Will there be 1600FSB Core 2 Quads?

[PeteRoy]

I know the Core 2 Quad Extreme will have 1600Mhz FSB but will there be regular quads with this FSB before Nehelm?

 

[Idontcare]

Its a good question. I know one of the "features" of the X48 chipset from Intel is supposed to be support for 1600MHz FSB. Now when will X48 be released, and will its release be accompanied by 1600MHz FSB quads is a good question. I doesn't seem to get discusses much.

 

[GuitarDaddy]

I wish they would stop already with the FSB increases, it gives minimal gains and kills the overclockability. Think what the multi's would be, probably 5.5 to 7.5

 

[PCTC2]

The QX9770 is on the roadmap. It's a 1600MHz FSB Quadcore Extreme Edition.

 

[aigomorla]

isnt the QX9770 1600fsb?

doe pctc beat me to it...

 

[alfa147x]

Originally posted by: GuitarDaddy
I wish they would stop already with the FSB increases, it gives minimal gains and kills the overclockability. Think what the multi's would be, probably 5.5 to 7.5

Im new to overclocking but isn't Bus Speed different the Front Side Bus?

Like i said im new to this whole thing, please dont bite my head off if im wrong

 

[lopri]

P45 is on the roadmap as well. If X48 and P45 as well as G45 actually happens, that could be a sign of initial short supply of Nehalem. (be it intentional or not)

 

[tvdang7]

Originally posted by: GuitarDaddy
I wish they would stop already with the FSB increases, it gives minimal gains and kills the overclockability. Think what the multi's would be, probably 5.5 to 7.5

expensiver ram too......... to overclock.

amd doesnt do that. we can use ddr2 400 prob still till the end of next year lol.

 

[DerwenArtos12]

Originally posted by: aigomorla
isnt the QX9770 1600fsb?

doe pctc beat me to it...

a deer? a female deer?

Yup, the extremes are on the roadmaps and that only means that it'll trickle into the lower chips. Quite frankly though. With nehalem and the IMC front side bus will basically be extinct as it has been with AMD for some time. It's all about the hyper transport speeds and internal scaling. They can call teh FSB whatever they want once they put the memory controller on board but, the memory frequency and the chipset interconnect frequency are not going to be directly tied to it, similarly to memory speeds are now, everything is based on a top tier frequency with dividers on them. *shrug* oh well.

 

[lopri]

Originally posted by: DerwenArtos12
a deer? a female deer?

Yup, the extremes are on the roadmaps and that only means that it'll trickle into the lower chips. Quite frankly though. With nehalem and the IMC front side bus will basically be extinct as it has been with AMD for some time. It's all about the hyper transport speeds and internal scaling. They can call teh FSB whatever they want once they put the memory controller on board but, the memory frequency and the chipset interconnect frequency are not going to be directly tied to it, similarly to memory speeds are now, everything is based on a top tier frequency with dividers on them. *shrug* oh well.

?

IMC front side bus = Hyper transport speeds = QuickPath

Of course the memory and IMC will be directly tied to each other, albeit via dividers.

 

[GuitarDaddy]

Originally posted by: alfa147x

Originally posted by: GuitarDaddy
I wish they would stop already with the FSB increases, it gives minimal gains and kills the overclockability. Think what the multi's would be, probably 5.5 to 7.5

Im new to overclocking but isn't Bus Speed different the Front Side Bus?

Like i said im new to this whole thing, please dont bite my head off if im wrong

Different yes, but linked. Front Side Bus = Buss speed x 4 (quad pumped)

1066 = 266
1333 = 333
1600 = 400

 

[Idontcare]

Originally posted by: GuitarDaddy

Originally posted by: alfa147x

Originally posted by: GuitarDaddy
I wish they would stop already with the FSB increases, it gives minimal gains and kills the overclockability. Think what the multi's would be, probably 5.5 to 7.5

Im new to overclocking but isn't Bus Speed different the Front Side Bus?

Like i said im new to this whole thing, please dont bite my head off if im wrong

Different yes, but linked. Front Side Bus = Buss speed x 4 (quad pumped)

1066 = 266
1333 = 333
1600 = 400

"Bus" speed is what is used to drive the clocks for everything else.

DDR2 is 2X (2x267, etc) multi off the bus speed (and can be higher with memory multipliers/dividers).

FSB is a 4X multi off the bus speed (4x267, etc, as GuitarDaddy pointed out).

