Given that stock cortex a8 chips were available one year before both Samsung and Apple released their version, you are saying that Samsung and Apple have caught up with the big silicon companies in one year's time.
Wow, way to miss the point and say something else unrelated and nonsensical. You must be channelling the BB CEO.
I'm saying that despite what you wrote Samsung already gave Apple priority for the A4 OVER their own Hummingbird product. Therefore what you wrote was stupid and therefore you completely evaded this by writing another stupid statement to attempt to divert attention.
Samsung is a "big silicon company". More so than Ti. So what you wrote was once again factually incorrect and stupid. Samsung didn't "catch up" to anyone. Intrinisty and Samsung enhanced their Samsung's A8 product used in the 3GS to be one of the very best SoC of the A8 generation: the Hummingbird.
And before you embarrass yourself further, the reason why Samsung > Ti is because Samsung is sampling 32nm A9s where Ti isn't going to move past 45nm themselves but probably end up using TSMC. Half of Ti's production was already outsourced in 2007. It's probably higher now and even more so when they go fabless for 32nm and beyond. So much for "big silicon".
Meanwhile:
"South Korea's Samsung Electronics Co. Ltd. said its foundry business has qualified a 32-nm low-power process with high-k metal gate technology. The company lays claim to being the first foundry to qualify a low-power process using high-k at 32-nm."
Given that stock cortex a8 chips were available one year before both Samsung and Apple released their version, you are saying that Samsung and Apple have caught up with the big silicon companies in one year's time.
You don't even know what the ARM Cortex line is do you. (yes that was a statement not a question)
ARM licenses IP, not chips. Companies contract to use the IP to make the chips with whatever customizations they want.
There is no such thing as a stock Cortex chip!
They don't exist!
Because the Cortex IP itself isn't the entire package of IP needed to generate a functional silicon CPU!
The IP was available in 2008, the first complete shipping implementations (not counting pre production tape-outs) didn't show until Feb/Mar 2009.
Anywhere.
You.
Still.
Fail...
As for big silicon. I only consider Intel and AMD in that company. If you haven't shipped 25nm or smaller memory runs you don't qualify as big silicon. 32nm is a generation old tech. Cheap to build the fab for that, only about a $ Billion and a quarter for the machinery. You want big silicon you don't start until your fab runs $3 Billion for less than 25 nm.
You don't even know what the ARM Cortex line is do you. (yes that was a statement not a question)
ARM licenses IP, not chips. Companies contract to use the IP to make the chips with whatever customizations they want.
There is no such thing as a stock Cortex chip!
They don't exist!
Because the Cortex IP itself isn't the entire package of IP needed to generate a functional silicon CPU!
The IP was available in 2008, the first complete shipping implementations (not counting pre production tape-outs) didn't show until Feb/Mar 2009.
Anywhere.
You.
Still.
Fail...
As for big silicon. I only consider Intel and AMD in that company. If you haven't shipped 25nm or smaller memory runs you don't qualify as big silicon. 32nm is a generation old tech. Cheap to build the fab for that, only about a $ Billion and a quarter for the machinery. You want big silicon you don't start until your fab runs $3 Billion for less than 25 nm.
I know that only 4 companies to-date has signed ARM architecture license --- Qualcomm, Infineon, Marvell and Microsoft. They can design their own cores.
I know that all the other players license the cores and put their own extra stuff around it.
I know that TI --- which has a long history of designing application processors --- will barely able to produce low 6 figure quantities of OMAP4430 dual core cortex A9 application processors for the RIM Playbook.
I also know that Samsung while they are big in silicon, they are heavy on memory chips, not actual application processors. That's why Samsung licensed the Hummerbird.
Now you are telling me that Samsung and Apple --- while they are both big companies with lots of cash ---- both are pretty new to this industry and somehow they are going to do even better TI --- which is to produce millions of dual core cortex A9 application processors for the iPad 2.
I know that only 4 companies to-date has signed ARM architecture license --- Qualcomm, Infineon, Marvell and Microsoft. They can design their own cores.
I know that all the other players license the cores and put their own extra stuff around it.
I know that TI --- which has a long history of designing application processors --- will barely able to produce low 6 figure quantities of OMAP4430 dual core cortex A9 application processors for the RIM Playbook.
