Intel Preparing Ivy Bridge-E for September Release

3 SKUs, early September release. That is what the rumor mill has to say about Intel’s upcoming high-performance desktop parts code named Ivy Bridge-E.

Ivy Bridge-E - Image courtesy Wccftech.com

Ivy Bridge-E (Courtesy Wccftech.com)

A supposedly leaked slide puts the launch of three processor parts in the range of 4-11 of September. There’s three parts that are to be released on that date. The top dog is i7-4960X, a six core (12 thread) part with a clock speed of 3.6 GHz and turbo up to 4 GHz. The runner up is i7-4930K. This one should also have six cores (12 threads) but with a slightly lower clock speed of 3.4 GHz and turbo up to 3.9 GHz. The final part that is expected is called i7-4820K and should come to the party with four cores (8 threads) with a base clock of 3.7GHz and turbo up to 3.9GHz. These parts will feature 15, 12, and 10MB of cache respectively, and support for quad-channel DDR3 memory up to 1866MHz.

The i7-4960X is expected to sell in the $999-1$,099 price range, as is the tradition at Intel for the top performance desktop processor. The wildcard here is the presence of a quad core part, which could end up priced closer to Haswell SKUs, but still be paired with the high-performance X79 platform, socket 2011 and quad-channel memory. Note the K moniker, which specifies this CPU is unlocked for added overclocking pleasure. All three CPUs are expected to fit in the 130W thermal envelope.

Intel is not expected to release a new chipset for these parts, relying on the trusted X79 for the time being, although it is not impossible X89 may arrive further down the road.

Read the full report on AnandTech

- dostov

Tags: , , ,

36 Comments:

xsuperbgx's Avatar
I am wondering if the 4820k, with quad channel memory and the adjustable straps will match haswell on s1150? With so many dud haswell retail CPU's will people be happier with IB-e?
EarthDog's Avatar
Quad channel ram means little to nothing though outside of bandwidth. It doesnt make a difference in benchmarking it seems and also not with any other applications...
xsuperbgx's Avatar
You don't think IB-e will have any 2d benching advantage over ib on 1155?
EarthDog's Avatar
Nope. sb-e didn't have any advantages over sb. They topped out lower and any extra cache didn't seem to matter on the hex
ivanlabrie's Avatar
3d on ln2 might be nice...quad 8970+49/30k-60x at 6ghz
xsuperbgx's Avatar
How about the heat spreader? Will that be soldered and if so, will that help performance at all, or just temps?
EarthDog's Avatar
Who knows... And temps only. I imagine its just like IB though.

As far as 3d, in 11 and vantage, sure, otherwise, nope as others dont respond to cores.
ivanlabrie's Avatar
Exactly...should be fun for those.
Sabishiihito's Avatar
Quad channel DDR3-2800? One can dream
ivanlabrie's Avatar
We're talking Ivy-E still...how many Ivies did 2800 with 4 sticks?

I guess you would know better than me anyway.
awktane's Avatar
I hope they get smart! If not well I've been itching to delid a processor for a while now...
Hippogriff's Avatar
If it's true, I am going to wait. I've been working with a crappy laptop for three years, so I can wait one more month and see how the 4820k performs compared to the 4770k.
EarthDog's Avatar
Its likely going to be a bit slower... That part is Ivybridge essentially at 4c/4t.
Hippogriff's Avatar
What about OC potential? My friend has a 3960x @4.9 with a H100i, so if IB-E is as good as SB-E...
EarthDog's Avatar
I'd imagine the same potential as Ivybridge...for that part, but not certain.
Sabishiihito's Avatar
My 3570K would happily do 4x4GB DDR3-2800. I almost hate to get rid of it, but I've sold my MVE so it has no home, and I'm obsessed with pushing RAM to 3000+ on Haswell
xsuperbgx's Avatar
That is a good IMC for ib. My 3570k would only do 2400.
ivanlabrie's Avatar
<--- Shiranui Gen-An has some nice kits and chips...My 3770ks were ****ty both of em
Rob94hawk's Avatar
So does this mean 1150 is a dead socket?
ATMINSIDE's Avatar
No... It'll replace SandyBridge-E processor lineup, which runs on the 2011 socket. IvyBridge-E will also run on the 2011 socket.

