[CPUWorld] Intel "Braswell" 14 nm systems-on-a-chip delayed (again)

[Fjodor2001]

Yet another Intel 14 nm CPU delay. This time for the Braswell SoC:

http://www.cpu-world.com/news_2014/2014082102_Intel_Braswell_systems_on_a_chips_delayed_again_.html

Originally, Braswell SoCs were scheduled for Q1 2015 release date. In July, we reported that the launch have been pushed to at least March 2015, and possibly to April or May 2015. As we recently learned, the processors have been delayed by another quarter due to slower, than expected, sample validation. Current "Ready To Ship" dates of Braswell microprocessors are June - August 2015. Their production is expected to start in April - June 2015.

I wonder what this means: "As we recently learned, the processors have been delayed by another quarter due to slower, than expected, sample validation."? Are they talking about "low yields"? Or do they mean that the chip Verification team is taking longer than expected to verify that the chip design works correctly?

 

[witeken]

It has nothing to do with yields. I think this is what the article is talking about: http://en.wikipedia.org/wiki/Post-silicon_validation

 

[Burpo]

Being that these will be <10 watt chips and mostly used by OEM, I'm sure Intel wants zero issues/problems with their new Pentium/Celeron offerings.

 

[*NixUser]

Give then time. We don't want another TSX bait and switch, right?

 

[witeken]

Give then time. We don't want another TSX bait and switch, right?

We don't want delays, either.

 

[liahos1]

its depressing. how is this the company that became chipzilla? They've owned every major compute segment they've gone after and they cant spin out some low power parts that all their competitors seem to push out scatter shot. i hope bk gets in there and restructures both the mobile comm's group and their atom group. What happened to only the paranoid survive?

 

[Phynaz]

Give then time. We don't want another TSX bait and switch, right?

Inflammatory much?

 

[witeken]

its depressing. how is this the company that became chipzilla? They've owned every major compute segment they've gone after and they cant spin out some low power parts that all their competitors seem to push out scatter shot. i hope bk gets in there and restructures both the mobile comm's group and their atom group. What happened to only the paranoid survive?

I think a huge amount of wrong decisions in the past 8 years led to this ("If we pretend it isn’t happening, will it go away?").

 

[monstercameron]

I kinda find it funny but alas Intel can do no wrong.

 

[liahos1]

I think a huge amount of wrong decisions in the past 8 years led to this ("If we pretend it isn’t happening, will it go away?").

fwiw i've had a few discussion with ex intel baseband engineers and they think Hermann is out of his seat at the end of the year and Aisha takes over.

 

[frozentundra123456]

Inflammatory much?

Well, I usually try to support intel, since they seem to take a lot of unwarranted (and warranted) criticism on these forums, but that was my first thought too: maybe they will avoid another TSX type debacle, or the haswell chipset bug, etc.

 

[PPB]

Or the P67 SATA 3 bug, which was a big downer in countries like mine where not every retailer agreed to replace your mobo with a b3 one free of charge

 

[monstercameron]

Well, I usually try to support intel, since they seem to take a lot of unwarranted (and warranted) criticism on these forums, but that was my first thought too: maybe they will avoid another TSX type debacle, or the haswell chipset bug, etc.

its interesting that you think Intel receives unwarranted criticism. The way I see it they always get a pass, then again opinions are just that. This isn't the first product Intel has delayed on these smaller nodes. Maybe that is a sign for things to come.

 

[Haserath]

Inflammatory much?

We don't, do we?()

 

[Arachnotronic]

its depressing. how is this the company that became chipzilla? They've owned every major compute segment they've gone after and they cant spin out some low power parts that all their competitors seem to push out scatter shot. i hope bk gets in there and restructures both the mobile comm's group and their atom group. What happened to only the paranoid survive?

The problem is that the paradigm for mobile processors is vastly different from PC chips.

You can't just have the best CPU on a great process and win...you need to have the best CPU, modem, graphics, image signal processor, video encode/decode blocks, and you need to be able to deliver to very strict product release cadences since the market is highly competitive.

Qualcomm realized this well before everybody else, and this is why they are as powerful and successful as they are today.

 

[Fjodor2001]

What happened to only the paranoid survive?

It died with the 8% annual performance increase policy.

 

[frozentundra123456]

its interesting that you think Intel receives unwarranted criticism. The way I see it they always get a pass, then again opinions are just that. This isn't the first product Intel has delayed on these smaller nodes. Maybe that is a sign for things to come.

To be fair, AMD receives its share of bashing to, but I certainly dont agree that these forums are pro-intel like a lot of people seem to think. But you are right, everyone is entitled to their opinion, and this is off topic. I dont think we should pursue this.

 

[Arachnotronic]

This Braswell delay is unfortunate, but the key will be for the chips to be ready to go in systems for sale during the back-to-school season. The current timeline still suggests that this is possible, though it'll be close.

 

[Roland00Address]

I thought the Core M were supposed to be pre christmas, according to this article they were q1 2015 which means past christmas and they are going to be delayed till at least Mar 2015 (aka very end of q1 2015 probably q2 2015)

are we talking the same chips here?

 

[witeken]

Braswell != Core M

Braswell is the name for the Celeron and Pentium (N suffix) laptop and desktop Atom SKUs.

Core M and Cherry Trail are still Q4.

 

[monstercameron]

I thought the Core M were supposed to be pre christmas, according to this article they were q1 2015 which means past christmas and they are going to be delayed till at least Mar 2015 (aka very end of q1 2015 probably q2 2015)

are we talking the same chips here?

I dont think core-m and brawsell are the same.

 

[kimmel]

are we talking the same chips here?

No we aren't. Core M is big core 14nm Broadwell. This is Braswell or 14nm small core.

 

[Exophase]

How much are Bay Trail-M and Bay Trail-D selling? I remember when Cedar Trail was substantially delayed due to a lack of interest. That's not what anyone's saying here, but amount of high volume demand will tend to correlate with how hard they're pushing it out the door.

 

[Arachnotronic]

How much are Bay Trail-M and Bay Trail-D selling? I remember when Cedar Trail was substantially delayed due to a lack of interest. That's not what anyone's saying here, but amount of high volume demand will tend to correlate with how hard they're pushing it out the door.

Extremely well.

BYT-M is now 20% of Intel's entire notebook mix, and BYT-M/D represent 60% of Intel's Pentium/Celeron mix.

 

[ElFenix]

No we aren't. Core M is big core 14nm Broadwell. This is Braswell or 14nm small core.

that's not confusing at all