Intel: There Is No Apple Foundry Deal [View article]
Rovo, This is certainly one of many interesting scenarios that may happen. I agree that it will be a two step process. Apple will not be ready to jump to an Intel based mobile based processor in the near term. It would be >1 year at least and even that is optimistic. However a long term conversion strategy is possible depending on the success of Intel's proprietary mobile technology which is far from given at this point in time.
The immediate Apple need is to get out of Samsung. Going fully to TSMC has a lot of jeopardy and possible very high cost in the form of additional wafer fab capacity. Intel already has that capacity with more of it going idle every day. There is going to be a lot of pressure on Intel's treasury and stock price if they do not find a way to take those fabs back to 90% capacity this year.
Intel: There Is No Apple Foundry Deal [View article]
"Intel: There Is No Apple Foundry Deal" < The title is correct for the time being but not necessarily for the longer term. Nothing is cast in concrete and there are very good business reasons for these two companies to work together. Never underestimate the huge market forces that are at work.
Apple: Is It Really Too Cheap To Ignore? [View article]
"Until there is a neat sign that Apple can turn its fortunes (and those of its shareholders) around, as much as the stock may appear irresistibly cheap, I wouldn't touch it." <= Who says Apple's fortunes are in the tank? Its the investors who are selling that have it wrong.
Kyle, Do you have any comment on the following article? It is based on data pre Oct 1, 2012 which is before iP5 volume shipments but the picture would still be similar:
Samsung also manufactures QCOM chips in its fabs and as a result has indepth insight into QCOM's technology. After all they are manufacturing it using their own process technology (wonder if that includes QCOM's latest LTE chips?). They may be end running QCOM just like they did with the insight they got into Apple's latest products using knowledge gained manufacturing Apple's A series microprocessors. I doubt if there are any truly effective Chinese Walls in Korea.
This is why Intel needs to humbly get into bed with Apple. They are not going to get Samsung's business. They are going to be relegated to the second tier unless they solve whatever problem is in the way of an Intel-Apple partnership.
An Apple partnership would cement Intel's credibility in the mobile chip world and potentially give Apple a cutting edge state of the art low power 14nm Trigate SOC that integrates processor and LTE baseband modem. Or alternatively in an Intel manufactured multichip module (MCM), with an SOC to follow later. That combo potentially beats Samsung at its best and brings jobs and $'s back to America.
Thanks to Stockgal87 and bob987654321 for correcting the A6 $10/chip error made by the author vs. the more correctly estimated $28.
Intel could gain an estimated "$10.6 billion revenue increase" by building 100% of Apple's A series processors "a 19.6% revenue increase." That's quite a pop. Those numbers make a lot more sense that the $10 assumed by Ashraf and change the whole picture. I continue to disupte the conclusion drawn in the title of this article.
If you add in the Baseband solution as Ashraf refers to it, and it becomes possible to make a SOC solution out of it, that makes it even more lucrative for an Intel-Apple deal. That could add another $20-30 per chip and make for an estimated $20B/year total.
I have to believe that the two of them must be talking. There has been no confirmation of a joint TSMC-Apple fab and maybe this is the reason why. Ottelini's forecast of 2 new Intel fabs this year may depend on such a deal either in the form of an SOC or a multichip module.
I would add another comment from Russ Fischer on anther article written by Ashraf, reference below, that states the potential financial gain may be significantly higher:
"Per the iSuppli iPhone BOM report, the application processor costs $17.50 which implies that Apple is only paying Samsung $7000 per 32nm wafer for which they are still in the learning curve with less than optimum yields. Figure closer to $30 for the A chip. The radio section is listed as $34 dollars. That includes wi-fi, power amps etc. That they grouped all this stuff again indicates that they don't have much of a clue, but guess the LTE+transceiver at $25 of the $34. Apple can (must) use a discrete design. For them that is a good approach because the rate they do upgrades on the A chips, they would spend their lifetimes re-certifying the chips if it included the LTE baseband. I would guess that Qualcom is really hosing those companies that chose to use the single chip solution....probably more than the two chip solution, so about $50-60"
Public Cloud Price War Starts Hurting Cloud Software Suppliers [View article]
Bubbrasil, I agree with your comment on Amazon but Bezos seems intent on a scorched earth zero margin business strategy. I also wonder how long Amazon can hold out without having earnings go significantly negative. That event might crater the shares big time due to their ridiculously high PE of 318.
The good report from Best Buy today suggests sales are moving back to bricks and mortar from Amazon now that buyers have to pay sales tax. This has leveled the playing field significantly. Fry's and others have price match guarantees. The tax free advantage Amazon had, 8% delta in California, may cause a lot more pain than investors are expecting.
So I wonder how long Amazon can continue the scorched earth cloud pricing model?
