Intel 14 nm delayed.

[roob]

Well any CEO that for sure knows that the problem actually is fixed perhaps?

You could also ask yourself what CEO would be stupid enough to state the prior and spread FUD about your own products even if you know for sure that the problem is actually fixed?

Any CEO of a US-listed company. Withholding material information doesn't tend to go over too well.

 

[SiliconWars]

You could also ask yourself what CEO would be stupid enough to state the prior and spread FUD about your own products even if you know for sure that the problem is actually fixed?

Yeah good point, I often miss these little nuances. If it was fixed why talk about it, surely they could absorb a month or two and nobody would notice? No, the problem here is it's not fixed. It might never be fixed. You have to treat your shareholders with care, not scare them away but also not make empty promises that can come back to haunt you.

He'll make a good Intel CEO.

 

[inf64]

So if Broadwell is H2 2014, what does this mean for the launch of Skylake? Skylake should be Q1/Q2 2015 or roughly 1 year after (initial/ pre-delay) Broadwell schedule.
If Skylake's schedule is "unchanged" then Broadwell will be one short lived design... Skylake will bring a number of features Broadwell won't have so it will be dumb to buy Broadwell if Skylake is so close (~6-8 months from Broadwell launch).

 

[liahos1]

So if Broadwell is H2 2014, what does this mean for the launch of Skylake? Skylake should be Q1/Q2 2015 or roughly 1 year after (initial/ pre-delay) Broadwell schedule.
If Skylake's schedule is "unchanged" then Broadwell will be one short lived design... Skylake will bring a number of features Broadwell won't have so it will be dumb to buy Broadwell if Skylake is so close (~6-8 months from Broadwell launch).

an interesting point. BK on the call said skylake timing was on track.

 

[Fjodor2001]

So if Broadwell is H2 2014, what does this mean for the launch of Skylake? Skylake should be Q1/Q2 2015 or roughly 1 year after (initial/ pre-delay) Broadwell schedule.
If Skylake's schedule is "unchanged" then Broadwell will be one short lived design... Skylake will bring a number of features Broadwell won't have so it will be dumb to buy Broadwell if Skylake is so close (~6-8 months from Broadwell launch).

Actually Broadwell was H2 2014 before this announcement. Now it's Q4 2014 or Q1 2015.

Your question regarding the lifetime of Broadwell is valid and interesting though, if it indeed is true that the Skylake release schedule is unchanged. Who knows, Skylake may also be delayed soon though. Based on the latest track record where SB, IB, Haswell, and now Broadwell all have been delayed for different reasons it would not be unexpected.

The tick-tock clock seems to be broken...

 

[jpiniero]

So if Broadwell is H2 2014, what does this mean for the launch of Skylake? Skylake should be Q1/Q2 2015 or roughly 1 year after (initial/ pre-delay) Broadwell schedule.
If Skylake's schedule is "unchanged" then Broadwell will be one short lived design... Skylake will bring a number of features Broadwell won't have so it will be dumb to buy Broadwell if Skylake is so close (~6-8 months from Broadwell launch).

Q1 2015 was always unrealistic for Skylake. It could just mean that Broadwell could be released in Sept-Oct 2014 and Skylake August 2015.

 

[SiliconWars]

We all know what happens to schedules once they are off. If Broadwell is delayed by 6 months then Skylake will follow a year later regardless. These things are set in motion way before yields are even known and the OEM's need to stick to a yearly cadence.

 

[OCGuy]

Intel is so far ahead...I am glad in their recent earnings report they are still dumping billions into R&D.

 

[liahos1]

i guess a more pertinent question is who cares? It't not like you're going to switch our your ivy bridge or haswell laptop for an AMD one just because broadwell is delayed 1 quarter. pc market seems to be stabilizing while intel is killing it in DCG.

Lets be real, for better or worse, Intel has no competition here. they could delay it a year and people who need to buy laptops or pcs will just get what intel is selling then. If on the other hand intel said airmont parts were delayed i'd be a little more paranoid.

 

[bullzz]

@liahos1 - if 14nm is delayed, so is broadwell high volume production. if that is delayed, it would have a ripple effect on airmont release as well

 

[liahos1]

@liahos1 - if 14nm is delayed, so is broadwell high volume production. if that is delayed, it would have a ripple effect on airmont release as well

Not necc true. If they wanted to prioritize atom they actually have better yield at 14nm vis a vis broadwell.

