Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Broadwell shows up. Kind of.

QuizzicalQuizzical Member LegendaryPosts: 25,531

Intel promised that Broadwell would be available in time for Christmas.  And it is, kind of.  New Egg now has two models in stock.

The cheaper model sports an 800 MHz dual core processor with turbo up to 2 GHz, Intel integrated graphics, 4 GB of memory, 128 GB of storage, and a tiny 2-cell battery.  The more expensive model offers a 1.1 GHz dual core processor with turbo up to 2.6 GHz, the same Intel integrated graphics, 8 GB of memory, 256 GB of storage, and the same 2-cell battery.  Both are 11.6" laptops with Windows 8.1 Pro.  Being ultrabooks, the memory, storage, and battery presumably aren't user upgradeable or fixable if they break.

Care to take a guess at prices?  Would you believe $1066, before shipping and taxes?  Oh, and that's the cheaper model.  Add more than $422 to get the more expensive model.  Yes, that's in US dollars.

If there is such a thing as a price tag saying "please don't buy me", this is it.

So what is going on here?  If you believe Intel's tick tock cadence, Broadwell was supposed to launch last year.  Yes, Ivy Bridge was delayed and Haswell was delayed more, but it's still a year and a half after the Haswell launch and Broadwell parts barely exist.

The logical possibilities are that OEMs simply declined to produce Broadwell-based laptops or that Intel simply can't produce enough chips soon enough.  Given how many hundreds of millions of dollars Intel has paid OEMs to produce completely stupid laptops in the past few generations that they wouldn't have otherwise built, it's unlikely to be the former.  Intel can dedicate their entire 14 nm process node to Broadwell production, so it's not like they're competing for fab space.

I see two possibilities--and they're not mutually exclusive:

1)  Broadwell has been delayed more than Intel is willing to admit, and September's paper launch was very much a paper launch.  This wouldn't preclude a bunch of Broadwell laptops showing up soon, hopefully at less than half the price of the ones above.

2)  Intel's troubled 14 nm process node is more troubled than they're willing to admit and they simply can't produce very many chips at all, or perhaps rather, it's very expensive on a per chip basis.

Whether (1) is true or not, I think there's a substantial dose of (2) involved here.  Intel has boldly forged ahead with a 14 nm process node while the rest of the industry thinks that they'll really need EUV lithography to sensibly go below about 20 nm.  Sometimes if marketing says you should do one thing and physics demands another, physics wins, no matter what management decides to do.

That said, it's not like Intel really needs Broadwell soon in order to be competitive with AMD.  They're got Haswell for higher performance laptops and Bay Trail Atom for cheaper, lower power models.  Broadwell tries to offer power consumption closer to Atom with performance closer to Haswell--or at least ULV versions of Haswell.  The only real market segment where Intel doesn't have a sensible product is if you want integrated graphics that are good.  That's not a huge market, and Broadwell wouldn't come remotely close to filling that hole, even if the market were drowning in Broadwell chips.

Now, it's entirely possible that the arrival of EUV lithography makes Intel's 14 nm fab problems go away.  Of course, it will likely do the same for TSMC, Samsung, and Global Foundries.  It will be interesting to see how much they can narrow the gap with Intel on fab capabilities and how quickly.

Comments

  • RidelynnRidelynn Member EpicPosts: 7,383


    Originally posted by Quizzical
    That said, it's not like Intel really needs Broadwell soon in order to be competitive with AMD.  They're got Haswell for higher performance laptops and Bay Trail Atom for cheaper, lower power models.  Broadwell tries to offer power consumption closer to Atom with performance closer to Haswell--or at least ULV versions of Haswell.  The only real market segment where Intel doesn't have a sensible product is if you want integrated graphics that are good.  That's not a huge market, and Broadwell wouldn't come remotely close to filling that hole, even if the market were drowning in Broadwell chips.

    I think this is the biggest factor.

    The "release" we see now is just something token - they said they would do it, so they did it.

    For the time being, they are really only competing with themselves in the x86 arena. Their real competition is in the ultra low power segement, where ARM pretty well has anything tablet and smaller wrapped up. Broadwell doesn't quite get into that market (yet - but it gets a lot closer), and as long as ARM isn't seriously pushing up into the laptop market (there have been rumors but no serious products yet), or AMD comes out from behind to push something dramatically better, I don't think Intel is in any hurry.

    And if you don't really have competition - you aren't under any time or financial pressure to push out a new product release. You need to keep moving just fast enough to stay ahead of the curve, but any faster than that and your just hurting yourself.

    For the x86 market - Intel isn't staying stagnant, but they aren't going to push any faster than their competition makes them. They are focused very highly on the mobile+wearable market, and they should be - they have a fairly embarassing showing in that segment for as large of a head start as they had in the desktop/laptop class market for so many years. And hopefully they are looking at what is past that (I'm sure they are thinking wearables (and hence the latest Google Glass announcement), but I think they need to be looking past that even, I think the idea of "transformable" x86's has been a flop) - but that's easier said than done.

Sign In or Register to comment.