Ivy Bridge-E and EP Coming in 2012

[Edrick]

They say by end of year, which I take to mean 2012. Although I suppose it could mean at least until the calendar turns 2013. Either way, Haswell will be right around the corner.

http://vr-zone.com/articles/ivy-bri...--the-multi-socket-platform-heaven/15488.html

The dual processor Socket 2011 platform, the Xeon E5, is expected to receive the ten core 25 MB L3 (some sources mention 30 MB L3 but that doesn't seem likely as of now) cache Ivy Bridge EP part update, which should also bring around 15%-20% clock speed update. The 22 nm tri-gate process benefits for both power consumption and clock speed will help Xeon E5 solve the high-TDP problems that bothered it a little bit in the Sandy Bridge EP generation right now. In my mind, based on the power and performance gains assessment, I'd expect 2.4 GHz 10-core parts at the 95W TDP level, 3 GHz 10-core parts at the 130W TDP level, and 3.3 GHz 10-core parts at the 150W workstation SKU grade, all plus minus 100 MHz or so.​

And, oh yes, the TDP problem solution should also enable the single Socket 2011 systems based on X79 chipset to have full speed 8-core and 10-core chips in this iteration - good news for those of you holding, say, entry level Core i7-3820 quad core, since this would make a nice speed doubling upgrade without increasing the system power consumption.​

Again, this will be just a CPU update on the same chipset and board platform, likely to be matched with a Patsburg C600 chipset stepping update for those 'minor problems' like say 6 Gbps SAS, for instance. Otherwise, the stable platform should provide a smooth transition from 8-core Sandy Bridge EP to 10-core Ivy Bridge EP.​

 

[cytg111]

Either way, Haswell will be right around the corner.

- Yea, right!

 

[Arachnotronic]

I told you, my dear Edrick, IVB-E will come and it'll be a beast. Think 8-10C unlocked extreme edition CPUs. Think...Haswell 4C/8T killer.

 

[tynopik]

if they can get it out this year that's fine, but previous articles indicated it wouldn't be out till 2H 2013 . . . which is not fine

 

[Arachnotronic]

if they can get it out this year that's fine, but previous articles indicated it wouldn't be out till 2H 2013 . . . which is not fine

I guarantee you that's not going to be the case. People saw this roadmap: http://cdn.overclock.net/1/15/15954714_intelhaswelllaunchdate_dh_fx57.jpeg

And thought, "ZOMG IVB-E IS NOT EXPLICITLY LISTED!!!"

But notice the greater-than-or-equal-to sign? Intel's previous roadmaps had that too after the 990X for Q4 2011. That CPU was SNB-E.

So we'll probably get an 8C/16T IVB-E or an 8C/16T SNB-E. Either way, the enthusiast lineup will hold a significant advantage over the mainstream for multithreaded workloads.

 

[cytg111]

IF haswell lands in H1 2013, it will be because there's prospect of making much green from tablets and deriatives to outweight the green they're going to loose from sales on IVB. Seeing there is no immidiate competition I am going to bet that the first generation of Haswell products will be tablet/phone oriented socs first .. then mainstream H2 2014

 

[gevorg]

Unlike in IVB vs SB-E, Haswell will not have a die shrink advantage over IVB-E, so quad-core IVB-E should stay on top even after Haswell's release. Based on what has been released/leaked so far about Haswell, the main improvements come in power usage for mobile/small form factor systems and IGP speed boost, rather than IPC improvement.

 

[exar333]

Bring it on! Waiting to drop-in an IB-E 8 or 10 core.

 

[Edrick]

Unlike in IVB vs SB-E, Haswell will not have a die shrink advantage over IVB-E, so quad-core IVB-E should stay on top even after Haswell's release. Based on what has been released/leaked so far about Haswell, the main improvements come in power usage for mobile/small form factor systems and IGP speed boost, rather than IPC improvement.

I think you are very far off track with that statement. As many have already stated clearly on these forums, myself included, Haswell will bring huge gains with its new instructions that will put IB to shame.

 

[Edrick]

I told you, my dear Edrick, IVB-E will come and it'll be a beast. Think 8-10C unlocked extreme edition CPUs. Think...Haswell 4C/8T killer.