I also know that Samsung while they are big in silicon, they are heavy on memory chips, not actual application processors. That's why Samsung licensed the Hummerbird.
Now you are telling me that Samsung and Apple --- while they are both big companies with lots of cash ---- both are pretty new to this industry and somehow they are going to do even better TI --- which is to produce millions of dual core cortex A9 application processors for the iPad 2.
You are showing yourself to be an absolute talk out your ass sideways idiot. Apple is an ARM licensor, as is TI- what do you think the OMAP series are? And Samsung is a licensor too--Hummingbird isn't made or designed by ARM, the CortexA8 core was licensed by Samsung so they could make the Hummingbird. There are also a handful of DoD and European defense contractors that contract for ARM IP directly too.
Your statements only highlight your utter lack of knowledge as to what ARM sells and what licensors get. At least you keep creeping a little closer to reality each time you try to cover your arse, but the rate is abysmally low given the learning opportunities you have been provided.
You are showing yourself to be an absolute talk out your ass sideways idiot. Apple is an ARM licensor, as is TI- what do you think the OMAP series are? And Samsung is a licensor too--Hummingbird isn't made or designed by ARM, the CortexA8 core was licensed by Samsung so they could make the Hummingbird. There are also a handful of DoD and European defense contractors that contract for ARM IP directly too.
Your statements only highlight your utter lack of knowledge as to what ARM sells and what licensors get. At least you keep creeping a little closer to reality each time you try to cover your arse, but the rate is abysmally low given the learning opportunities you have been provided.
What made up un-facts will you try next?
I never said otherwise.
I said that big merchant established chip companies are barely able to produce low quanities of dual core cortex A9 application processors --- these are stock merchant chips. Now you are telling me that Samsung and Apple who are both relatively new to the scene are going to be virtually certain to provide millions of these same kind of chips at the same time as TI. Remember these are newcomers who didn't have the expertise and had to license the Hummerbird from someone else.
Is it possible? Sure, it is possible. But you are making it sound as if it is a virtual certainty, and I disagree with that assessment.
I know that only 4 companies to-date has signed ARM architecture license --- Qualcomm, Infineon, Marvell and Microsoft.
Quote:
Originally Posted by Hiro
You are showing yourself to be an absolute talk out your ass sideways idiot. Apple is an ARM licensor, as is TI- what do you think the OMAP series are? And Samsung is a licensor too--Hummingbird isn't made or designed by ARM, the CortexA8 core was licensed by Samsung so they could make the Hummingbird. There are also a handful of DoD and European defense contractors that contract for ARM IP directly too.
Quote:
Originally Posted by samab
I never said otherwise.
wow.
just wow.
Either you have the worlds shortest memory or convenient lying is considered a valid debate tactic in your world. Your post's above are QFT.
Quote:
Originally Posted by samab
I said that big merchant established chip companies are barely able to produce low quanities of dual core cortex A9 application processors --- these are stock merchant chips. Now you are telling me that Samsung and Apple who are both relatively new to the scene are going to be virtually certain to provide millions of these same kind of chips at the same time as TI. Remember these are newcomers who didn't have the expertise and had to license the Hummerbird from someone else.
Is it possible? Sure, it is possible. But you are making it sound as if it is a virtual certainty, and I disagree with that assessment.
You keep saying that, but you have no evidence. You are just making up words hoping something will stick.
There are shipping CortexA9 chips in the wild. Not small feature-size production runs for low power, but there isn't any reason to expect that at less than the 1 year point from IP availability. The implementations are actually going swimmingly well based on both current producers of plug-in device cores shipping well respected units. This means there are no architectural gotchas for the small feature size crowd to deal with, they only have to deal with setting up the fab process.
Apple isn't a fab, so your assessment of their experience in producing chips is utterly irrelevant and worthless. Because Apple simply doesn't do that! And the opposite side of that, the diss on Samsung is way off base too, because Samsung is a proven 32nm memory producer for a couple years now, meaning they have the proven technical ability to run the fab at the feature size the low-power CortexA9 based cores will be fabricated on.
