Socketed LGA1150 Intel Broadwell-K to be released in Q4 2014

[FlanK3r]

The main part of 95W TDP is iGPU load. Without this could be around 75W maybe. And still, few Wats up or down, doesnt matter (its few rolls more or less in month, no expensive).

 

[FlanK3r]

Its a platform refresh. Not a CPU refresh. If lucky we just get a speedbin.

So something as +100 MHz in clock and maybe few fix of bugs? I bought until today 2 chips 4770K. One avarage and from friend one very good for LN2. So I skip refresh maybe....

 

[mikk]

You say the TDP isn't specified, yet it's specified in the slides?

It is not specified. 95W is just a placeholder.

If you say the TDP is not for Broadwell, then I guess you mean it's for Haswell Refresh? So the Haswell Refresh CPUs would not be possible to use on existing 84 W TDP Haswell motherboards. I.e. the 95 W TDP motherboards would only be needed for the Haswell Refresh CPUs sold in the short timeframe between the release of Haswell Refresh and Broadwell-K?

Once again, TDP is meaningless unless Intel confirmed it.

 

[Fjodor2001]

There is a note you admit not knowing what is.

Which you also admit you don't know what it means either.

 

[Fjodor2001]

It is not specified. 95W is just a placeholder.

Any source stating that it's just a placeholder?

Once again, TDP is meaningless unless Intel confirmed it.

It's in the slides by produced by Intel.

Sure, it's not 100% definite, but it's the best info we have so far. But if you want to be certain we have to wait until it's actually released and verified. No point discussing anything until then...

 

[mikk]

Haswell:
http://www.3dcenter.org/dateien/abbildungen/Intel-Roadmap-Haswell-Slide14.png
http://cdn4.wccftech.com/wp-content/uploads/2012/08/7s.jpg

Ivy Bridge:
http://www.techpowerup.com/img/11-04-12/67a.jpg
http://pics.computerbase.de/4/3/5/5/7/6.png

Do you see this? 95W!!! In the end it came with a TDP of up to 77W and 84W. There is a footnote beside 95W in this Broadwell-K Roadmap. Most likely the same note as usual from Intel. One example from the current mobile Roadmap. There is noted that TDP will be communitcted later.

 

[jpiniero]

I think that slide was before the delay. So the i5 BGA models will be released in 4Q 2014, but that might be it.

 

[Aikouka]

I thought the FIVR being on die was to allow more finite control of voltage? So, wouldn't moving it off die hamper the power reductions gained in Haswell?

 

[magomago]

I don't get this whole thing about upgrade path; Intel for the longest time lets a socket last 1-2 generations. I can't place an Ivy in my Sandy, and I can't place a Haswell in an Ivy. What upgrade path did I really have, an extra 100Mhz on a refresh 6 months to 1 year later??

 

[Ajay]

I thought the FIVR being on die was to allow more finite control of voltage? So, wouldn't moving it off die hamper the power reductions gained in Haswell?

Yes, more fine grained control of various voltage domains. It is a bit of a hot spot, and hence suspect to some as being cause for some clock limitation - but I don't think anyone (except Intel) knows that for sure. In any case, FIVR is staying as Intel continues the move away from strictly high performance processors towards more balanced designs.

 

[ShintaiDK]

I thought the FIVR being on die was to allow more finite control of voltage? So, wouldn't moving it off die hamper the power reductions gained in Haswell?

Its not moving away, it stays.

 

[mikk]

I think that slide was before the delay. So the i5 BGA models will be released in 4Q 2014, but that might be it.

Delay is already in there. Looks like Q3 for i7 and i5 models and Q4 for i3 models. This is in line with Intels statement that they expect production start in Q1.

 

[Fjodor2001]

Do you see this? 95W!!! In the end it came with a TDP of up to 77W and 84W. There is a footnote beside 95W in this Broadwell-K Roadmap. Most likely the same note as usual from Intel. One example from the current mobile Roadmap. There is noted that TDP will be communitcted later.

There are other slides correctly specifying 77 W for IB and 84 W for H. So it really doesn't prove anything. The 95 W TDP specified for B in the OP can be incorrect, but it can also be correct. We just don't know for sure yet.

