[TechSpot] IB-E delayed until Mid 2013

[[XC] Oj101]

Don't forget that LGA2011 was not designed as an enthusiast platform, it was designed as a workstation/server platform. As such, new technologies and features are validated on the consumer platform (LGA1155 in this case) before moving on to the workstation/server platform.

[Tenknics]

What would you say if I told you that IB-E will be based on SB-E?

Uh is this a joke? Of course it is, just like regular mainstream Ivy is based of regular mainstream Sandy.

Every tock is always based off a Tick...that's kinda the whole point... this isn't news unless you were being sarcastic?

[[XC] Oj101]

No not quite, I'm talking a 22 nm SB-E. SB-E is based on SB, IB-E is based on SB-E.

[Mats]

Don't forget that LGA2011 was not designed as an enthusiast platform, it was designed as a workstation/server platform. As such, new technologies and features are validated on the consumer platform (LGA1155 in this case) before moving on to the workstation/server platform.

Nothing stopped Intel from launching Bloomfield nine months before Lynnfield, so I can't really see why it suddenly would be impossible to launch Haswell-E say six months after Haswell.

[Tenknics]

No not quite, I'm talking a 22 nm SB-E. SB-E is based on SB, IB-E is based on SB-E.

Wow this is very interesting for so many reasons. I mean it makes sense in that fact that the major upgrades and bullet points for Ivy were the process and the beefed up iGPU. Well the later doesn't matter for the -E platforms, so maybe it was just easier to take an SB-E core and shrink it...But at the same time does this mean IB-E won't have the same core and uncore improvements that the desktop version got and even the slight bump in IPC?

SB-E was botched cause of the chipset. Maybe IB-E was going to be too close to Haswell so, they're releasing the updated SB-E/respin as IB-E. I'm very curious as to whats going on. But dangit I want my 8cores already!

If such is the case, I imagine IB-E being the proper SB-E Intel wanted to release, with the full 8 cores active @ 22nm.

[Liam_G]

No not quite, I'm talking a 22 nm SB-E. SB-E is based on SB, IB-E is based on SB-E.

Hmmm, very interesting if this comes about, hope to hear more news on this platform soon. Maybe x89 or whatever it will be plus 22nm SB-E 8 core will be released at same time as Haswell next year. 8 Core 22nm SB-E would make sense if it was TDP keeping them from enabling all cores on 32nm. I'd like to see the 4 core 22nm SB-E as well to see if intel has improved their tri-gate transistor implementation, although haswell will probably be the first indicator there.

[lutjens]

Intel owns the performance market...what incentive is there for them to release Ivy Bridge-E? After Intel released the E5-2687W multiplier locked and has refused to give us an un-neutered CPU (multiplier unlocked, 8 Core, full cache, all QPI links enabled, top bin) at any price, I'm seriously questioning my aggressive computer upgrading strategy. It's very hard to throw big dollars at a company that works hard to intentionally screw you over because you want the best. You want a run-of-the-mill Ivy i7? No problem, here's your CPU based on our latest technology. You want Intel's best and are willing to pay big dollars for it? Tough. Here's your last generation CPU that's locked up tighter than Fort Knox that we've been selling to our buddies since the middle of last year. Now, since they've had their fill, we've decided to let you have one....now give us our $2000.

Stuff it up your cornhole Intel...

[BeepBeep2]

Nice necro guys

[Liam_G]

Nice necro guys

Haha, I admit I was fishing for some new info, certain people seem to know whats going on unofficially and aren't adverse to letting things slip I haven't kept up with the latest so thought I might go fishing looks like I'm going home hungry

[GoldenTiger]

Haha, I admit I was fishing for some new info, certain people seem to know whats going on unofficially and aren't adverse to letting things slip I haven't kept up with the latest so thought I might go fishing looks like I'm going home hungry

Hehe, ...

[Tenknics]

Agreed. Despite wanting an updated SB-E/Ivy-E REALLY BADLY, I have to admit Intel has no reason to release anything in the high-end market right now..or for awhile. It's kinda unfortunate that Intel needs real competition to show us whats really possible.


This whole news section is dead. I'm sure it has a lot to do with Windows 8 launching in the fall and a lot of companies are timing their new products with the launch of that.

[lutjens]

Agreed. Despite wanting an updated SB-E/Ivy-E REALLY BADLY, I have to admit Intel has no reason to release anything in the high-end market right now..or for awhile. It's kinda unfortunate that Intel needs real competition to show us whats really possible.


This whole news section is dead. I'm sure it has a lot to do with Windows 8 launching in the fall and a lot of companies are timing their new products with the launch of that.

Also, with Intel in firm command of the high-end market, having a 6 core "Extreme Edition" also gives Intel a handy, high margin dumping ground for all the Xeons that have a defective core or two, or that have a portion of the cache that's defective, but otherwise are a high bin chip. Simply deactivate the offending section of the chip and voila, much of the profit is saved. I'd also surmise that the 3930K also only exists for that reason...to make it more palatable to enthusiasts. They just disable an additonal 3MB of cache to differentiate the 3960X from the 3930K.

Which is fine...Intel needs to do what it needs to do to ensure maximum profit. What incenses me to no end is the castration of the very best chips...the E5-2687W. Absolutely ignorant move on Intel's part, one that effectively flips the bird to the highest spending enthusiasts, some of whom have given Intel huge $$$ over the years.

[Tenknics]

Also, with Intel in firm command of the high-end market, having a 6 core "Extreme Edition" also gives Intel a handy, high margin dumping ground for all the Xeons that have a defective core or two, or that have a portion of the cache that's defective, but otherwise are a high bin chip. Simply deactivate the offending section of the chip and voila, much of the profit is saved. I'd also surmise that the 3930K also only exists for that reason...to make it more palatable to enthusiasts. They just disable an additonal 3MB of cache to differentiate the 3960X from the 3930K.