I know that only 4 companies to-date has signed ARM architecture license --- Qualcomm, Infineon, Marvell and Microsoft. They can design their own cores.
This is completely wrong as usual. Intel, Apple, and Samsung are also architecture license holders. Apple is unconfirmed given their traditional secrecy but most folks believe so. Especially given ARM's coy statement and Apple's early history in creating ARM for Newton with Acorn.
Quote:
I know that all the other players license the cores and put their own extra stuff around it.
I know that TI --- which has a long history of designing application processors --- will barely able to produce low 6 figure quantities of OMAP4430 dual core cortex A9 application processors for the RIM Playbook.
This seems highly unlikely. Given your penchant for making "facts" up I'd want a link.
Quote:
I also know that Samsung while they are big in silicon, they are heavy on memory chips, not actual application processors.
See what I mean? A made up "fact". Samsung is only the second largest semiconductor company behind intel and the largest supplier of application processors.
"Scemama notes that Samsung is the world's largest vendor of application processors."
Again wrong. They didn't license Hunningbird. They designed it in collaboration with Intrinsity.
Quote:
Now you are telling me that Samsung and Apple --- while they are both big companies with lots of cash ---- both are pretty new to this industry and somehow they are going to do even better TI --- which is to produce millions of dual core cortex A9 application processors for the iPad 2.
Wrong again. Neither Samsung nor Apple are new. Apple has been a fabless designer for a while (mostly northbridges prior to buying pa semi but they also owned a big chunk of ARM). And samsung didn't become the second largest semi company expected to overtake Intel overnight. Samsung passed Ti a while ago.
I don't think you wrote a single correct statement in this entire post.
Again wrong. They didn't license Hunningbird. They designed it in collaboration with Intrinsity.
That is corporate talk semantics. It's like the xbox CPU's. Microsoft licensed the IBM powerpc architecture and then hire IBM to help them design the xenon cpu. The end result is that the CPU itself is a Microsoft's own IP. Who did most of the work? It's IBM.
That is corporate talk semantics. It's like the xbox CPU's. Microsoft licensed the IBM powerpc architecture and then hire IBM to help them design the xenon cpu. The end result is that the CPU itself is a Microsoft's own IP. Who did most of the work? It's IBM.
You are completely wrong. On all counts, well except that it was IBM that did the work, but sometimes even a blind squirrel finds a nut.
Once again you prove that you can make crap up and spout it off with the best of them, being utterly incorrect on any important point the whole while.
Microsoft did not design the Xenon. IBM did, completely. Microsoft contracted IBM to design and produce the chip, providing requirements and feedback in the process. IBM started with their existing PPC muli-core design and then wrapped three of them together with VMX optimizations and communications crossbars to meet the MS XBox CPU spec requirements. Microsoft didn't license and design anything, they bought finished, custom specified hardware. That's all.
Don't let yourself be confused by MS having a couple employees working with IBM. That doesn't constitute design any more than it did between Apple, Motorola and IBM when the AltiVec/VMX design was created. Apple's engineers had the grand picture and knew what they wanted working with Moto and IBM every day. They were even very capable, participating completely in the architecture process. But all the actual decisions were Moto's/IBM's, so very explicitly AltiVec belongs only to Moto and is a Moto design just as VMX is for IBM.
Your idea of corporate semantics is the thing that's f*%-ed up here. Samsung DID design the Hummingbird (not the Hummerbird as you have repeatedly called it throughout the thread). Samsung licensed the core IP from ARM, then contracted for engineering assistance from Intrinsity to create a complete CPU design. After they were done Samsung fabbed their own proprietary CPU design.
You don't understand any of this. At all.
Get the clue that you will not be allowed to disseminate incorrect information at your every whim. And when you make stuff up with such an obviously deficient background it isn't difficult at all to contradict by those of us who do happen to have appropriate backgrounds.
Get the clue that you will not be allowed to disseminate incorrect information at your every whim. And when you make stuff up with such an obviously deficient background it isn't difficult at all to contradict by those of us who do happen to have appropriate backgrounds.
"Hummingbird represents an interesting business model. Samsung has a license for the Cortex-A8 from ARM, and a license for Hummingbird from Intrinsity. Intrinsity received a fee to develop Hummingbird, and both ARM and Intrinsity will collect royalties when the core is used in SoCs. Hummingbird belongs to Samsung; Intrinsity can?t license it, re-use it, or port it to another process."
