Sandy Bridge E not being released untill Q4 2011?

[dangerman1337]

http://www.nordichardware.com/news/...oards-being-tested-by-intel-and-partners.html

Woah Looks like we'll have to wait even longer, lets hope this doesn't mean AMD's bulldozer is going to be released later than that and i hope it comes with Lightpeak and/or USB 3.0 native aswell if the date is true :\.

 

[ilkhan]

AFAIK the date we "had" was H2 2011, so not really a surprise. 3 years on s1366 should make the early adopters feel good. Even a launch i7-920 is pretty near top of the line, still. Itd cost $900 to get more than just a few hundred Mhz better.

 

[dangerman1337]

AFAIK the date we "had" was H2 2011, so not really a surprise. 3 years on s1366 should make the early adopters feel good. Even a launch i7-920 is pretty near top of the line, still. Itd cost $900 to get more than just a few hundred Mhz better.

Werid that LGA1366 has been out LGA1156 but the later is going to be discountiuned for LGA 1155 eariler than LGA2011 coming out.

I also wonder what will this mean for Haswell/rockwell.

But i hope Sandy bridge E is 22nm though.

 

[IntelUser2000]

They won't be ready with a 22nm shrink by that time, not only because the process won't be ready but because the higher end parts changes layout. Sandy Bridge means 32nm, Ivy Bridge and Haswell will continue the way of introducing mainstream platforms first.

 

[dangerman1337]

They won't be ready with a 22nm shrink by that time, not only because the process won't be ready but because the higher end parts changes layout. Sandy Bridge means 32nm, Ivy Bridge and Haswell will continue the way of introducing mainstream platforms first.

Yeah that's true but then again there's the possibility of Sandy bridge to turn into Ivy Bridge E, it has happend before (Clarkdale and arrandale replaced a 45nm nehalem derative dual core but forgot what it was called though).

 

[Idontcare]

Yeah that's true but then again there's the possibility of Sandy bridge to turn into Ivy Bridge E, it has happend before (Clarkdale and arrandale replaced a 45nm nehalem derative dual core but forgot what it was called though).

This may seem like a pedantic argument of semantics to you, but just to clarify Sandy Bridge can not "turn into" Ivy Bridge E.

Intel can elect to not bring Sandy Bridge E to market but they can't turn it into Ivy Bridge E.

Clarkdale and Arrandale are 32nm products that were intended to be the successors to Havendale and Auburndale 45nm products.

Havendale and Auburndale did not "turn into" Clarkdale and Arrandale, they simply were never brought to market in the first place.

I dote on this technicality because some folks have the false perception that migrating chip designs across process nodes is as simple as editing the project folder name in Windows Explorer and pressing some button in a software app that says "re-optimize chip for Node XYZ". It don't work that way.

 

[Nemesis 1]

Well what ever It is . It makes sense that It will be a 22nm chip. It would also likely be able to scale to 12 cores. Call it SB if no core changes made . Call it Ivy . If there are core changes made. But Ivy to me will have core changes in the IGP . The core don't know don't care . SB lowend is being fabbed now and has been for some time . Intels not performing any mircles here in the time required to shrink to another node . Its taken them over a year . And they have 22nm core going threw the fab as we speak . So . Being released a year from now shows only hard work . No magic here the time factor is evident .

I think that what we may be seeing is something new from intel. I think that we may have to face the reality that we need to think outside the box. What if . It is highend SB going to 22nm . That would mean something new. I am thinking that the shrink to 16nm won't happen 2 years after 22nm . I think it will take longer. Lets say the 22nm release is High end sandy . What about Ivy bridge for the desktop and notebooks. Is it possiable that Ivy bridge for desktop is still on 32nm . I follow what IDC is saying . I for the most part would have to agree with him . Based on what intel has done in the past. Who is to say intel isn't going to capitilize on its investments in the fabs.

