Nehalem's butt - im not joking.

[JAG87]

http://www.xtremesystems.org/f...howthread.php?t=177122

 

[Idontcare]

Nice. When are we supposed to reasonably expect to see some benches? Sept?

 

[aigomorla]

omg... mine..

Dayam... i hate you JAG...

?Bloomfield
·4-core
·LGA1366
·3ch DDR3
·Support QPI(Quick Path Interconnect)
·L2=8MB
·Support SMT
·2008 Q4

that looks like xmas time.

 

[Ika]

Is it a typo that Bloomfield shows 3ch DDR3, which I assume is triple-channel?

 

[dmens]

Originally posted by: Aflac
Is it a typo that Bloomfield shows 3ch DDR3, which I assume is triple-channel?

no it is not a typo

 

[n7]

I may be skipping Yorkfield to save the pennies for this.

 

[Dadofamunky]

Good chip porn there. Looks like the Bloomfield will cost a bloomin' fortune.

 

[manimal]

Someone in that original post said something about this chip being the end of overclocking? what gives?

 

[Denithor]

Bloomin' fortune indeed, as you will need the chip itself, along with motherboard and DDR3 at the very minimum even if you recycle your other parts.

Damn, I don't want to be an early adopter again.

 

[n7]

Originally posted by: manimal
Someone in that original post said something about this chip being the end of overclocking? what gives?

IMC kills extreme overclockability [is what they're saying].

(Intregrated Memory Controller)

It's somewhat true, but it's not known how an IMC will affect Intel yet...some people are drawing conclusions already though, based on nothing

 

[Idontcare]

Originally posted by: manimal
Someone in that original post said something about this chip being the end of overclocking? what gives?

It is a question at the moment as to exactly what "clock" an enthusiast would change on a quickpath system to affect the CPU clock.

Right now we know we can change the clock of the FSB via the Northbridge, and this modulates the clock of the CPU.

Without FSB clock to change, what confidence is there that Intel will leave open the option of changing the clock of whatever the CPU takes its multiplier off of.

There is no proof of this being a concern, but because there is no proof that it shouldn't be a concern it causes some consternation amongst those more paranoid posters in the audience.

 

[Duvie]

Originally posted by: n7

Originally posted by: manimal
Someone in that original post said something about this chip being the end of overclocking? what gives?

IMC kills extreme overclockability [is what they're saying].

(Intregrated Memory Controller)

It's somewhat true, but it's not known how an IMC will affect Intel yet...some people are drawing conclusions already though, based on nothing

I agree...they are likely just basing it off of AMD chips, and IMO they still have overclocked well enough that us hard-core oc junkies will still have something to shoot for.....



Wow socket 1366? I see the huge pin count, but how does it relate in actual dimensions of the package? It says bigger, but how much?

 

[wwswimming]

why don't we have pictures here at Anandtech ? i would say, "conscious decision to keep
bandwidth & hosting costs down", but usually when people insert an image in a post using
software that allows it, they use an image hosted somewhere other than the forum anyway.

hard to count that many gold dots. i'll take your word for it.

 

[aigomorla]

Originally posted by: wwswimming
why don't we have pictures here at Anandtech ? i would say, "conscious decision to keep
bandwidth & hosting costs down", but usually when people insert an image in a post using
software that allows it, they use an image hosted somewhere other than the forum anyway.

hard to count that many gold dots. i'll take your word for it.

because the people who have these ES chips are usually super high volumn vendors OR motherboard company's needing ES samples to make platforms for the chips.

 

[VirtualLarry]

Originally posted by: Idontcare

Originally posted by: manimal
Someone in that original post said something about this chip being the end of overclocking? what gives?

It is a question at the moment as to exactly what "clock" an enthusiast would change on a quickpath system to affect the CPU clock.

Right now we know we can change the clock of the FSB via the Northbridge, and this modulates the clock of the CPU.

Without FSB clock to change, what confidence is there that Intel will leave open the option of changing the clock of whatever the CPU takes its multiplier off of.

There is no proof of this being a concern, but because there is no proof that it shouldn't be a concern it causes some consternation amongst those more paranoid posters in the audience.

I'm pretty sure that this won't be a problem. All CPUs take their clock from an external source. As long as that's true, and mobo makers continue to offer the features that they have, then this clock source will be adjustable. Whether or not Intel offers dividers and whatnot to keep other clocks within stable ranges when this external clock is increased is another question entirely.

 

[Idontcare]

Unfortunately the way the pendulum seems to swing in this industry is that whomever has the market lead with current gen products tends to "engineer out" the wiggle room that enthusiasts have to "buy cheap and clock high" the next gen products.

I'd like to believe Intel has no reason to engineer us out of the loop but this has been consistently proven to not be the case in the past.

When one side is down and out then they have nothing to lose by letting an enthusiast crowd blossom, but as soon as they get the upper-hand you see them "take measures" that seem intentional to restrict the overclockers.