It is the EXACT business model as the console chips --- IBM designed them, and Microsoft owns the IP.
Considering that the latest rumors floating around is Samsung going to use the Tegra 2 on its Galaxy Tab 2 --- which means that Samsung's own dual core A9 chip is not ready yet. And if Samsung is not ready, then probably Apple's own dual core A9 chip is not ready as well.
"Hummingbird represents an interesting business model. Samsung has a license for the Cortex-A8 from ARM, and a license for Hummingbird from Intrinsity. Intrinsity received a fee to develop Hummingbird, and both ARM and Intrinsity will collect royalties when the core is used in SoCs. Hummingbird belongs to Samsung; Intrinsity can’t license it, re-use it, or port it to another process."
It is the EXACT business model as the console chips --- IBM designed them, and Microsoft owns the IP.
Yes, Samsung licensed ARM IP from ARM (because you have to in order to make ARM chips) and custom IP from Intrinsity. They then hired Intrinsity to help apply that IP to the A8 that Samsung ALREADY BUILT AND USED IN OTHER PRODUCTS. Which part of this is hard to understand?
Of course Intrinsity had IP to use...otherwise what is the value add to Samsung to hire Intrinsity?
This in no way supports your idiotic contention that Samsung cannot build the A9 on its own, especially given it has already done so and is sampling the chip. Just like it did with it's original A8 product. Which part of Samsung is the 2nd largest semi company and largest app processor company confuses you? Moreover Samsung is a architecture license holder and Ti is not.
It is also not like the business model of console chips given that Samsung makes chips unlike Microsoft (or Sony for that matter). This is like saying that Intel followed the console business model because it had to get a license from ARM and buy the StrongARM IP from DEC and therefore didn't have the chops to design chips.
Considering that the latest rumors floating around is Samsung going to use the Tegra 2 on its Galaxy Tab 2 --- which means that Samsung's own dual core A9 chip is not ready yet. And if Samsung is not ready, then probably Apple's own dual core A9 chip is not ready as well.
The Tegra 2 is first to market so Samsung might go that route, especially for the gpu performance. On the other hand its the same rumor that has been circulating since earlier in the year.
Of course, if true, that completely demolishes your other argument of how Korean and Japanese companies act since this is Samsung Mobile dumping Samsung Semi in favor of nVidia who traditionally uses TSMC. Google has been pushing Tegra 2 for Honeycomb.
What does that mean for Apple? It depends. As an ARM license holder it can design the A5 alone and get any decent fab to build it for them. After all, it owns all of Intrinisty's IP.
It would be interesting if the A5 is being fabbed by Samsung Semi's new 32nm HKMG fab. Samsung announced their 32nm fab was ready for volume production and Samsung developed 32nm HKMG ARMs as a part of the CPA and was highlighted at TechCon 2010 by ARM.
So Samsung Mobile going Tegra 2 might really mean that 45nm Orion isn't ready and likely never will be ready because it'll go 32nm. Apple being Apple they probably pre-bought a huge amount of 32nm fab capacity from Samsung and Samsung sure as hell isn't going to fab its own processors at Global Foundries.
Yes, Samsung licensed ARM IP from ARM (because you have to in order to make ARM chips) and custom IP from Intrinsity. They then hired Intrinsity to help apply that IP to the A8 that Samsung ALREADY BUILT AND USED IN OTHER PRODUCTS. Which part of this is hard to understand?
Of course Intrinsity had IP to use...otherwise what is the value add to Samsung to hire Intrinsity?
This in no way supports your idiotic contention that Samsung cannot build the A9 on its own, especially given it has already done so and is sampling the chip. Just like it did with it's original A8 product. Which part of Samsung is the 2nd largest semi company and largest app processor company confuses you? Moreover Samsung is a architecture license holder and Ti is not.
It is also not like the business model of console chips given that Samsung makes chips unlike Microsoft (or Sony for that matter). This is like saying that Intel followed the console business model because it had to get a license from ARM and buy the StrongARM IP from DEC and therefore didn't have the chops to design chips.
