[NEW STUFF] SANDYBRIDGE LGA2xxx

[Olivon]

Where does it show an IGP on these guys?

Hope not ...

[radaja]

the future looks very exciting.great thing to come from both Intel and AMD

[-Sweeper_]

wow nice find, we'll be able to keep our current coolers!!

can you tell anything about the frequency JC?

[terrace215]

Nehalem has also sliced L3 cache.

Indeed its shared, but it still is physically allocated to a core.

Well, the -EX does.

[terrace215]

Hope not ...

I think it's rather unlikely...

[Hornet331]

Well, the -EX does.

27.jpg

Not only the EX

[terrace215]

Not only the EX

Check out:

http://www.hotchips.org/archives/hc2...tel-NHM-EX.pdf

The N-EX L3 is like what comes with SB. The other Nehalem/Westmere L3s are not, "sliced" though they may also be.

[Olivon]

wow nice find, we'll be able to keep our current coolers!!

can you tell anything about the frequency JC?

They are blanks but maybe 3200 MHz + Turbo

[informal]

Are you seriously expecting 3.2Ghz from early ES of 8 core SB? Sure ,it's possible but looking at QC offerings and TDPs,3.2Ghz 8 core would be above 130W TDP.

[jogshy]

What I won't understand is why Intel is naming them as i3/i5/i7 and not something new like s3/s5/s7 ( s of Sandy :p and i > s in the alphabet too ! )

[saaya]

1200 would be too low
2200 would make sense, in the past early es chips from intel were always 2400
3200 sounds like too much, possible but unlikely... maybe its 3200 or even 4200, but then its because somebody played with the unlocked multipliers

[sin0822]

i bet hes dying to tell us how this benches but he cant. I mean hes got everest right there on his desktop.

[terrace215]

1200 would be too low
2200 would make sense, in the past early es chips from intel were always 2400
3200 sounds like too much, possible but unlikely... maybe its 3200 or even 4200, but then its because somebody played with the unlocked multipliers

Erm, there wasn't much load on it. Heard of "turbo" ?

[R101]

20 mb l3 :d

[Manicdan]

my bet is 2200mhz, unless its been overclocked

[informal]

my bet is 2200mhz, unless its been overclocked

2.2Ghz is the most probable frequency.With no load,Turbo won't kick in.

[Manicdan]

2.2Ghz is the most probable frequency.With no load,Turbo won't kick in.

yup

i was hoping that maybe its 200mhz, because at idle it would be nice to see incredible down clocking. but i dont believe thats the case due to the Everest image, in the top right the blank space for multiplier number is big enough for 2 digits, the question is 12 vs 22/32, and i wonder if someone can find out if starts with a 1 or not based on pixels

[FlanK3r]

2200 MHz think for 8 cores with HTT, ES and TDP about 130W its ok

[-Sweeper_]

its just an ES, early SB quad-core samples also run at low clocks (~2.5ghz)

around 3ghz is my bet for the final silicon, but they could surprise me...

[ajaidev]

No comments other than still under kinda NDA.

[Manicdan]

did you really have to block the intel logo too, lol

[terrace215]

No comments other than still under kinda NDA.

Well, hmm, same cpu-z version, but yours has Xeon branding, JC's doesn't and is presumably the high-end desktop part... with shockingly similar properties...

Maybe you were attempting to hide "Xeon" with those somewhat misplaced white boxes???

[IDANPE]

intel have 2.26GHZ 8C/16T now at 45NM..
3.0GHZ+ for the final clock should be Possible with SB and 32NM..

and the pic from jc is in idle so i think 12X100-> 1.2GHZ idle instead of 1.6GHZ in Nahelem/Westmere.

[FlanK3r]

ajaidev: where did u it found?

[saaya]

Erm, there wasn't much load on it. Heard of "turbo" ?

why so c0cky?

With no load,Turbo won't kick in.

i was hoping that maybe its 200mhz, because at idle it would be nice to see incredible down clocking.

i think we wont ever see that happen... everytime the clocks change, voltages change, you lose a couple clock cycles... actually a couple hundred or thousands... and 200mhz is so low that youd have to up the clocks as soon as you move the mouse cursor or some background task starts doing something

intel have 2.26GHZ 8C/16T now at 45NM..
3.0GHZ+ for the final clock should be Possible with SB and 32NM..

and the pic from jc is in idle so i think 12X100-> 1.2GHZ idle instead of 1.6GHZ in Nahelem/Westmere.

very good point...
12x100 in idle sounds about right for a desktop cpu...