Public Cloud Price War Starts Hurting Cloud Software Suppliers [View article]
Dana, Thanks for heads up. When you say public cloud services I assume you are not including companies like Salesforce.com and Netsuite. I suspect you are talking about the use of the cloud by companies for their IT system needs. Do you think Citrix and Oracle will be affected? I need to do some more thinking on this. Any insight you have appreciated. Thx, Robert
Intel: There Is No Apple Foundry Deal [View article]
The immediate Apple need is to get out of Samsung. Going fully to TSMC has a lot of jeopardy and possible very high cost in the form of additional wafer fab capacity. Intel already has that capacity with more of it going idle every day. There is going to be a lot of pressure on Intel's treasury and stock price if they do not find a way to take those fabs back to 90% capacity this year.
Intel: There Is No Apple Foundry Deal [View article]
Apple: Is It Really Too Cheap To Ignore? [View article]
Intel Won't Build Apple's iPhone Chips [View article]
Intel Can't Let Samsung Win [View article]
Will Chinese Consumers Save Apple? [View article]
Year of the China Droid
http://bit.ly/XXnbHG
Intel Can't Let Samsung Win [View article]
Intel Can't Let Samsung Win [View article]
Modem algorithm design in LTE advanced systems
Location: Suwon, South Korea
http://bit.ly/13ykTWl
Intel Can't Let Samsung Win [View article]
Ashraf, Here you go:
QUALCOMM Selects Samsung Electronics Co., Ltd. for New Supply Partnership, November 22, 2005
http://prn.to/ZZDrOv
Let's hope QCOM has a more effective noncompete agreement in place than Apple did.
Intel Can't Let Samsung Win [View article]
Intel Can't Let Samsung Win [View article]
Intel Can't Let Samsung Win [View article]
Samsung announces 8-core Exynos 5 Octa mobile chipset
http://bit.ly/13yherq
Galaxy S3 teardown reveals Samsung LTE modem
http://bit.ly/12G3YkB
Samsung Announces Milestone in Development of 14nm
http://bit.ly/13yhczP
This is why Intel needs to humbly get into bed with Apple. They are not going to get Samsung's business. They are going to be relegated to the second tier unless they solve whatever problem is in the way of an Intel-Apple partnership.
An Apple partnership would cement Intel's credibility in the mobile chip world and potentially give Apple a cutting edge state of the art low power 14nm Trigate SOC that integrates processor and LTE baseband modem. Or alternatively in an Intel manufactured multichip module (MCM), with an SOC to follow later. That combo potentially beats Samsung at its best and brings jobs and $'s back to America.
Intel Won't Build Apple's iPhone Chips [View article]
Intel could gain an estimated "$10.6 billion revenue increase" by building 100% of Apple's A series processors "a 19.6% revenue increase." That's quite a pop. Those numbers make a lot more sense that the $10 assumed by Ashraf and change the whole picture. I continue to disupte the conclusion drawn in the title of this article.
If you add in the Baseband solution as Ashraf refers to it, and it becomes possible to make a SOC solution out of it, that makes it even more lucrative for an Intel-Apple deal. That could add another $20-30 per chip and make for an estimated $20B/year total.
I have to believe that the two of them must be talking. There has been no confirmation of a joint TSMC-Apple fab and maybe this is the reason why. Ottelini's forecast of 2 new Intel fabs this year may depend on such a deal either in the form of an SOC or a multichip module.
I would add another comment from Russ Fischer on anther article written by Ashraf, reference below, that states the potential financial gain may be significantly higher:
"Per the iSuppli iPhone BOM report, the application processor costs $17.50 which implies that Apple is only paying Samsung $7000 per 32nm wafer for which they are still in the learning curve with less than optimum yields. Figure closer to $30 for the A chip.
The radio section is listed as $34 dollars. That includes wi-fi, power amps etc. That they grouped all this stuff again indicates that they don't have much of a clue, but guess the LTE+transceiver at $25 of the $34.
Apple can (must) use a discrete design. For them that is a good approach because the rate they do upgrades on the A chips, they would spend their lifetimes re-certifying the chips if it included the LTE baseband.
I would guess that Qualcom is really hosing those companies that chose to use the single chip solution....probably more than the two chip solution, so about $50-60"
Comment from the following reference:
http://bit.ly/YJCZPE
Public Cloud Price War Starts Hurting Cloud Software Suppliers [View article]
The good report from Best Buy today suggests sales are moving back to bricks and mortar from Amazon now that buyers have to pay sales tax. This has leveled the playing field significantly. Fry's and others have price match guarantees. The tax free advantage Amazon had, 8% delta in California, may cause a lot more pain than investors are expecting.
So I wonder how long Amazon can continue the scorched earth cloud pricing model?
Public Cloud Price War Starts Hurting Cloud Software Suppliers [View article]
Thx, Robert