8C IB is very tempting. And I most likely will end up with one. But I am still much more excited about Haswell and its new instructions. And if IB can produce 8C and 10C CPUs, then Haswell can and should be able to do the same (Haswell-E). To think I may have to wait until Q4 2013 for that just depresses the hell out of me.

 

[gevorg]

I think you are very far off track with that statement. As many have already stated clearly on these forums, myself included, Haswell will bring huge gains with its new instructions that will put IB to shame.

It would take quite a bit of time, possibly years, for these instructions sets to trickle down to common software. Just look at current use of AVX set. But I hope you're right, and that more and more software developers would aggressively push out software that take advantage of latest CPU instructions.

 

[Edrick]

Just look at current use of AVX set. But I hope you're right, and that more and more software developers would aggressively push out software that take advantage of latest CPU instructions.

AVX was essentially a prerequisite for AVX2. By itself was not going to offer huge gains, but it laid the groundwork for AVX2. In a perfect world, they would have been released together, but that would have been a huge undertaking for even Intel. It was smart to break them up like this over multiple Tocks. So once completed and we have the entire set of instructions in place, I expect to see huge gains.

 

[Ferzerp]

More cores only helps with specific workloads. Faster cores help with every workload. Don't get all excited about 8-10 core processors unless you find yourself with extremely parallel workloads all the time that you're waiting on.

 

[IntelUser2000]

Unlike in IVB vs SB-E, Haswell will not have a die shrink advantage over IVB-E, so quad-core IVB-E should stay on top even after Haswell's release.

It may be just the opposite. Tocks can be seen as an architecture that's built for the process. Ivy Bridge is a mere shrink with architecture based on Sandy Bridge so it doesn't extract full potential of 22nm, but Haswell does.

AVX was essentially a prerequisite for AVX2. By itself was not going to offer huge gains, but it laid the groundwork for AVX2.

They put AVX for servers. It's important, just that desktop users overestimated the impact floating point had on their usage. Boosting integer performance is the big thing, however that's not what AVX offered.

 

[GammaLaser]

More cores only helps with specific workloads. Faster cores help with every workload. Don't get all excited about 8-10 core processors unless you find yourself with extremely parallel workloads all the time that you're waiting on.

Maybe the increased die size from the extra cores over 4-core Ivy would help with the heat density issues the early OC reviews have had?

 

[nenforcer]

I'm still just waiting for a Socket 2011 motherboard with a native USB 3 controller. Is that too much to ask?

 

[bunnyfubbles]

well this is excellent news considering I just upgraded to s2011

 

[Edrick]

I'm still just waiting for a Socket 2011 motherboard with a native USB 3 controller. Is that too much to ask?

Who cares? Does it matter in any way, shape, or form if the USB3 is native or not?

 

[Arachnotronic]

8C IB is very tempting. And I most likely will end up with one. But I am still much more excited about Haswell and its new instructions. And if IB can produce 8C and 10C CPUs, then Haswell can and should be able to do the same (Haswell-E). To think I may have to wait until Q4 2013 for that just depresses the hell out of me.

Yeah, but Intel seems to be building even more features into the system architecture of their -E chips, so the wait will be worth it.

 

[exar333]

I'm still just waiting for a Socket 2011 motherboard with a native USB 3 controller. Is that too much to ask?

Who cares? Buy a USB 3.0 card for $20 and call it a day. If you are 'dead-set' on USB 3.0, S2011 probably isn't your cup of tea anyways...silly comment IMHO.

 

[Edrick]

Who cares? Buy a USB 3.0 card for $20 and call it a day. If you are 'dead-set' on USB 3.0, S2011 probably isn't your cup of tea anyways...silly comment IMHO.

No need. Every X79 MB comes with USB 3.0. Heck, every SB MB comes with USB 3.0.

 

[Conscript]

Who cares? Buy a USB 3.0 card for $20 and call it a day. If you are 'dead-set' on USB 3.0, S2011 probably isn't your cup of tea anyways...silly comment IMHO.

Still, now that the 7 series boards have it, it would be nice to get a premium chipset when you pay a premium price. Agreed, it's not a biggie...but it's there. Me, I'd prefer to see thunderbolt or whatever they're going to call it on the PC.