Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Given AMD still hasn't fully released their desktop 28nm "Southern Islands" ( 77xx , 78xx , 79xx ) line up, it would not be surprising if mobile 28nm 7730M , 7750M, 7950M to arrive late in April or early May rather than when the Ivy bridge desktop chips arrive. All the more so, if Intel isn't dropping the mobile Ivy's until May/June ( i.e., there is lower demand for PCI-e v3.0 mobile offerings till then. )

As for being "behind" the Core i7 Ivy's. Not much if any. In the line up above the top end is the

i7-3770K: 4 cores, 8 threads, 8 MB of cache, TDP of 77W, 3500 MHz with Turbo mode:3900 MHz,

Whereas the Xeon 1620 comes in at

Xeon E5-1620 4 8 3.6 GHz 10 MB 130 Watt

The Xeon has an extra 50W of power , 100MHz higher base, 2MB more cache, and 2 more memory controllers to throw at the "tick behind" problem. The i7 will certainly have better integrated graphics. :) It will also certainly be "greener".

I'd guess that Apple with go with the 3700 which is a 200MHz gap and only in the "deluxe" build to order model.
 
Given AMD still hasn't fully released their desktop 28nm "Southern Islands" ( 77xx , 78xx , 79xx ) line up, it would not be surprising if mobile 28nm 7730M , 7750M, 7950M to arrive late in April or early May rather than when the Ivy bridge desktop chips arrive. All the more so, if Intel isn't dropping the mobile Ivy's until May/June ( i.e., there is lower demand for PCI-e v3.0 mobile offerings till then. )

At least according to the roadmaps, mobile chips will be released simultaneously with the desktop chips. I.e:

Quad core SKUs = April
Dual core SKUs = May

The Xeon has an extra 50W of power , 100MHz higher base, 2MB more cache, and 2 more memory controllers to throw at the "tick behind" problem. The i7 will certainly have better integrated graphics. :) It will also certainly be "greener".

Ivy Bridge brings some architectural improvements too (well, Intel claims so... :rolleyes:) so in terms of raw CPU performance, i7-3700(K) will probably outperform i7-3820/E5-1620.

I'd guess that Apple with go with the 3700 which is a 200MHz gap and only in the "deluxe" build to order model.

i7-3700 is most likely 3.5GHz as well (there's a typo in the roadmap). If you look at the frequencies, all except low-end SKUs offer a maximum Turbo of four multipliers. Yet the Turbo of i7-3700 is 3.9GHz, which would translate to five multipliers. That doesn't make much sense.
 
Why is Intel scared of 4GHz? Nothing will pass this barrier right now at stock or turbo. Pretty soon we will have 3.8GHz chips that turbo to 3.9GHz;)
It's not like the speed of light or anything.
Also pretty excited as a Mac owner for 7950 even though the PC camp is complaining that it only competes with GTX580 after 1 years time and the price is too high. Apple and Nvidia are still apparently pissed at each other so anything AMD can do to compete with lower power requirements is promising for the Mac users. Also Nvidia's top performers do not fit in with the Apple view of things. They drain way too much power for the performance compared to AMD. They may always take the crown but they are not very elegant doing it.
 
Last edited:
Why is Intel scared of 4GHz? Nothing will pass this barrier right now at stock or turbo. Pretty soon we will have 3.8GHz chips that turbo to 3.9GHz;)
It's not like the speed of light or anything.

It seems that power consumption will go up substantially when breaking the ~4GHz barrier.

graph3.jpg


core-2-duo-e7600-1.png


core-i7-950-1.png


It doesn't make sense to break that barrier "just for numbers" if the real world gain is negligible. What's causing that is another question and which I unfortunately cannot answer.
 
Why is Intel scared of 4GHz? Nothing will pass this barrier right now at stock or turbo. Pretty soon we will have 3.8GHz chips that turbo to 3.9GHz;)
It's not like the speed of light or anything.

Looking at the charts (from another poster), it looks like its the thermal management challenge ... and more to the point, the challenge of how to do the thermal management cheaply, through aircooled designs instead of water. DARPA does have a "Thermal Management" program ongoing right now where they're looking at some of this stuff, but having just personally attended their 'showcase' meeting back in December, I can assure you that the solutions are only "less cheap" in comparison to the even more exotic solutions that preceeded them.

In addition, I'm sure that smart companies like Intel are looking at their technology roadmaps along with the state of the economy ... if I were them, I'd hold back on any 4GHz labels until Enterprise is revving up and in a stronger buying mood where they're more concerned about pumping up performance than its initial price tag. As such, if 4GHz were to appear today, there's not a huge demand...and by the time that the buyers do finally go on a spending spree, 4GHz would have been out for awhile, so the buyers would be expecting that it should by then also be cheap.

As such, don't be surprised if 3.933GHz and 3.966 Turbo modes show up too :)



-hh
 
Looking at the charts (from another poster), it looks like its the thermal management challenge ... and more to the point, the challenge of how to do the thermal management cheaply,

4ghz+ has been relatively easy to achieve on PC's for years now, even using stock cooling to some extent. Problem is PC stock cooling is rather noisy and 4ghz does run hot and needs extra power to run reliably.
 
Maybe it IS like the speed of light:p

Well, it looks like from Hellhammer's post that for most processors the 4 GHz "barrier" is still in the linear range. It would be interesting to see the curve plotted out further to see if the slope changes again to something a little closer to infinate, as is the case with F vs speed near the speed of light. The problem is really just where that "inflection point" (I know its not really an inflection point) is going from a basically horizontal line to a linear line of substantially greater slope. And this seems to occur somewhere in the 3.8 range. So, that's probably why we see turbo boost that only go just slightly into this range.
 