 

[Nemesis 1]

I clicked on your link . I am not happy . The ozone advertizement stays up on the screen . I clicked on it several times . Each click is a payday for Nordic. I have read there stuff befor . Its not real reliable. Than this Ozone scam just shows what this article is really about.

 

[theAnimal]

I clicked on your link . I am not happy . The ozone advertizement stays up on the screen . I clicked on it several times . Each click is a payday for Nordic. I have read there stuff befor . Its not real reliable. Than this Ozone scam just shows what this article is really about.

There are ads on the Internet? You must be using the wrong one.

 

[Nemesis 1]

Ya but this ad stays onscreen so you can't read the full article . So you click the ad to get rid of it but nothing happens . Other than going to another page . Each click is money in nordics pocket . I clicked on everthing I seen in graphics but ad stays.

 

[RussianSensation]

Well what ever It is . It makes sense that It will be a 22nm chip. It would also likely be able to scale to 12 cores. Call it SB if no core changes made . Call it Ivy . It is highend SB going to 22nm .

SB will be 32nm as has been said many times. LGA2011 being released in Q3 2011 at the earliest was reported by BitTech.net 6 months ago. Refresh of SB will be 22nm, but not SB. SB will be 6- and 8-core derivatives. No idea where you got your information that it will scale to 12 cores on 32nm for the consumer market (maybe servers).

 

[Nemesis 1]

^ That would be 12 cores @22nm . That is speculation based off of what intel said at IDF. 10 cores at 32nm has been around for sometime.Intel does have an 8core already on the market at 45nm. Also why didn't bit tech publish SB specs 6 months ago Intel has been pretty good about plugging leaks. The asian market is the leak source and I for one don't like their agenda. At no time did I say these chips were for desktop.

Lets stop and think . The GPU . Those cards are all made in the Asian markets . Behold intel/AMD start making SoC with good graphics . The American markets gain from this . The asian markets lose. I am all for the asian markets losing products to sell . Of course NV loses so to does AMD . But AMD has no choice in the matter. Intel will keep pushing the SoC chips to higher and higher performance. Amd has to stay with intel or lose CPU share. So in effect AMD has to kill a product it sells seperately in the market place. AMD payed alot for ATI . But looking at SB it was a smart buy at the wrong price.

 

[RussianSensation]

10 cores at 32nm has been around for sometime.

You are confusing the server market and the consumer market. Just because Intel can release a 32nm 10-core SB on LGA2011, doesn't mean it is their market strategy at this time.

Nordic Hardware did a poor job in that article in providing any sort of tangible info. This article provides much more information.

 

[Nemesis 1]

Me personnally don't care about any of it . The only thing I am concerned with is the M/B pricies. If M/B pricies don't go lower than we have a problem with price fixing which can easily be proven in a court of law. Robert has sworn to carry on my work . So our organization will fight for all of us. I should add here that if intel tries selling the south bridge at high pricies they will get pounded by everyone the EU included. No one will alow that to happen, Asian marketers might tell you the Sbridge is going to be $45 dollars . But its a flat out lie. These SoC are going to hurt M/B makers and everyone better dam well understand that.

 

[Nemesis 1]

You are confusing the server market and the consumer market. Just because Intel can release a 32nm 10-core SB on LGA2011, doesn't mean it is their market strategy at this time.

Nordic Hardware did a poor job in that article in providing any sort of tangible info. This article provides much more information.

No i not confused at all . I read the title . Its a server subject . Now when I speak of SoC clearly were talking about consumers market. I read your link . Its ok in the fact its more about M/B than SB tech details. Thats the whole point . Asian markets leak Info that they have access to . AMD NV Intel are fully aware of this. So Asian are now more limited to the actual factual facts the Anerican companies release to them . How was that SB preview by coolar possiable . Simple enough he works for a M/B maker and is high up or he would be out on the streets by now. Just for the record. I don't care if its the high end desktop . Its still server type chip.

