Intel brings forward Nehalem Launch

[terrace215]

Good to see that they are bringing forward the launch, but doesn't this mean it will be a paper launch because the parts won't be arriving?

Presumably the launch has been brought forward because Intel's roadmaps are conservative, and Nehalem progressed to production-worthiness earlier than the end of the allotted schedule. Why would they move it forward to have a paper launch? It's not like AMD is providing any competition at the high-end.

And remember they did get first silicon back a few weeks before last year's IDF, and it worked a lot better than they expected for that stage, per one of the lead architects.

[stangracin2]

Presumably the launch has been brought forward because Intel's roadmaps are conservative, and Nehalem progressed to production-worthiness earlier than the end of the allotted schedule. Why would they move it forward to have a paper launch? It's not like AMD is providing any competition at the high-end.

And remember they did get first silicon back a few weeks before last year's IDF, and it worked a lot better than they expected for that stage, per one of the lead architects.

yeah its just like video cards and not having enough to ship in the first 2 weeks of launch

[IvanAndreevich]

QFT

I'll probably pick up a second-hand 45nm Yorkie when people start upgrading and save most of my cash for video cards a watercooling setup.

GPU > CPU

[Hornet331]

yeah its just like video cards and not having enough to ship in the first 2 weeks of launch

we are talking about intel here, intel has 2 45nm fabs online with 2 more in the pipeline, while tsmc has one or two 55nm fab and they are shared between variouse clients.

[adamsleath]

harhar
oh no here we go again
another new product; the problem isnt with Intel the producer; it is the distributers / gougers that are the problem.

how much for a courier from the factory dispatch door? but if it wasnt for marketing i wouldnt even know there was anything to buy
ho hum i better go sort through the other quadrillion adverts around the place

[JumpingJack]

Not being a troll here, but are a lot of you guys really going to jump on this in Q4? I think I'll be waiting until I see some results that can beat current e86/x48/P45 benchmark scores. I am really curious to see what tweaks will be available in the high end bioses. Until then it's kinda boring IMO.

What makes you think you are being a troll? Very lucid and intelligent position.

I suspect -- and this is my opinion, that Nehalem in terms of adoption will be slower than in prior generational architecture change overall. The transformation of the support architecture is just very radical.

Frankly, what Intel has done AMD pulled off in 2003. And, as you may recall, the initial iteration of K8 had their hiccups, bugs, and kinks to work out. I would expect similar growing pains here. I think I will be waiting for awhile before taking a leap.

[JumpingJack]

Nehalem FTW.

The Quad is going to be a native-quad right? so bring it on! That means it's easier to cool, since it's not 2 dies smashed together.

Actually, it will be harder to cool -- if this is your only parametric for cooling. If anything, the dual die enable spreading the heat sources farther apart and helps spread the head load over the IHS.

In terms of cooling Nehalem, the question isn't is it native or dual die, it is die size and heat density. The die size is larger, but if the heat density is lower then it is a push.

[gOJDO]

@Warboy

Removing the heat from one large die is harder than removing the heat from two smaller. Also, when making a MCM you can choose two dies with similar thermals and frequency potentials. With a monolithic die it is harder to find a sample with relatively good thermals and frequency potentials.

[ziddey]

any final word yet on overclocking mainstream nehalems? still supposedly locked down?

[chuckbam]

to be honest i dont see a soild reason to change from a Q6600 witch is strong enough for
everything today.
i think i will stay with Q6600 untill it will be weak in the games/softwares or at least a 50% improvment new Cpu..
Cpus This days.. the same..

I am going to wait about 6 months. As for DDR3, Is going down in price now. cas7

http://www.newegg.com/Product/Produc...82E16820227309

[Shintai]

any final word yet on overclocking mainstream nehalems? still supposedly locked down?

You can OC them.

