Haswell: News, Rumors & Reviews

[EXT64]

Trust me, I'm not buying it (at least not anytime soon). I doubt I could afford it anyway :lol:
Good ole 1366 is good enough for me for a while.

 

[metallifux]

It wouldn't make any sense for Intel to release LGA2011 for consumers and put LGA1356 only for servers.

There will be a large number of consumer processors and boards available just from GLA 1155 and LGA 2011, do consumers really need another socket? Do board makers want to cater to one? Where could it fit in between the mainstream/enthusiast cross over. Two sockets have worked for a long time and Intel has also held its price points for many years, I'd not expect much change.

I reckon 2 sockets is more than enough for consumers. I really don't understand why they need more than one? Isn't that why we have more than one chipset so that we can have different levels of awesomeness/price.

 

[JAYDEEJOHN]

Pull-ins are generally possible "if needed", though usually at a loss of potential profits and/or performance. And, no, that doesn't translate to "We must soak them on the old chips before we begin selling them the new!"

Rather, it's a matter of ongoing process and design tweaks to make the chip higher performing, higher yielding, better bin split, lower power, and cheaper to manufacture... all of which need to be delayed or bypassed altogether to bring a processor out "if needed". Think the G0 stepping of Conroe, but in reverse: instead of a later stepping with significant improvements, you get an earlier stepping which is still being improved upon. Sure, it can be done to face a major market threat, but generally you stick to your roadmap because it's a tried and true amount of time that it takes to get things dialled in for the marketplace.

That said, sometimes the design or process surprises you and the A-step silicon is remarkably healthy. But there are nearly always improvements which can be made which are worth taking the time for.

Osborne anyone?
Yes, it certainly can mean soak, and going thru the entire process dev is as important to Intel as it is to anyone.
Its convenient if you have a comfortable place in any market, but alot of markets, this simply isnt the case, and those players in those markets also do quite well, and their consumers are also treated quite well.
To totally deny it simply cant be done, no matter what market youre in.

Generally I dont think Intels soaking us much, but since the urgency isnt there in this market, its too vague to make such absolutes.
Now, going to the other side of this, the "if needed", this also can be construed differently by my above comments, as it depends on your position in a given market, where needs vary, both from consumers, competitors and partners or trends, even regs, such as USB3
But it sounds good tho

 

[archibael]

Not sure I'm getting your Osborne reference.

This isn't vague yammering, this is experience. I can't speak for any of the non-mainstream business units, but I have been involved in or witness to the decision-making process on every major product release since Tualatin, and never once has a decision been made to delay or even to "fail to pull in" a ready product release because the current product on the market was doing so well. The faster you get the new product out the door, in fact, the faster you start recouping your R&D costs and get your engineering team working on the next product.

This goes doubly for a product which debuts a new fab process, as the business accounting usually lays the entire cost of engineering the new fab process on the first product to use it, and therefore you want to turn a profit on that product ASAP to make your return-on-investment for the process node positive. (Stupid, and funny money, it seems like, but that's corporate finance. It's probably a tax thing.) And generally speaking, a shrink uses less silicon, so you want to get to that node ASAP for obvious cost reasons (more die per wafer).

From my experience, I can say pretty categorically that if the product's delayed, it's not because there's a warehouse of "the good stuff" waiting for some competition to materialize.

 

[Leaps-from-Shadows]

Some info on LGA 2011, SNB-E and even the mythical IVB-E: Linky

 

[ghnader hsmithot]

Socket 2011 Futures: A Difficult Road to Perfection
Reported by Nebojsa Novakovic on Saturday, July 30 2011 11:31 pm

Socket 2011 will be Intel's new solution for high-end desktop, workstation and server processors. While the full potential of this platform will be unleashed only with next year's 22nm Ivy Bridge-E processors, Sandy Bridge-E is still likely to dominate the top-end desktop market till then.
Tweet
0
REVIEW

LGA2011: The Biggest Socket Ever

When news of Socket 2011 first appeared over a year ago, many were quite excited. This was hardly surprising since Socket 2011 will be the new solution for high-end desktop, workstation and server processors. This behemoth of a socket is so huge it requires two retention levers, one on each side, for the very first time. Despite this, Socket 2011 will retain compatibility with LGA1366 cooling solutions. The competition didn't - and still doesn't - have anything comparable, save trying to put multiple dies onto a single chip in the G34 socket.

Why such a huge socket? Sandy Bridge-E is a huge chip: the full-flavour edition will have eight cores with 20 MB of shared L3 cache on a ring bus, a quad-channel DDR3 memory controller and two QPI links. Socket 2011 also has to ensure platform scalability beyond eight cores and 20 MB of cache. Even on the 32 nm process, all this needs lots of space, not to mention power.

The new CPUs, which had originally been expected to arrive around now (Q3 2011) have supposedly been pushed back to end 2011, with some sources even estimating early 2012. It seems unlikely that Intel would further delay Sandy Bridge-E since AMD (after all those delays) is expected to launch Bulldozer no more than seven weeks hence.

Read more: http://vr-zone.com/articles/socket-2011-futures-a-difficult-road-to-perfection/13147.html#ixzz1TebWuTPA

Wow!

 

[ghnader hsmithot]

Hopefully,
With that big of a socket we have that much bigger a performance.