Which is fine...Intel needs to do what it needs to do to ensure maximum profit. What incenses me to no end is the castration of the very best chips...the E5-2687W. Absolutely ignorant move on Intel's part, one that effectively flips the bird to the highest spending enthusiasts, some of whom have given Intel huge $$$ over the years.

its like Im reading a page out of Intel's 2012 playbook. This is spot on.

[Nikolasz]

Feels like pention 1 era.

[Brice MJ]

finally some new info on IB-EP/IB-EN:

Based on that information, which is about 2 - 3 months old, the Ivy Bridge EP and EN CPUs will have up to 10 cores. 10-core SKUs are going to have TDP as low as 70 Watt, and there will be also 95 Watt versions of these chips. 8-core -EP and -EN processors will have TDP as low as 60 Watt. Low-power quad- and hexa-core Xeons will come with 45 Watt TDP. Ivy Bridge-EP/-EN products should be available in late first half, or early second half 2013. Bear in mind that this information is somewhat old, and more recent roadmap doesn't show Ivy Bridge Xeons in the second quarter 2013.

http://www.cpu-world.com/news_2012/2...Xeon_CPUs.html

70-95w TDP 10-core/20-thread CPUs, coming in early H2/2013

[Aberration]

:O

Good

Lord

[Kain665]

Try 12c/24th

[lutjens]

its like Im reading a page out of Intel's 2012 playbook. This is spot on.

Thanks for the compliment...I'm just calling it as I see it and its nice to see others see the validity in it as well. The more I think about it though, the more I come to the conclusion that Intel doesn't want us enthusiasts buying E5-2687W's, which are still selling well in the server market. They'd rather us buy the defective chips that are the i7-3960X, ensuring a steady demand for their defective crap, which is probably why the E5-2687W is locked...to force overclockers to buy the i7-3960X. Only when Intel has more non-defective, high yielding 8 cores than they can sell as Xeons will they ever consider an unlocked 8 core processor. Even then, considering that even the 2GHz E5-2650 has a higher MSRP than the i7-3960X, I think they'd just remark any high yield 8 cores down to lower speed grade Xeons before they'd ever produce even an 8 core "Extreme" CPU. By locking out overclocking as they've done, they've largely dictated the CPUs that enthusiasts will consume, who are now wholly dependent on the i7-3960X and the i7-3930K for any serious overclocking on the LGA2011 platform. By selling all their 8 cores as Xeons, they've ensured that they get more money for any 8 core than they could at the $999 Extreme price point. By keeping all overclockable CPUs the 6 core version, they've also ensured a steady, forced enthusiast demand for their defective chips, keeping the margin up on them.

What Intel needs to get through their thick head is that some enthusiasts aren't willing to purchase a defective chip, no matter how they gussy it up and that by trying to ram crap down our throats, they are marginalizing and alienating high paying customers. These people will only tolerate it for so long before they begin to find new hobbies. Forcing your most loyal customers to buy crap, is not the way you keep them as customers.

[SabreWulf69]

Just as most speculation comes I shall add my simplistic opinion of this into the fray. Basing recent-ish release type speculation on recent events such as the release of the 2700K only months just before Ivy was released leads me to believe something similar could happen here as the release of 3970X has just come about... based on this, and it may as well be just as tangible evidence as anything else that has come along, I'm thinking maybe a Q4 2K12 - Q1 2K13 release of a 2011 socket new 'line' of CPU's to be released. Maybe with our guesstimations, we should have some fun and place some bets hahaha.

[Karak]

Hopefully there will be consumer 8 core chips before the end of this year. I really want to buy them, Intel just does not let me I was planning to buy 8 core Sandy Bridge E, since there were not released as desktop products, no cake for them >.<

Information about 8 core Sandy Bridge were circulating from around 2008 or 2009 with even some ES benchmarks, probably if AMD chips were more successful Intel would not done it the way it is now. If you check price/performance of _high_end_ segment, things were going downhill very fast since 2008 Nehalems...

[lutjens]

Hopefully there will be consumer 8 core chips before the end of this year. I really want to buy them, Intel just does not let me I was planning to buy 8 core Sandy Bridge E, since there were not released as desktop products, no cake for them >.<

I wouldn't hold my breath waiting for one, if I were you. If Intel released an 8 core desktop chip, they'd have to lower the prices on the defective junk that they're currently pawning off as a high-end CPU (the i7-3960X). There's no reason for them to do that when they can continue to sell these defects at $999 each and the 8 cores for much more than that as Xeons. The current situation is a bit like the P4/Celeron scenario that happened last decade. Except now the Xeon represents the P4 and the Celeron has become an Extreme Edition chip, being sold for $999.

This is what happens when an inherently ignorant company, with absolutely no concern whatsoever for the consumer, has no credible competition at all...

[Gorki]

This is what happens when an inherently ignorant company, with absolutely no concern whatsoever for the consumer, has no credible competition at all...

Actually market is bored for some time, Intel e-peen is on scene for some time like from i7 generation and prices topped out a long time back and it's still flooded by intel! Intel is really packing thing isn't it?
I'm actually cheering for AMD hoping it would spice thing just a little bit. Oh well...

[aussie-revhead]

Guys, any more news on Ivy-E ? I just heard they are very close to release and we will see them soon, is there any truth to this?

[paulbagz]

Heard from whar? :O

Sauces!

-PB

[aussie-revhead]

I am following a lead so I will report if I find anything, but I doubt I would be the one to bring breaking news to this forum, I would think many others would have leaked it before I heard of it. It's from a reviewer (one of the first to receive an SB-e sample last year too), updates will be posted when they become available