Array
(
    [content] => 
    [params] => Array
        (
            [0] => /forum/index.php?threads/intel-needs-another-andy-grove.17376/page-2
        )

    [addOns] => Array
        (
            [DL6/MLTP] => 13
            [Hampel/TimeZoneDebug] => 1000070
            [SV/ChangePostDate] => 2010200
            [SemiWiki/Newsletter] => 1000010
            [SemiWiki/WPMenu] => 1000010
            [SemiWiki/XPressExtend] => 1000010
            [ThemeHouse/XLink] => 1000970
            [ThemeHouse/XPress] => 1010570
            [XF] => 2021370
            [XFI] => 1050270
        )

    [wordpress] => /var/www/html
)

Intel needs another Andy Grove

When both PC and non-PC demand are down, having the foundry is not an advantage. The foundry also does not have enough process bandwidth (16/3/18A), yet there's more demand at mature nodes. Outsourcing some tiles is another bad sign.
Mostly agree but some thoughts, if anything wouldn’t foundry make you far more resilient to semi downturns then being a pire IDM since you widen your product base? Historically fabless merchant chip vendors operate on thin margins (although this could just be due to the nature of FPGAs and ARM socs). So while they don’t have to worry about underutilized fabs, they will still be under a lot of financial pain due to the high volumes needed when your products are low margin.

As for the foundry bandwidth, I don’t think I understand what you mean Fred. Did you mean that intel doesn’t have capacity for major external and internal customers? If so then I mostly agree. There should be plenty of i16 capacity that can be made available at the drop of a hat given how hard intel ramped 14nm. As for the EUV nodes things will undoubtedly be tight for the next couple of years, but disagrigated designs and a little bit of time should allow for capacity issues to maybe not be sorted but at least significantly softened.

In my opinion outsourcing isn’t really a red flag yet. At least for mtl what are your options to replace N5 or N7. Nothing and i7. That doesn’t really reflect poorly on future nodes, rather the complex nature of i7 making it poorly suited for this role.
 
Last edited:
Mostly agree but some thoughts, if anything wouldn’t foundry make you far more resilient to semi downturns then being a pire IDM since you widen your product base? Historically fabless merchant chip vendors operate on thin margins (although this could just be due to the nature of FPGAs and ARM socs). So while they don’t have to worry about underutilized fabs, they will still be under a lot of financial pain due to the high volumes needed when your products are low margin.

As for the foundry bandwidth, I don’t think I understand what you mean Fred. Did you mean that intel doesn’t have capacity for major external and internal customers? If so then I mostly agree. There should be plenty of i16 capacity that can be made available at the drop of a hat given how hard intel ramped 14nm. As for the EUV nodes things will undoubtedly be tight for the next couple of years, but disagrigated designs and a little bit of time should allow for capacity issues to maybe not be sorted but at least significantly softened.
Having the foundry would be advantageous if the PC demand is down, but non-PC is not, just that now everything is affected. Here the foundry of Intel has only three nodes announced so far while ideally, it should have as many as possible. For example, the 22nm and larger, where there is a larger market. Intel 4 was only used for the CPU tile, while GPU, SoC, and I/O tiles were outsourced to TSMC, not done internally. So on one hand, it is fab-lite at these more advanced nodes, but it still wants to be a foundry provider at the same nodes? Capacity for itself or its foundry customers would be too low both ways.
 
Please no more Andy Groves. The last thing Intel needs is another paranoid narcissist engaged in channel system structural restraints, artificial accelerators administered on weighed field effects, market gaming, rackateering, and Dr. Grove sending his goon squad after whomever he perceived as threatening to his personality type, and or Intel, by refusing to tow the Intel line. Although I will say in support of Dr. Grove, unlike today, then you paid for every unit of production coming out of the fab and today half of Intel's production volume is bundled in as sales close for no charge that is seriously destroying Intel in terms of losing cost optimization and revenue generation. Dr. Grove certainly understood the benefits of bribes, Intel Inside for example of throwing some free product at this and that country for sales access but today the amount of product walking out the back door is absurd. This has always been as issue in Intel and similar design manufacturer / producers when lithography hurdles appear insurmountable and employees in fear of their livelihood begin to rob the enterprise blind. Mike Bruzzone, Camp Marketing
 
Intel-dropped-E-logo-300x156.jpg


When Intel takes the lead on silicon performance, they should drop the 'e' again ..
 
Back
Top