Regarding Note 2 in the inline slide in your previous comment saying "Exact Skus/TDP for Broadwell will be communicated in a future roadmap publication", did you notice that slide did not specify any TDP at all? So it could just as well be that they are referring to the "future roadmap" slide now published in the OP, where TDPs now are actually specified.

 

[dahorns]

There are other slides correctly specifying 77 W for IB and 84 W for H. So it really doesn't prove anything. The 95 W TDP specified for B in the OP can be incorrect, but it can also be correct. We just don't know for sure yet.

Given that Intel and the industry as a whole are currently focused on power efficiency and mobility, we have fairly convincing evidence that Intel will not be releasing a 95 W TDP desktop part; i.e. the road map is most likely incorrect.

We obviously can't know for certain, this early on it is doubtful that Intel can either. Of course, you also have the thorny epistemological problem of whether we can truly know anything.

 

[Fjodor2001]

Given that Intel and the industry as a whole are currently focused on power efficiency and mobility, we have fairly convincing evidence that Intel will not be releasing a 95 W TDP desktop part; i.e. the road map is most likely incorrect.

Yes, but primarily for the Mobile segment. Remember that Intel increased the TDP from 77 W to 84 W going from Ivy Bridge -> Haswell for the Desktop CPUs.

 

[ShintaiDK]

Yes, but primarily for the Mobile segment. Remember that Intel increased the TDP from 77 W to 84 W going from Ivy Bridge -> Haswell for the Desktop CPUs.

But the platform power went down. FIVR was the reason.

 

[Fjodor2001]

But the platform power went down. FIVR was the reason.

No, the max load power consumption did not go down with Haswell. See e.g.:

http://www.anandtech.com/show/7003/the-haswell-review-intel-core-i74770k-i54560k-tested/2

3770K: 101W
4770K: 113 W

 

[Fjodor2001]

Sure, Intel can make a 65 W 14 nm node shrink "copy" of the 4770K with some more & redesigned EUs and 30% increased iGPU performance perhaps.

But really, most people sitting on a SB 2500K or later has no reason to upgrade to that. If they want to give those people a reason to upgrade they'll have to provide something more than that, like 6 CPU cores and/or a seriously beefed up iGPU / eDRAM. And such a change would likely require around 95 W TDP as in the OP slides, despite the node shrink to 14 nm.

Then they can have low power BGA Broadwell CPUs for the mainstream AIO PC desktop PCs and similar.

Since the desktop plans for Broadwell has shifted back and forth from LGA, to BGA only, and now back to LGA (and BGA), it looks like Intel has been reconsidering their plans a couple of times. So I guess anything is possible.

 

[ShintaiDK]

No, the max load power consumption did not go down with Haswell. See e.g.:

http://www.anandtech.com/show/7003/the-haswell-review-intel-core-i74770k-i54560k-tested/2

3770K: 101W
4770K: 113 W

A fine example of comparing apples and oranges.
Let me show you another one:

http://www.hardwarecanucks.com/foru...ntel-haswell-i7-4770k-i5-4670k-review-25.html

3770K: 113W
4770K: 101W

Now try figure out why.

Platform power was reduced.

 

[Phynaz]

No, the max load power consumption did not go down with Haswell. See e.g.:

http://www.anandtech.com/show/7003/the-haswell-review-intel-core-i74770k-i54560k-tested/2

3770K: 101W
4770K: 113 W

From your link:
"Here I’m showing an 11.8% increase in power consumption, and in this particular test the Core i7-4770K is 13% faster than the i7-3770K. Power consumption goes up, but so does performance per watt."

Noticed you skipped the idle power consumption numbers too.

 

[ChronoReverse]

Besides, the FIVR is what allows the Haswell to idle really well (as well as partial loads).

 

[Techhog]

FVIR won't be removed because Intel doesn't give a crap about thermals. If they did, they'd just go back to solder.

 

[Phynaz]

You realize FVIR reduces system thermals, right?

 

[Techhog]

You realize FVIR reduces system thermals, right?

I mean CPU thermals obviously. And they only care about system thermals because of laptops and tablets.

 

[ShintaiDK]

I mean CPU thermals obviously. And they only care about system thermals because of laptops and tablets.

Desktop users care too. Just not the microscopic niche you belong to.