Array ( [content] => [params] => Array ( [0] => /forum/index.php?threads/intel-q12022-beats-but-offers-disappointing-guidance.15953/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 has no choice, other than be a foundary, if it wants to survive long term.
That's for sure, but Intel also allowed themselves to fall behind in CPU, GPU, networking, and FPGA technologies. These products would not win in the datacenter, HPC, laptops, or cloud networks, not to mention mobile devices. And there's the loss of Apple's Mac business, while Mac volumes are climbing dramatically. Amazon and Ampere have enabled ARM to make inroads into cloud datacenters. I don't think for any of these businesses a "one generation ahead" fab process (as Craig Barrett used to say) would have made a difference. Intel has basic merchant chip product architecture problems, and IMO the pain won't stop until the real innovation starts.As just another fabless processor design company, it could be even more competitive for Intel. But fab power doesn't seem to provide Intel with the advantage it once had.
I think there's a deeper justification. Graviton gives Amazon a huge cost advantage over merchant chips, from anyone, and their users in AWS get to specify what they want without competing with broad-market requirements. And I think Google and Microsoft are working as fast as they can to have their own server CPUs too, because they're watching Amazon succeed, especially with EC2, and thinking, "Hey! We can do that!".re amazon choosing arm, to be fair, they faced no supply from amd vs crap from intel. not nice.
Recently I started thinking if Intel does retake the technology leadership in semiconductor manufacturing around 2025, will that solve Intel's problems?By way of getting our thinking straight, it bears noting that intels MO as a fab has been very different from the pure fab model of tsmc.
intel has used its fab lead as leverage to dominate markets for the end product. Advances with no market advantage gained held little interest for them.
even recently, optane, a fine product, was made so exclusive that the market gave up on it and it ultimately failed to gain traction.
They may have done some sub-contract fabbing, but most was intel brand chips.
This mindset was part of them ignoring the inevitable demise of moores law.
with 100% of the server market anyway, & several year lead in process, why do a complete over haul of products (a switch to chiplets) for no gain to intel?
they didnt see zen coming - it (32 core epyc & 8/16 core desktop) gained a foot hold even w/ crap Glofo process initially, & then the 10nm debacle & tsmc turned attrition into a route.
Samsung have a foot in both models - consumer SSDs eg., but they seem troubled too
I’m not so sure about that. I think it’s more likely AMD is achieving higher product gross margins than Intel because their products are just plain better.After paying TSMC huge fabrication cost, AMD have the same gross margin as Intel.
You can tell TSMC cost is a lot lower than Intel cost.
NXP Semiconductors also released their Q1 2022 earnings on Monday.AMD just released their Q1 2022 earnings. They achieved a stunning 71% gross revenue growth. Excluding the new acquired Xilinx revenue, AMD still achieved about 55% quarterly revenue growth YoY. I have updated the Q1 earnings comparison for major semiconductor companies below:
Q1 2022 vs. Q1 2021 Revenue (Excluding equipment manufacturers):
Intel: -1% (-1% non GAAP, -7% GAAP)
TI: 14%
Micron: 25% (Fiscal Quarter ended March 3, 2022)
MediaTek: 32.1%
UMC : 34.7%
TSMC: 36%
Samsung Memory: 39%
Qualcomm: 41%
SK Hynix 43%
AMD: 71% (GAAP and Non-GAAP, 55% if excluding Xilinx)