 

[cbn]

Not necc true. If they wanted to prioritize atom they actually have better yield at 14nm vis a vis broadwell.

That is how I understand it as well:

 

[OBLAMA2009]

who even knew they were planning to cum out with broadwell this quarter in the firstplace, it seems like they just released haswell

 

[AtenRa]

Four months after Haswell release and you can only find 10-15 laptops in NewEgg. Even if Broadwell will be released in Q4 2014 there will be no products on the market until Q1 2015 at the earliest.

 

[Arachnotronic]

Actually Broadwell was H2 2014 before this announcement. Now it's Q4 2014 or Q1 2015.

Your question regarding the lifetime of Broadwell is valid and interesting though, if it indeed is true that the Skylake release schedule is unchanged. Who knows, Skylake may also be delayed soon though. Based on the latest track record where SB, IB, Haswell, and now Broadwell all have been delayed for different reasons it would not be unexpected.

The tick-tock clock seems to be broken...

Broadwell was "first half of 2014" according to management's previous remarks. It is now Q3..

 

[AtenRa]

Ohh one more thing,

22nm IvyBridge was released in Q1 2012 (April), Haswell was released in Q2 2013(June) 14 mothns later. If Broadwell will be released in Q4 2014(October) that means 22nm to 14nm will take place in 30 months, more than 2 years.

 

[biostud]

As long as X99 + Haswell-E is not delayed...

 

[ShintaiDK]

So if Broadwell is H2 2014, what does this mean for the launch of Skylake? Skylake should be Q1/Q2 2015 or roughly 1 year after (initial/ pre-delay) Broadwell schedule.
If Skylake's schedule is "unchanged" then Broadwell will be one short lived design... Skylake will bring a number of features Broadwell won't have so it will be dumb to buy Broadwell if Skylake is so close (~6-8 months from Broadwell launch).

Skylake was never H1 2015. Skylake is still H2(early Q3) 2015.

 

[myocardia]

Ohh one more thing,

22nm IvyBridge was released in Q1 2012 (April), Haswell was released in Q2 2013(June) 14 mothns later. If Broadwell will be released in Q4 2014(October) that means 22nm to 14nm will take place in 30 months, more than 2 years.

One year has 12 months. 1½ years would then be how many months? Feel free to take your shoes off if you need to, but it's just a hair over half of your figure.

edit: 6 quarters is 1½ years

 

[Abwx]

You could also look at it as, if they are still at the "data" stage they are a long way behind the norm. Up till this point I have simply quoted what Krzanich was quoted as saying.

Absolutely not being "obtuse" in other words, you lot are just in denial of what he actually said.

It takes six to height weeks to run a single manufacturing cycle,
if the faulty step occur one month after the waffer starts and
even it s actualy corrected on the fly for the waffers that are
in a previous step you ll still have to wait a full cycle to be sure
that your process can be fully certified assuming your you have
no other issue.

 

[NTMBK]

You could also look at it as, if they are still at the "data" stage they are a long way behind the norm. Up till this point I have simply quoted what Krzanich was quoted as saying.

Absolutely not being "obtuse" in other words, you lot are just in denial of what he actually said.

They determined that there were yield issues based on data, and they determined that these issues were fixed based on data. It's fixed, and you're being obtuse.

 

[Eddward]

Are you guys talking about desktop Broadwell SKUs ?
Because last time what I saw on leaked roadmaps is Skylake LGA 1H 2015 (on track according to CEO) and Q2/Q3 2014 Haswell refresh - Lynx Point Platform...so I don't see much space for desktop Broadwell here.
Moreover at IDF, as far as I know Intel talked only about mobile - low power/ultra low power chips.
Also Z97 series chipset for Haswell refresh have Intel ME version only 9.1 (according to leaks), not 10.0 like it should be for new platform. (Sandy -> Ivy, 7.0->8.0)
Maybe I am wrong or missing something, don't know.

 

[ShintaiDK]

Are you guys talking about desktop Broadwell SKUs ?

There is no desktop Broadwell. Only E3 xeons and mobile.

 

[USER8000]

They determined that there were yield issues based on data, and they determined that these issues were fixed based on data. It's fixed, and you're being obtuse.

The way its worded seems to indicate that they "think" they might have got over the problem,and initial tests indicate it,but until the solution is enabled on a larger scale it is not a 100% certainty. At least from my own experience of working in the research field,the language pretty much fits that scenario,and people have used it before.

 

[guskline]

Intel is on track. Release it when it's ready.