure what we'll end up paying. The motherboards might not be so friendly on the wallet though, especially the models of the X58 that also include the nForce 200 chip. I guess they'll eventually do a "performance" grade chipset (P55?) for the rest of us. At least DDR3 has been on a steady price decline.

 

[Extelleron]

Originally posted by: AmberClad
Me? I like the projected price - $284 for the lowliest Bloomfield model. Course, that's the price per unit with 1000 unit lots. Not sure what we'll end up paying. The motherboards might not be so friendly on the wallet though, especially the models of the X58 that also include the nForce 200 chip. I guess they'll eventually do a "performance" grade chipset (P55?) for the rest of us. At least DDR3 has been on a steady price decline.

$284?

If that's true then that is an amazing price/performance right there. Especially if these things OC as well as Core 2 does now.

 

[Idontcare]

Originally posted by: AmberClad
Me? I like the projected price - $284 for the lowliest Bloomfield model.

That's just crazy talk right there. :shocked:

The very same kind of crazy talk that me loves to hear! :laugh:

 

[Gary Key]

Originally posted by: taltamir
what worries me is... why is it only 2.9ghz on an engineering sample?

2.66, 2.93, 3.2 at launch... which considering the new design is actually incredible from our viewpoint. Even more incredible is just how high they clock at this point on somewhat immature X58 platforms. Hopefully that headroom remains at launch. That is all I can say...