CPU is its own multi (6X, 7X, 9X, etc) off the bus speed (9x267, etc).

 

[DerwenArtos12]

Originally posted by: Idontcare

Originally posted by: GuitarDaddy

Originally posted by: alfa147x

Originally posted by: GuitarDaddy
I wish they would stop already with the FSB increases, it gives minimal gains and kills the overclockability. Think what the multi's would be, probably 5.5 to 7.5

Im new to overclocking but isn't Bus Speed different the Front Side Bus?

Like i said im new to this whole thing, please dont bite my head off if im wrong

Different yes, but linked. Front Side Bus = Buss speed x 4 (quad pumped)

1066 = 266
1333 = 333
1600 = 400

"Bus" speed is what is used to drive the clocks for everything else.

DDR2 is 2X (2x267, etc) multi off the bus speed (and can be higher with memory multipliers/dividers).

FSB is a 4X multi off the bus speed (4x267, etc, as GuitarDaddy pointed out).

CPU is its own multi (6X, 7X, 9X, etc) off the bus speed (9x267, etc).

Until they integrate the memory controller. *IF* it's done similarly to AMD's then there will be the CPU speed with it's associated refrence clock and everything will be based off the CPU speed with dividers instead of multipliers. *shrug* different means to the same end.

 

[tigersty1e]

Originally posted by: GuitarDaddy
I wish they would stop already with the FSB increases, it gives minimal gains and kills the overclockability. Think what the multi's would be, probably 5.5 to 7.5

Me too.

We've been on 1600 FSB for more than 1 year now.

All they have to do is repackage the E6320 with 400FSB and label them as new processsors at 2.8GHZ.

 

[Idontcare]

Originally posted by: tigersty1e

Originally posted by: GuitarDaddy
I wish they would stop already with the FSB increases, it gives minimal gains and kills the overclockability. Think what the multi's would be, probably 5.5 to 7.5

Me too.

We've been on 1600 FSB for more than 1 year now.

All they have to do is repackage the E6320 with 400FSB and label them as new processsors at 2.8GHZ.

No doubt Intel marketing knows this and it is on their timeline for Q4 SKU refresh. Get ready for 3X multi's!

 

[GuitarDaddy]

Yep, no doubt they are trying to force the enthusiast/overclocker to the high end or extreme chips with unlocked multis

Which is a sad thing for people wanting to get that extra performance out of budget hardware. For hardcore clockers seeking benchmark records, they are already using high end so they don't care

 

[bunnyfubbles]

A more important question would be if Intel is going to be releasing any 45nm chips with FSB speeds lower than 333MHz

A 45nm replacement for the E2xxx line would be sweet, although most likely only dual core (although a ~$70 4GHz chip wouldn't be anything to sneeze at )

While it might still be in the realm of possibility to see a penryn based quad core with a 266MHz fsb, it seems even more unlikely to see one with a much coveted multiplier higher than 9x (and even if we should see a 9 x 266 situation, it would most likely have extremely nerfed cache)...while 10 is most likely a sweet spot, you'll most likely have to pay a pretty penny to get it no matter what.

 

[Tempered81]

think its supposed to be e3xxx and e5xxx and e7xxx. you wont see em for a while though, the most confirmed hot chip on the horizon, IMO is the e8600 with the 10x multi.

10x400 = 4ghz with only pc2=6400 800mhz ram Price may be high 200's though and wont be out til mid summer. ;(

 

[Idontcare]

Originally posted by: bunnyfubbles
A more important question would be if Intel is going to be releasing any 45nm chips with FSB speeds lower than 333MHz

A 45nm replacement for the E2xxx line would be sweet, although most likely only dual core (although a ~$70 4GHz chip wouldn't be anything to sneeze at )

While it might still be in the realm of possibility to see a penryn based quad core with a 266MHz fsb, it seems even more unlikely to see one with a much coveted multiplier higher than 9x (and even if we should see a 9 x 266 situation, it would most likely have extremely nerfed cache)...while 10 is most likely a sweet spot, you'll most likely have to pay a pretty penny to get it no matter what.

It's a great question. One question would be what would prompt Intel to offer such a product line. Competition is usually the answer. Let's see what competition will prompt Intel to migrate the budget SKU's to 45nm. I am not too hopeful, I see Nehalem being the competition to Penyrn and that means when Nehalem comes along we'll just see the existing 45nm product lines relegated to lower pricing, but not necessarily reduced capabilities.