[naokaji]

we are talking about intel here, intel has 2 45nm fabs online with 2 more in the pipeline, while tsmc has one or two 55nm fab and they are shared between variouse clients.

to be honest, I expect the worst if it comes to nehalem avalaibility at launch, it took them monthes to make the 45nm quads avalaible in the 3rd world ehh sorry uk, but Intel is welcome to prove me wrong by having it avalaible in masses at launch, I certainly woudnt complain if it would be possible to buy them.

[Movieman]

You can OC them.

[Hornet331]

to be honest, I expect the worst if it comes to nehalem avalaibility at launch, it took them monthes to make the 45nm quads avalaible in the 3rd world ehh sorry uk, but Intel is welcome to prove me wrong by having it avalaible in masses at launch, I certainly woudnt complain if it would be possible to buy them.

well they wern't exactly forced to hurry with the 45nm consumer cpus and server + ee cpus where available within a month after announcement.

I personally still think that 45nm delay was a money decision rather then a issue with the process.

[twwen2]

You can OC them.

Good news! I guess it'd be kinda like OCing an AMD?

[Chickenfeed]

Once DDR3 becomes commonplace, the prices will fall, not go up anyway. Thats how it has always been and will always be. It's not mainstream and that's why it cost more. Heck, look at DDR1 prices?

The man speaks the truth

[Shintai]

I dont see DDR3 costing much more than DDR2 here. Specially not when you think on its DDR3-1333 vs DDR2-800. Even 4-6GB is peanuts.

Seems like a pretty mood point to moan about.

[NoobCake]

Hey guys,

What would the sizes of these boards be?

Bit bigger then standard ATX boards?

If they are, wouldn't that mean manufacturers have to make more cases?

[Shintai]

Hey guys,

What would the sizes of these boards be?

Bit bigger then standard ATX boards?

If they are, wouldn't that mean manufacturers have to make more cases?

Standard ATX. (Why would they be bigger?). Dual socket will be eATX tho.

And I think mainstream nehalems will primary be mATX.

[Shintai]

Good news! I guess it'd be kinda like OCing an AMD?

Depends what you think on. It will be the external clock thats the key factor. just like HT or FSB.

[NoobCake]

Standard ATX. (Why would they be bigger?). Dual socket will be eATX tho.

And I think mainstream nehalems will primary be mATX.

wah??? i heard somewhere they would be a bit bigger???

[Shintai]

wah??? i heard somewhere they would be a bit bigger???

We already seen X58 boards, and they aint bigger. And why would they be?

And the Lynnsfield/Havendale boards could be alot smaller.

[Neuuubeh]

Not being a troll here, but are a lot of you guys really going to jump on this in Q4? I think I'll be waiting until I see some results that can beat current e86/x48/P45 benchmark scores. I am really curious to see what tweaks will be available in the high end bioses. Until then it's kinda boring IMO.

I'll wait a bit to see first scores, and for people to start selling their "old" PCs or shops clearing out, and will probably try grab me another q9450 pc . My gf is also whining for HDD space, so I guess it could double as a fileserver of some sorts. Also been thinking of liquid cooling for some time..

I sure as hell dont want to be the very first adopter of any tech, I cant really think of any advantage to that. The disadvantages on the other hand..

[iboomalot]

interesting people complaining about ram when current DDR3 will work and if you have to upgrade the prices by OCT should be even lower.

I would think having one choice in mobo the X58 and with mobo prices climbing should be what the people complain about.


Kinda said when a $200.00+ P45 mobo is considered a good buy or GPUs that run over 300.00


I guess its better than the 2,500 I paid for a 1984 256K MAC

[Zaskar]

Not being a troll here, but are a lot of you guys really going to jump on this in Q4? I think I'll be waiting until I see some results that can beat current e86/x48/P45 benchmark scores. I am really curious to see what tweaks will be available in the high end bioses. Until then it's kinda boring IMO.

Ill be waiting for the second round of boards.

MB manufactures always release a basic "high end" board at first for new architectures, and then a month or 2 later release one thats usually much better since they try to freeze the design soon enough for the first so they can take it to market as soon as possible. They also like getting people who want the cutting edge to buy 2 boards by holding back some features from the first :P