When did AMD start releasing their Black Edition chips? When they had the lead over Intel, or after they had measurably fallen from grace?

We've seen Intel take steps to "corral" the enthusiasts into this uber-expensive upper end with their Extreme edition CPU's. If I were a product director at Intel interested with extracting as much wealth from the marketplace as possible I would continue this theme into the chipset markets...releasing "extreme" versions of chipsets which have unlocked chipset multipliers for ram and FSB, etc but have the mainstream chipsets be "locked".

Only time will tell what Intel will do with their current lead. I am not too hopeful given the history of this industry.

 

[Nemesis 1]

Based on available info. I have to agree 100% with Idontcare. Intel seems to building A platform for the high end . and than the rest. But if the highend platform is all that. I don't see what differance it makes. As the people who buy the highend will finally get something better.

 

[RussianSensation]

I have to say having owned E6400 @ 3.4ghz since mid 2006 and Q6600 since mid 2007, Core 2 architecture doesn't feel that exciting/fresh anymore Oh no! the Upgrade bug... Let's hope DDR3 prices come down fast enough!

 

[superstition]

Originally posted by: RussianSensation
I have to say having owned E6400 @ 3.4ghz since mid 2006 and Q6600 since mid 2007, Core 2 architecture doesn't feel that exciting/fresh anymore Oh no! the Upgrade bug... Let's hope DDR3 prices come down fast enough!

Until I built my e2140 system I was using a 1.8 GHz Sempron 3100+ (64-bit with SSE-3), socket 754, with an ECS motherboard that couldn't be overclocked because it had no locks. My other system was a 1 GHz G4 iBook that used SDRAM. So, I don't have a lot of sympathy for people who say their quad-core chips aren't fresh enough.

lol

 

[Tempered81]

Yeah running core 2's in our custom pc's is pretty fresh. Log on to your grandmas pentium II or try out the old p4 from HP at work thats clogged with the 2003 pre-installed partner software, and 6-million windows updates/spybot search & destroy, windows defender, macafee, AND symantec & it takes seriously 11 minutes to fully boot. And then after booted, it runs slower than a 486 in windows 3.1

When you can install the entire windows xp OS in less than 11 minutes, it's fresh!

 

[RussianSensation]

Originally posted by: superstition

So, I don't have a lot of sympathy for people who say their quad-core chips aren't fresh enough.

lol

Well I upgrade my cpu every 2-3 years so:

2001 = XP1600+
2003 => P4 2.6@3.2ghz (new cpu type)
2006 => C2D 2.13@3.4ghz (new cpu type)
2007 => C2Q 2.4@3.4ghz (same cpu type but 2x the cores). Problem is I still consider this the same system as the 2006 I had since in a lot of applications it's not faster than the dual core system I had in 2006. Therefore, this C2D/Q setup has served me for almost 2 years now which is getting "old." Yes I get 2x the speed in video encoding and Winrar is about 40% faster, but it's not like I am getting 2x the framerates in games or programs install 2x faster.

With Nehalem, not only will you be able to get 8 cores down the line (which helps in threaded apps), but it's going to be a lot more efficient per each clock cycle. This means you'll get faster performance even if a program is not coded well for multi-cores.

 

[zephyrprime]

Originally posted by: n7

Originally posted by: manimal
Someone in that original post said something about this chip being the end of overclocking? what gives?

IMC kills extreme overclockability [is what they're saying].

(Intregrated Memory Controller)

It's somewhat true, but it's not known how an IMC will affect Intel yet...some people are drawing conclusions already though, based on nothing

If anything, I think it will make overclocking easier. It's well known that the A64 could run with lower speed memory dividers when overclocking the fsb. Current intel chips couldn't do this without a performance penalty because of their separate memory controller.

 

[taltamir]

oooh, shake it you.... err.... is a CPU a male or a female?

 

[dmens]

Originally posted by: n7

Originally posted by: manimal
Someone in that original post said something about this chip being the end of overclocking? what gives?

IMC kills extreme overclockability [is what they're saying].

(Intregrated Memory Controller)

It's somewhat true, but it's not known how an IMC will affect Intel yet...some people are drawing conclusions already though, based on nothing

imo it is entirely untrue. logical criticality is almost certainly not at the memory controller.

 

[lopri]

Originally posted by: Nemesis 1
Based on available info. I have to agree 100% with Idontcare. Intel seems to building A platform for the high end . and than the rest. But if the highend platform is all that. I don't see what differance it makes. As the people who buy the highend will finally get something better.

You got it backwards. It's not about Intel giving better stuff that's worth more money, but it's about NOT giving more for mid-range buyers. A crude imaginative example: It's not like making QX6700 better, but more like throwing an artificial FSB wall (say at 300FSB) on Q6600. Of course the purpose is to squeeze more money out of enthusiasts and raise the overall prices therefore their margins.