The 1150 socket released about 2 weeks ago.
awktane's Avatar
Errm no. Socket 1150 likely has at least one more round of CPUs upcoming. Ivy Bridge E replaces the enthusiast line of sandy bridge E processors.

Edit: ATM beat me to it
Rob94hawk's Avatar
I know it just came out. I have the 4770k/1150. I was being sarcastic in a way. Gone are the days where there are 20 different cpu's for one particular socket. Maybe if I'm lucky Intel will make a 14nm Broadwell for 1150....
ATMINSIDE's Avatar
Right now, Intel's track record is to have two generations of CPUs on one socket. So, most likely, the next desktop chip will also be on 1150.

The SB-E and IB-E are on a different socket from the desktop chips because of RAM/PCIe lanes.
ivanlabrie's Avatar
Oh, look: http://translate.google.com/translat...B%2592&act=url

Them biotches could indeed solder 22nm parts and where intentionally hampering the consumer line performance...

Also: http://www.xbitlabs.com/news/cpu/dis..._Offering.html
ATMINSIDE's Avatar
Solder O.o

That's some WIN.
ivanlabrie's Avatar
Yup, for regular water/air overclockers, big one.
(some) Delidded Ivy chips could run 5ghz+ as daily oc's no problem with temperatures in check...If they had soldered those, well, you know how it goes
Artificial market segmentation at its finest.
ATMINSIDE's Avatar
Yeah, if those were soldered, you'd see SB-like clocks everywhere again.
ivanlabrie's Avatar
Yup...this pisses me off, and yeah, many benchers will say there's no real problem with that since you can still overclock using ln2 (and in fact, delidding normally hurts ln2 oc's) but still, this is either clear manipulation of the consumer markets, deliverately hampering mainstream chips OR maybe they couldn't solder 22nm parts till now (which I have a hard time believing)
trueblack's Avatar
4820k ???? are we sure not just 4820, without the K?
ivanlabrie's Avatar
Even the quads are unlocked now it seems...
dudleycpa's Avatar
@trueblack - we sure: http://www.cpu-world.com/CPUs/Core_i...0i7-4820K.html

I don't know if it is deliverable, premeditated, clear manipulation, of the prosumer market. It seem that Moore's Law (what was it 12 months). Intel I believe quoted 24 months. Is broken.

I don't want to bore anyone with boring/dated history lessons (we're all adults here) Look up Alan Turing - great read!
ivanlabrie's Avatar
Same kind of silicon as Ivy bridge (like SB to SB-E) yet now they do bother soldering them?
Doesn't sound reasonable...
trueblack's Avatar
this makes the 4820k the best buy easily.

having less core means can use higher voltage for faster base clock.
might be the best one to OC with.


that's why I wonder is it really unlocked... this might cannibalize the market for 4930k
EarthDog's Avatar
IMO, I highly doubt it would cannibalize the market for 4930K. I mean, think about it. Nobody should be buying a quad on this platform really as it offers no benefits over Haswell at this time (outside of additional PCIe lanes for tri/quad CFx/SLI). In fact, it is likely it will be slower like IB is to Haswell currently. I mean quad channel memory bandwidth is shown to be a non factor, so why would one go quad core CPU over Haswell there? To pay more for the entire system? I can't say I would understand such a move personally.
trueblack's Avatar
quite true.


Intel should have launched IB-e FIRST.

then launch Haswell..


the reverse order does feel like it makes no sense.
ivanlabrie's Avatar
I'd go soldered Ivy hexa...but only since I'd like to fold with it while not benching.
Hmmm, I'll upgrade by the time it launches so it's a posibility.
Leave a Comment