 

[ghnader hsmithot]

So is a dual cpu setup possible?
With intel you know anything is possible.

 

[ghnader hsmithot]

I guess i dont know about this much but according to information i have that the Ivy bridge would most likely have their Clock generators on the die itself.

No one is interested in this thread....

 

[ghnader hsmithot]

2015 Version Of Intel Atom "Faster Than AMD Phenom II" CPU

Read more: http://www.itproportal.com/2011/07/26/2015-version-intel-atom-faster-amd-phenom-ii-cpu/#ixzz1ThGJ8EW1

 

[JAYDEEJOHN]

My referencing release timings are more in the minutia, where it does happen, and in smaller windows of time.
To broaden it out only strengthens what you said, and is not what Im refering to.
You can run into an Osborne effect if your supply is loaded, and with stiff competition, and it does happen.
The opposite does as well, when your market is casual with a comfortable lead.
Im just not accepting it never happens, and it makes things look better than they really are.

 

[jimmysmitty]

So is a dual cpu setup possible?
With intel you know anything is possible.

Yes. In fact they already have some dual socket LGA2011 mobos in the making. Thats the whole purpose of having 2 QPI links as well, is for multi CPU support so each CPU can have the fastest conection possible.

2015 Version Of Intel Atom "Faster Than AMD Phenom II" CPU

Read more: http://www.itproportal.com/2011/07/26/2015-version-intel-atom-faster-amd-phenom-ii-cpu/#ixzz1ThGJ8EW1

I would not be suprised but then again it will be 4 year old CPU vs brand new by then.

I am more interested in the 22nm Atom and the power savings it can do. I think Intel needs to push that out next year instead of 2013.

 

[dkjazz]

On motherboards that don't have on-board graphics, there will be no support for QuickSync and there will not be Lucid Virtu on them. These are some of the features that users have been waiting for the the Z68 boards

 

[mayankleoboy1]

http://www.phoronix.com/scan.php?page=article&item=mesa_711&num=1

There is also initial support for Intel Ivy Bridge within Mesa 7.11. Ivy Bridge is the Sandy Bridge successor that's expected to be released by the end of the calendar year, but initial support for IVB graphics are already in place within Mesa, the xf86-video-intel DDX, and the Linux kernel to ensure a better "out of the box" experience this time around.

 

[ghnader hsmithot]

LGA2011 real outbreak: Ivy Bridge-E

Although not yet seen any public road map, but Sandy Bridge-E should be the next generation after 22nm Ivy Bridge-E, is expected to bring more of the core, optimized caching, richer features, the interface will continue to use the LGA2011 interface, while the release time at least until the end of 2012 or even 2013.

http://tech.163.com/digi/11/0801/15/7ACND19E00162OUT.html

 

[fazers_on_stun]

Hopefully,
With that big of a socket we have that much bigger a performance.

Sheesh, that is a HUGE socket. However I wonder if that is a full-size ATX board or the MATX board?? Hard to tell from the picture..

 

[ghnader hsmithot]

So what can we expect tp achieve from a big socket??

 

[Leaps-from-Shadows]

With that big of a socket we have that much bigger a performance.

Sheesh, that is a HUGE socket.

So what can we expect to achieve from a big socket??

According to an informal and entirely unscientific survey of random women in a local mall, the size of the socket doesn't matter...

According to an equally informal and unscientific survey of random men in a local bar, the women from the first survey were lying through their teeth.



Anyway...

The socket is big, yes. But it's not that big. Remember ... 1366 coolers work on the 2011 socket.

 

[ghnader hsmithot]

According to an informal and entirely unscientific survey of random women in a local mall, the size of the socket doesn't matter...

According to an equally informal and unscientific survey of random men in a local bar, the women from the first survey were lying through their teeth.



Anyway...

The socket is big, yes. But it's not that big. Remember ... 1366 coolers work on the 2011 socket.

Leaps!I think its big! its big combined with 32nm we might see great performance numbers.Go Intel!

Below as you can see.Top left socket 1155 top right is 1156, bottom right is 1366 and of course LGA 2011 is bottom left.

 

[ghnader hsmithot]

I guess i might wait for Ivy Bridge E.This sucks!

Why in the world do they have to have a new socket!

 

[ghnader hsmithot]

I guess no one on this forum is an intel fanboy..

 

[Leaps-from-Shadows]

Good news!
SNB-E and X79 to be released in November.

Bad news...
More features cut out of X79. Basically, it's now just an upgrade of P67 with two SATA3 and four SATA2 ports.

Linky!

SNB-E and IVB base clock options detailed:
Linky

 

[Umbongo]

I guess i might wait for Ivy Bridge E.This sucks!

Why in the world do they have to have a new socket!

They are moving from LGA 1366 because you need more pins for more features present in the Sandy Bridge and, in the future, Ivy Bridge architectures.

 

[Ubrales]

I guess no one on this forum is an intel fanboy..

Ghnader, keep on trucking! I am a die-hard Intel guy!

 

[metallifux]

Good news!
SNB-E and X79 to be released in November.

Bad news...
More features cut out of X79. Basically, it's now just an upgrade of P67 with two SATA3 and four SATA2 ports.

Linky!

SNB-E and IVB base clock options detailed:
Linky

what about native usb3?