All your talk about me getting it wrong that Samsung licensing the Hummingbird --- has been proven that you are wrong.
Another thing that you got it wrong is that you think Samsung is a architecture licensee. There are only 4 architecture licensees --- Marvell, Qualcomm, Infineon and Microsoft.
I never claim that Samsung can't build their own A9 --- I said that they can't do it within the time line given. And the recent rumors about Samsung going to use the tegra 2 on their own Galaxy Tab 2 tablet reinforces my arguments.
I never said that Samsung has a business model like Microsoft. I said that the arrangement between Intrinsity and Samsung is like IBM and Microsoft. Intrinsity/IBM did all the designing but Samsung/Microsoft owns the IP.
Quote:
Originally Posted by nht
The Tegra 2 is first to market so Samsung might go that route, especially for the gpu performance. On the other hand its the same rumor that has been circulating since earlier in the year.
Of course, if true, that completely demolishes your other argument of how Korean and Japanese companies act since this is Samsung Mobile dumping Samsung Semi in favor of nVidia who traditionally uses TSMC. Google has been pushing Tegra 2 for Honeycomb.
What does that mean for Apple? It depends. As an ARM license holder it can design the A5 alone and get any decent fab to build it for them. After all, it owns all of Intrinisty's IP.
It would be interesting if the A5 is being fabbed by Samsung Semi's new 32nm HKMG fab. Samsung announced their 32nm fab was ready for volume production and Samsung developed 32nm HKMG ARMs as a part of the CPA and was highlighted at TechCon 2010 by ARM.
So Samsung Mobile going Tegra 2 might really mean that 45nm Orion isn't ready and likely never will be ready because it'll go 32nm. Apple being Apple they probably pre-bought a huge amount of 32nm fab capacity from Samsung and Samsung sure as hell isn't going to fab its own processors at Global Foundries.
I was talking about time line, time line, time line. I never said that they couldn't design it. I never said that they couldn't manufacture it. I said that they can't design it within the time line given. I said that even if they can design it, they can't produce it in quantities that would satisfy Samsung's own internal demand, let alone produce in massive quantities that Apple demands.
Considering the fact that Cortex A15 is design for 32nm, by the time that the fabs are ready to do 32nm --- they would be manufacturing A15 chips then.
"Hummingbird represents an interesting business model. Samsung has a license for the Cortex-A8 from ARM, and a license for Hummingbird from Intrinsity. Intrinsity received a fee to develop Hummingbird, and both ARM and Intrinsity will collect royalties when the core is used in SoCs. Hummingbird belongs to Samsung; Intrinsity can?t license it, re-use it, or port it to another process."
It is the EXACT business model as the console chips --- IBM designed them, and Microsoft owns the IP.
Ummm. No. I find it very interesting how you cherry-pick which sentence to use, esopecially when the very next sentence of YOUR SOURCE says this :
Quote:
Hummingbird belongs to Samsung; Intrinsity can?t license it, re-use it, or port it to another process.
Taking statements out of context and gratuitous information hiding allows all kinds of whacky things to be said.It's know as being dishonest.
The model of core development in that article is exactly what I said it was, Samsung contracted Intrinsity to work on Hummingbird. The method of payment includes some royaly-like mechanism, but that didn't change the ownership of the Hummingbird core itself.
Ummm. No. I find it very interesting how you cherry-pick which sentence to use, esopecially when the very next sentence of YOUR SOURCE says this :
Taking statements out of context and gratuitous information hiding allows all kinds of whacky things to be said.It's know as being dishonest.
The model of core development in that article is exactly what I said it was, Samsung contracted Intrinsity to work on Hummingbird. The method of payment includes some royaly-like mechanism, but that didn't change the ownership of the Hummingbird core itself.
I didn't cherry-pick --- I specifically stated that it is the same as IBM designing the Xenon chip and Microsoft owning the IP. Microsoft bought an architecture license for the PowerPC chip, then hire IBM to design the Xenon. Microsoft keeps the intellectual property for the Xenon --- IBM can't sell the Xenon to another customer and can't license the Xenon to another customer.
All your talk about me getting it wrong that Samsung licensing the Hummingbird --- has been proven that you are wrong.
