Thursday, November 1st 2012
Top Intel Ivy Bridge-E Core Processors To Still Pack Six Cores
Intel's 2011-launched Core i7 "Sandy Bridge-E" HEDT platform is based on a 32 nm silicon that's common with Xeon E5 series processors. While the silicon physically packs eight CPU cores and 20 MB last-level cache (LLC, or L3 cache), client Core i7 processors are configured with only a maximum of six cores, and up to 15 MB L3 cache. According to a MyDrivers.com report, the maximum core count won't change with next-generation 22 nm Ivy Bridge-E Core i7 processors.
Ivy Bridge-E will be an upscale of Ivy Bridge. Similar to Sandy Bridge-E, the silicon will feature up to eight cores and 20 MB L3 cache. In its Core i7 avatar, however, the chip will be configured with no more than six cores, and no more than 15 MB L3 cache. The new chip will introduce IPC improvements, PCI-Express Gen 3.0 certified root complex (one which NVIDIA will approve of), higher CPU core clock speeds, and support for faster memory.
TDP could be the only reason Intel isn't willing to unlock cores 7 and 8 on client processors. Eight core, 20 MB LLC-laden Xeon E5 models based on today's 32 nm silicon, with 130W TDP, barely manage to scrape the 3.00 GHz mark. Given that, the prospects for Ivy Bridge-E client CPUs to run with all cores and LLC enabled, and yet deliver higher clock speeds than predecessors were always going to be low.
Intel Core i7 "Ivy Bridge-E" HEDT processors are compatible with existing socket LGA2011 motherboards (subject to BIOS update), and are slated for Q3-2013.
Source:
MyDrivers
Ivy Bridge-E will be an upscale of Ivy Bridge. Similar to Sandy Bridge-E, the silicon will feature up to eight cores and 20 MB L3 cache. In its Core i7 avatar, however, the chip will be configured with no more than six cores, and no more than 15 MB L3 cache. The new chip will introduce IPC improvements, PCI-Express Gen 3.0 certified root complex (one which NVIDIA will approve of), higher CPU core clock speeds, and support for faster memory.
TDP could be the only reason Intel isn't willing to unlock cores 7 and 8 on client processors. Eight core, 20 MB LLC-laden Xeon E5 models based on today's 32 nm silicon, with 130W TDP, barely manage to scrape the 3.00 GHz mark. Given that, the prospects for Ivy Bridge-E client CPUs to run with all cores and LLC enabled, and yet deliver higher clock speeds than predecessors were always going to be low.
Intel Core i7 "Ivy Bridge-E" HEDT processors are compatible with existing socket LGA2011 motherboards (subject to BIOS update), and are slated for Q3-2013.
111 Comments on Top Intel Ivy Bridge-E Core Processors To Still Pack Six Cores
In fact, in all the avx tests I've seen, even my chip beats the AMDs without the need for AVX. But I haven't seen my chip in those AMD test with encoding. Those were pure AVX synthetics.
I can't find the tests you are referring to.
www.lostcircuits.com/mambo//index.php?option=com_content&task=view&id=103&Itemid=1&limit=1&limitstart=11
Keep in mind that Intel's chips with HyperThreading only use unused CPU resources where AMD has dedicated extra hardware to run these threads. In turn AMD CPUs will have a more consistent amount of performance per thread, where with Intel, as soon as you start hitting those HT threads, you're performance goes down very quickly (even more so if you're doing a lot of calculations that are the same, so shared resources are now in high demand.) So you will see near linear scaling with AMD's CPUs but with Intel, you only have linear scaling on the real cores, once you hit HT threads, your performance per core starts taking a dive.
AMD has a better multi-core setup where Intel has better single threaded performance with just enough multi-threading bells and whistles to keep up with AMD's multi-core strategy. Consider for a moment how many more cores (or modules if you will,) that AMD could fit on to a CPU once they move away from the 32nm process and start producing smaller CPU circuitry. The size of a "Module" is only something like 26% bigger to add a second thread in comparison to the Phenom II's CPU core. Before you know it, CPUs won't get smaller and optimization of CPU resources will become a lot more important than it is right now and I think AMD realizes that.
AMD has nothing out now that can even beat Gulftown when it comes to encoding.
And that is without avx.
Wow, Are people really comparing a $500(3930k) CPU to a $200(FX8150). They aren't even CPUs built to compete with each other. the FX8150 was built to compete with the 2600k lol!
Call it cherry picking call it whatever you want. Point is at the price point an 8350 is a very good chip. Be able to compete with a $500-1000 chip in any application should scream foul to you Intel owners. If I am going to be paying double to quadruple the price for a processor it had better perform the best in all applications hands down. Back when Wile E got his 980X it was that way there was not a single benchmark or actual usage of the chip that it was beaten in. That is not the case right now the simple fact is AMD competes on occasion with Intel's LGA2011 offerings that is great for the consumer and should breed competition from Intel.
:shadedshu
Seriously;
1. resize
2. save some graphs for another post, because it sounds like you'll need them if Steamroller doesn't eventuate in 2013. (See!, links! much easier to scroll past)
Besides, Gulftown is still faster for the encoding I do. AMD would be a downgrade.
Which is why I'm unhappy that Intel is not releasing an unlocked 8 core. The 3960 is the fastest at the tasks I perform, but not enough of a speed boost to justify the purchase. The IB-E is likely to be no different. At $1000+ for my CPU upgrades, I expect a significant increase before I make said upgrades.
All this tells me is that my 980X will still be relevant for some time to come. I can worry about storage and video cards instead. lol
we can say with full confident that amd has the upper hands in multithread, im not necessarily saying they perform better as far as the chip goes today, but from a design perspective they definitely knocked intels HT down, and with steamroller having dedicated decode units scaling will only get better and any slight single thread improvements in that power envelope will only make the multithread performance skyrocket. and one thing also i believe amd has a upper hand in is power gating, the power states on amd processors aswell as clock speed is much more dynamic than intel ivy or sandy, thats why ive had a few friends buy intel laptops and realizing how slow of an experience they had due to the mediocre power states. maximum peak performance isnt everything
so while you may be disapointed, piledriver already delivered a leap forward in multithread, now if steamroller addresses the single thread and get a good 20-30% ipc improvement + the 10-20% better scaling then you are talking a good 50%+ better multithread performance. intel on the other hand with their architecture can only go on to improve their single thread which is pretty complex considering where they are now, and with haswell mostly being an amd like move of powergating everything to get the most out of the chip
and before I wrap this up i will end with this
www.phoronix.com/scan.php?page=article&item=amd_fx8350_visherabdver2&num=10
to truly see what the architecture is capable of try looking at it in an environment which optimizes it properly, here under linux is much closer to the 3770k in performance and leaves the i5s in shame
Anyway, seeing how the practice of disabling cores is nothing new and the energy saving thing going on in the world this isn't very surprising, or even shocking, to me.
Related question (that might have been asked already): Can you run those eight core Xeons on desktop boards?
Current offerings just don't up the performance enough for me to consider.
He means that AMD hasn't been able to produce a serious compedetor to Intel CPUs, so Intel doesn't have much reason to add more cores or really have a push to make their CPUs faster because their only competitor that means anything is AMD. I think we will see this change over the course of the next couple years.