By: Daniel Bizo (fejenagy.delete@this.gmail.com), February 4, 2013 3:45 pm
Room: Moderated Discussions
Richard Cownie (tich.delete@this.pobox.com) on February 3, 2013 6:10 am wrote:
> Michael S (already5chosen.delete@this.yahoo.com) on February 3, 2013 2:51 am wrote:
> > Richard Cownie (tich.delete@this.pobox.com) on February 2, 2013 7:04 pm wrote:
> > >
> > > Yes. Which is still much too high to compete against current ARM-based
> > > chips, at $17-21.
> >
> > Where do your numbers are coming from?
> > If exynos4/5 and Tegra 3/4 are so cheap then how much does ST-Ericsson charge for their NovaThor chipsets?
>
> google "iPhone 5 BOM" and "Nexus 7 BOM". If you know which phones contain the NovaThor,
> you may be able to find it similarly.
>
Those are mostly just best guesses, especially for vertically integrated design manufactureres, and depending on the source you chose, they can vary significantly. Secondly, smartphone SoC ASP have been on the rise as a combination of increased integration (sorry if this was established somewhere in the thread already) AND emphasise on pushing up performance for ever better customer experience and richer apps.
Not only that Intel has made huge leaps towards much lower perf and power design points, but at the same time, these smartphone design made similarly significant progress to Intel's traditional domain and comfort zone -- high-perf logic. While it's obvious that a smartphone as a form factor will never allow such high power points as a notebook or PC, the direction is very clear. This pursuit of performance and energy efficiency plays surprisingly well to Intel's hand, which basically has to competitive advantages: microarchitecture design and semiconductor manufacturing.
Even Qualcomm can only have a fraction of the financial and engineering resources of Intel's to compete against the overlapping 4-5 Atom and low power 'big core' design teams (for tablets), and the process disadvantage is going to be painful in 12, and demoralising is 24 months time.
I don't think the question is how Intel is going to compete in chips. The question is the route to the market for smartphones. With Apple and Samsung locking down the overwhelming majority of the market, and with only a handful of other players, it will take a lot of persuasion and commitment on Intel's part to convince any of the larger players to shift. Intel was too late for both Nokia's and RIM's re-launch, Apple prefers in-house, with Samsung also showing such attitude with the flagship products. Intel will have the best performing chips soon, but who will care enough to buy it?
> Michael S (already5chosen.delete@this.yahoo.com) on February 3, 2013 2:51 am wrote:
> > Richard Cownie (tich.delete@this.pobox.com) on February 2, 2013 7:04 pm wrote:
> > >
> > > Yes. Which is still much too high to compete against current ARM-based
> > > chips, at $17-21.
> >
> > Where do your numbers are coming from?
> > If exynos4/5 and Tegra 3/4 are so cheap then how much does ST-Ericsson charge for their NovaThor chipsets?
>
> google "iPhone 5 BOM" and "Nexus 7 BOM". If you know which phones contain the NovaThor,
> you may be able to find it similarly.
>
Those are mostly just best guesses, especially for vertically integrated design manufactureres, and depending on the source you chose, they can vary significantly. Secondly, smartphone SoC ASP have been on the rise as a combination of increased integration (sorry if this was established somewhere in the thread already) AND emphasise on pushing up performance for ever better customer experience and richer apps.
Not only that Intel has made huge leaps towards much lower perf and power design points, but at the same time, these smartphone design made similarly significant progress to Intel's traditional domain and comfort zone -- high-perf logic. While it's obvious that a smartphone as a form factor will never allow such high power points as a notebook or PC, the direction is very clear. This pursuit of performance and energy efficiency plays surprisingly well to Intel's hand, which basically has to competitive advantages: microarchitecture design and semiconductor manufacturing.
Even Qualcomm can only have a fraction of the financial and engineering resources of Intel's to compete against the overlapping 4-5 Atom and low power 'big core' design teams (for tablets), and the process disadvantage is going to be painful in 12, and demoralising is 24 months time.
I don't think the question is how Intel is going to compete in chips. The question is the route to the market for smartphones. With Apple and Samsung locking down the overwhelming majority of the market, and with only a handful of other players, it will take a lot of persuasion and commitment on Intel's part to convince any of the larger players to shift. Intel was too late for both Nokia's and RIM's re-launch, Apple prefers in-house, with Samsung also showing such attitude with the flagship products. Intel will have the best performing chips soon, but who will care enough to buy it?