Why is everyone explaining the charts to me? Slightly insulting. They speak for themselves. Light speed was a joke on diminishing returns and e=mc2.
 
ummm, your computer (a 13" MBP) is basically a portable mini

Yes it is. With a screen. And a keyboard. And a trackpad. And a battery. And a DVD drive. And an SSD. And I got 15% off it with the academic discount.

Not seeing your point, I'm afraid.
 
So apparently Sandy Bridge Xeon cores should be out March 6th? Does this mean Mac Pro next month hopefully?
 
So apparently Sandy Bridge Xeon cores should be out March 6th? Does this mean Mac Pro next month hopefully?

I think at very best, maybe an announcement. You never know, but I think April-May is more likely unless something more drastic is in the cards.
 
Perhaps Apple will announce the Mac Pro a month before its availability, just like they did with the thunderbolt display. I could imagine this happening since the waiting period has been longer than the average.
 
I can't even recall the last time the Mac Pro was even mentioned by Apple in any regards. It's lonely out here . . . . . :(
 
OK all this talk about dropping the Desktop " BECAUSE OF PROFITS" is one of the most ludicrous statement i have ever heard.

Think about it. Who would DROP A PRODUCT so they can make MORE MONEY. One of the most important think you need in order to make MORE PROFITS is........wait for it....., yep MORE PRODUCTS. !!!!! Not less products.

So lets pretend I have a tiny little store and I sell candy and potato chips and I decide I want to make a whole lot of money, yes even MORE money, so my great idea is to eliminate some of my product, see economics 101.

There is very little R&D involved in building a Desktop over what they already have. It's just another case, it's that simple. Oh wait, and they already have the case in fact they have warehouses full of cases.

Apple doesn't warehouse anything. Everything is built on time. They don't need to buy a million chips and get stuck with them, they order what they are going to sell.

The biggest reason Apple wouldn't dump the Desktop is people would migrate to the PC desktop and then migrate all their software and once they do that they would stop buying ANYTHING Apple.

Do you produce both the candy and the chips as well?
 
Today is day 575 without Mac Pro refresh. I feel we will see an update in the next 2-3 weeks for sure :):D:)
 
I said 600 because I think the Mac Pro has never been updated after so long? So, 600 days would be a decent number of days to take a decision regarding what to do, if I should find an alternative (Pc etc)..

Everyone who says I should buy a refurb. and update it or just a get a Mac today is just denying that for Apple the Mac Pro doesn't have the importance it had years ago. I'm not going to buy an old machine and update it because it will still cost a lot and do you do that with cars? No.

Anyway, we are all in the same situation here..waiting :)

Ivy Bridge is suspected to be announced March 6th (1 day before the iPad 3). It will then only be a matter of another month or two before the Mac Pro gets refreshed. June would be the latest that it would go before being refreshed IMHO. I'm runing a 2006 Mac pro, and other than a HD failure last night, i haven't had any problems with it. Just hold off a little longer.
 
Ivy Bridge is suspected to be announced March 6th (1 day before the iPad 3). It will then only be a matter of another month or two before the Mac Pro gets refreshed. June would be the latest that it would go before being refreshed IMHO. I'm runing a 2006 Mac pro, and other than a HD failure last night, i haven't had any problems with it. Just hold off a little longer.

I think it's Sandy Bridge that will be released March 6th. Apparently Ivy Bridge Xeons won't be out till late 2012/early 2013 (someone correct me if I'm wrong)
 
No, Ivy Bridge "consumer" i7 will be released AND Sandy Bridge based Xeon's. Ivy based Xeon's are another year out most likely. The Xeon's release later after more rigorous testing. Or Intel likes to say that is why:) So for Mac Pro Ivy Bridge release means nothing unless they change directions and put i7's in the Mac Pro's.
 
I think it's Sandy Bridge that will be released March 6th. Apparently Ivy Bridge Xeons won't be out till late 2012/early 2013 (someone correct me if I'm wrong)

Like many of you, I have been waiting for, what seems like, an eternity. Hopefully Apple will announce something in the beginning of March month.
 
Ivy Bridge or BUST!

I think it's Sandy Bridge that will be released March 6th. Apparently Ivy Bridge Xeons won't be out till late 2012/early 2013 (someone correct me if I'm wrong)

Ivy Bridge Xeons are supposed to hit this Spring. So when they do, the Sandybridge Xeons will be obsolete. I'd rather have new tech in an updated Mac Pro. Apple has dogged this machine with obsolete video cards practically since its inception. We need not add CPU's to the roster.

In the second quarter of 2012, Intel will release 11 Ivy Bridge-H2 series CPUs including Xeon E3-1290v2, priced between US$189-884 and seven more processors including Xeon E5-2470, priced between US$192-1,440.
via Digitimes
 
Ivy Bridge Xeons are supposed to hit this Spring.

Ivy Bridge Xeon E3's will hit this Spring.

http://www.cpu-world.com/news_2012/2012020701_Prices_of_Xeon_E3-1200_v2_CPUs.html

This is not particularly surprising since Sandy Bridge Xeon E3's hit last Spring.

http://www.cpu-world.com/news_2011/2011040301_Intel_introduces_Xeon_E3-1200_CPUs.html


So when they do, the Sandybridge Xeons will be obsolete.

New Xeon E3's do not in any way obsolete New Xeon E5's. Even between generations.

I'd rather have new tech in an updated Mac Pro.

You'd like to be capped at 4 cores and only 20 PCI-e lanes ?
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.