 

[ilkhan]

s1156/s1155 = consumer = 2/4c with GPU
s1366/s2011 = 1-2 socket workstation = 4/6/8c without GPU
s1567 = 4-8 socket server = 8/10c without GPU

We have no idea what socket sandy/ivy's s1567 equivalent will be (if it does get a new socket, if they leave the clock gen on the board it may not).

 

[WaitingForNehalem]

AFAIK the date we "had" was H2 2011, so not really a surprise. 3 years on s1366 should make the early adopters feel good. Even a launch i7-920 is pretty near top of the line, still. Itd cost $900 to get more than just a few hundred Mhz better.

I've been enjoying my i7 920 since November 16, 2008 .

 

[dangerman1337]

This may seem like a pedantic argument of semantics to you, but just to clarify Sandy Bridge can not "turn into" Ivy Bridge E.

Intel can elect to not bring Sandy Bridge E to market but they can't turn it into Ivy Bridge E.

Clarkdale and Arrandale are 32nm products that were intended to be the successors to Havendale and Auburndale 45nm products.

Havendale and Auburndale did not "turn into" Clarkdale and Arrandale, they simply were never brought to market in the first place.

I dote on this technicality because some folks have the false perception that migrating chip designs across process nodes is as simple as editing the project folder name in Windows Explorer and pressing some button in a software app that says "re-optimize chip for Node XYZ". It don't work that way.

Well i was just saying :whiste:.

Though i have one thought, why replace LGA1156 but not LGA1366? Confusing that 1156 is going to have a short socket life and 1366 will have limited upgradability, unless intel can release a Gulftown in the price range of the 950 or 960 at least but i doubt it. Hell if you think about it Bloomfield has stayed on a long time now on the market compared to say conroe or penryn.

 

[ilkhan]

Only reason bloomfield is still around is to optimize production capability. Gulftown (on 32nm) is actually a smaller die than bloomfield (on 45nm). But I expect Intel has more 45nm capacity (and its a more mature process) than 32nm, so the chips are still around. Plus it keeps gulftown margins high.

 

[OBLAMA2009]

i didnt bother reading that article but does that mean that the current i7 designs are going to be the top of the line for the next year????

 

[Idontcare]

Well i was just saying :whiste:.

Though i have one thought, why replace LGA1156 but not LGA1366? Confusing that 1156 is going to have a short socket life and 1366 will have limited upgradability, unless intel can release a Gulftown in the price range of the 950 or 960 at least but i doubt it. Hell if you think about it Bloomfield has stayed on a long time now on the market compared to say conroe or penryn.

Replacing LGA1156 with LGA1155 is necessary because of the IGP on LGA1155 SB's.

Conroe and Penryn had more competition from Phenom and PhenomII. Bloomfield on LGA1366 really hasn't had much competition, that increases the shelf-life of the products.

I don't think we can read the tea leaves of bloomfield's market presence and extract anything meaningful in terms of what Intel could have done in the same timeframe had AMD fielded a 970X or 980X competitor. All we can divine from the tea leaves is what Intel elected to do in that timeframe for that market segment given the absence of AMD in that price and performance segment.

 

[ilkhan]

i didnt bother reading that article but does that mean that the current i7 designs are going to be the top of the line for the next year????

Remember that s1366 for desktop use is just a "we're already kicking your ass so this is really just for the higher margins" type of part. The chips/chipsets/boards are already in production for the xeon 3xxx series, it doesn't cost them much to sell those chips to consumers at double-triple the s1155 margins.

Sysmark http://www.anandtech.com/bench/CPU/2?i=142.192.186.146.52
h.264 http://www.anandtech.com/bench/CPU/54?i=142.146.192.191.186
crysis http://www.anandtech.com/bench/CPU/50?i=142.107.146.102

 

[theAnimal]

Ya but this ad stays onscreen so you can't read the full article . So you click the ad to get rid of it but nothing happens . Other than going to another page . Each click is money in nordics pocket . I clicked on everthing I seen in graphics but ad stays.

No ad for me. Proxomitron FTW!