But node aside, Samsung and Qualcomm can put much more money on chip R&D than AMD can, and that will reflect rather sooner than later, when ARM SoCs grow in size and complexity.
And since when this is an advantage? It would be an advantage if AMD could offer something competitive in both fronts. They clearly can't on x86, will a vanilla ARM core be enough on ARM?
Here's the scary part - the same will soon be said about these two (at least Qualcomm) and Intel.
It could be an advantage on the server front. Also AMD's overall server expertise and Seamicro fabric should be a large advantage initially. None of the ARM players can offer anything close.
You have to remember that Intel's sheer size also means they fall harder. For many people it's unimaginable that a company so large could fall, but it's possible.
You have to remember that Intel's sheer size also means they fall harder. For many people it's unimaginable that a company so large could fall, but it's possible. Qualcomm has 1/4 the staff numbers, will soon be bringing in more profits and has a fraction of the capex. This threat to Intel is real - AMD aren't even in the same league and simply have to diversify to get out of their way.
Ian made a good interview with the Intel UK site manager:
http://www.anandtech.com/show/7210/interview-with-rod-oshea-site-manager-for-intel-uk
He makes strategic thinking, that AMD in no way have the luxury of doing - not the slightest. Having people and ressources to do that, ofcourse helps the outlook for Intel in the future. Future stable cash flow from server side, b2b, and the development of new vertical business is what Intel will have to do in the future not fighting ARM ecosystem and Samsung. Its the modern thinking of embedded - unlike the old embedded we understand from eg. AMD consoles.
mrmt, we have to be carefull about saying that eg. the ARM way leads to bad situation for AMD. AMD situation is the choice between half bad, more bad, or doing more of the same bad. Pick your choice. They simple dont have a dime to move a foot. Its in that context their situation need to be asessed.
What does that even mean in a business accounting sense? "fall harder"
I understand the metaphor itself, the farther in distance you have to fall under the influence of an accelerating force such as gravity results in the farthest extremes experiencing all the more peak velocity before impacting the ground, but it just doesn't have a clear application as an analogy to Intel or the accounting realities of any business as far as I can determine.
Just because Intel has lots of revenue and a broad headcount doesn't mean they are somehow magically more susceptible to bankruptcy than is Qualcomm or AMD.
If you are just using it as a catchy phrase for adding a little punch to your posts then I have no intent to nitpick you over it and I can leave it be. But if there really is something to this in an accounting sense then I'd very much like to know more about it.
The more ground they lose, the harder the cuts will need to be in order to stave off bankruptcy. A 10% revenue loss to Intel would hurt far more than a 10% revenue loss to Qualcomm.
Interesting interview indeed. Did anyone else wonder what was up with the silly repetitive "I think...I think...I think" that started nearly every sentence? (total of 28 times, according to google, in the published interview itself)
I noticed it. It looks bad in written language but for coversations and discussions inside (lol) Intel its an advantage. Softening your opininion make it easier for people to hear and kind of accept without turning it into an agressive discussion.
As he sure have many crazy perspectives that a way to get it moving.
Secondly he have been in his position for only 6 weeks and its a way to stress that - beeing carefull. But Ian made him talk non the less !
I fail to see how anything he says can't be accomplished by ARM.
Well in these terms it means that Intel is where they are by virtue of their sheer size, their manufacturing clout, headcount etc. But their size is only supported by the same profits of much smaller competitors.
The more ground they lose, the harder the cuts will need to be in order to stave off bankruptcy. A 10% revenue loss to Intel would hurt far more than a 10% revenue loss to Qualcomm.
Qualcomm now has almost twice the cash of Intel and has no debt. Intel's profit is basically being spent on capex, yet they are still losing ground.
Sitting in a pile of cash isn't a sign of strenght, it just means that management could not find anything better to do with that pile of money and now has to leave it at the bank earning peanuts. Money in the balance sheet buys you time to change your business or acquire a competition, and it gives the muscle to shift the company in the direction management wants, but it doesn't say anything about success or failure in the future.
And look at what you are saying of Qualcomm. You are essentially saying that it is easier for anyone to make another Qualcomm than to make another Intel. And this is true. If anybody comes with better IP than Qualcomm and has access to the same fabs, they will take Qualcomm over no matter the size of their balance sheet. But to take Intel over they would need to spend a lot more of cash for foundries and R&D structure.
Intel is betting that once they can make their structures work as a single body nobody will have the will and the balance sheet to beat them, and that weakness in a part of a chain can be mitigated by the overall strength. Qualcomm has not such a luxury, they have to always be on the bleeding edge or else they will be overtaken, like TI or AMD were.
Without cash there is no opportunity of change. Capex for Intel is a serious matter.
Intel could double their profits in a single day by spending less on R&D or Capex, but they choose not too. Just like Qualcomm could become a negative cash-flow company in one day but pumping all of their profits into R&D. Intel still makes more money than Qualcomm, but they spend it differently.
They aren't spending $10 billion+ yearly capex for no reason though. The problem for Intel is they need to spend it or they'll lose their manufacturing advantage. To be honest, I don't believe that would be a bad thing for Intel but they seem hell-bent on maintaining it regardless of the cost.
Why wouldn't a manufacturing company - which is what Intel is - spend money on being the best manufacturing company??
CAPEX is not a problem for Intel. Capex would be a problem for Intel if they weren't able to execute their investiment plan, and nobody is questioning they ability to execute. Were CAPEX a problem they would be looking for financing or for a partner because you solve CAPEX problems throwing more money at the problem, but that's not the case.
What you and Pablo are questioning is the strategic direction of the company, meaning become a high volume bleeding edge mobile chips manufacturer.
We agree its not Capex in the strict sense because they can execute their investment plan. But i regard the Capex that Intel have as very fixed cost that take more than 5 years to change. If the mobile attack fails, or more like turns into an eternal loss, you stand with tons of extremely highly skilled and speciallized labor, that does not nessesary count in the amount for the next fab.
We have been presented with no argument that Intel can beat Samsung or Apple in the ARM ecosystem. Nothing but pages of PR and Antutu. Its like they think its the old gym class.
They would ask what was going wrong when they are spending more on maintaining their position than they are getting in income.
This is supposed to be an advantage right? So why is Qualcomm (soon) making more money without their own fabs, paying customer rates to TSMC, a node behind?
Why isn't Intel's process advantage counting? New nodes just get more and more expensive and Intel has already cut capex twice this year. They can no longer afford to keep going down this line. Their whole strategy appears to be hoping that the rest will stop progressing. With Apple and Qualcomm both dependent on TSMC's progress, and both having enough cash to build their own fabs if they needed to, that will never be allowed to happen.
What choice do they have? Ruiz's Asset Lite? Keep with current fab profile which is unable to fill the mobile market?
Instead you have to point at Intel's lack of a baseband LTE solution, lack of SoC component integration, and ancient CPU architecture as to reasons why Intel has failed to dominate the mobile market.
They had no competition. The rest combined didn't even have 1/10th of the money so the comparison is BS, frankly.Intel's billions a year in manufacturing technology has played a huge role in systematically destroyed each of it's major competitors in the x86 space over the last 20-30 years, so history doesn't seem to agree with your assertion.
Intel's "bleeding edge" fabs are likely to be no more bleeding-edge than the rest in 3-5 years, because they can no longer afford to maintain their position. They were too late, they missed the mobile boat while others got very rich.3-5 years from now if Intel still can't compete with mobile SoCs on it's bleed edge fabs than I'd be willing to concede to your point, but until such a time you have no ground to stand on because you are making a fallacious argument.