No this statement only proves you either A) cannot read for context or b) are so embroiled in your fantasy that you cannot admit you are still 100% wrong, so you delude yourself, or C) realize the core of B but are so egotistical that you intentionally attempt to keep up subterfuge to win a lost point by volume of falsehood and opportunistic out-of-context interpretation..
Quote:
Another thing that you got it wrong is that you think Samsung is a architecture licensee. There are only 4 architecture licensees --- Marvell, Qualcomm, Infineon and Microsoft.
You are clearly just making shit up here. And you don't even know what the flavors of ARM licenses are. As for contradiction of 4 licensees--from the ARM Ltd Annual report:
Quote:
In 2009, we signed 87 licences, the highest number of licences signed in a year. This takes the licensing base to more than 660 licences.
I never claim that Samsung can't build their own A9 --- I said that they can't do it within the time line given. And the recent rumors about Samsung going to use the tegra 2 on their own Galaxy Tab 2 tablet reinforces my arguments.
Hey, buying the farm on the advice of a rumor. Great way to lose your ass.
Quote:
I never said that Samsung has a business model like Microsoft. I said that the arrangement between Intrinsity and Samsung is like IBM and Microsoft. Intrinsity/IBM did all the designing but Samsung/Microsoft owns the IP.
Actually you still have it wrong, I said the Microsoft part on the 12th in this post:
Quote:
Originally Posted by Hiro
Microsoft did not design the Xenon. IBM did, completely. Microsoft contracted IBM to design and produce the chip, providing requirements and feedback in the process. IBM started with their existing PPC muli-core design and then wrapped three of them together with VMX optimizations and communications crossbars to meet the MS XBox CPU spec requirements. Microsoft didn't license and design anything, they bought finished, custom specified hardware. That's all.
Are you so vacuous that you can't remember what arguments were used against you and suddenly think they sprung from your own brain? You don't get credit for this, But I think its absolutely hilarious that you now think the stick that beat you is your savior.
The rest of that post dealt with associated issues. IBM still owns the Xenon and its IP, by license they just cannot sell the Xenon product to anyone else. In contrast Samsung OWNS the Hummingbird product, it isn't Intrinsity's to sell.
Quote:
I was talking about time line, time line, time line. I never said that they couldn't design it. I never said that they couldn't manufacture it. I said that they can't design it within the time line given. I said that even if they can design it, they can't produce it in quantities that would satisfy Samsung's own internal demand, let alone produce in massive quantities that Apple demands.
Shifting goalposts. Everything else you have said is so broken or malappropriated this is inconsequential.
Quote:
Considering the fact that Cortex A15 is design for 32nm, by the time that the fabs are ready to do 32nm --- they would be manufacturing A15 chips then.
Coretx A15 is a desktop/plug-in laptop class chip. Not a mobile device centric chip. Your lack of ability to realize that impeaches all you have to say regarding the relationships between desktop chip manufacturing and mobile CPU manufacturing.
I cannot I say this any more clearly. You are entirely out of your element here.
You are exhibiting all the weaknesses of a typical internet blowhard: 1) Find an article in the internet via Google; 2) Quote a sentence out of context to use as support for some assertion you really have no background in; 3) Screw it up by not having understood the quoted precisely because you are lacking the background necessary to actually understand the article, and so missing important, relevant, information; 3a) or maybe you do understand it, but hope others are sloppy and just believe you because you say so; 4) Repeat as often as necessary hoping the masses will get tired and let you "win".
No this statement only proves you either A) cannot read for context or b) are so embroiled in your fantasy that you cannot admit you are still 100% wrong, so you delude yourself, or C) realize the core of B but are so egotistical that you intentionally attempt to keep up subterfuge to win a lost point by volume of falsehood and opportunistic out-of-context interpretation..
I am right in this case. It has nothing to do with being egotistical --- it has to do with you two trying to slam me for not understanding the business arrangement. But I was right all along.
Quote:
Originally Posted by Hiro
You are clearly just making shit up here. And you don't even know what the flavors of ARM licenses are. As for contradiction of 4 licensees--from the ARM Ltd Annual report:
You don't even know what I am talking about. All the other licensees are "core" licensees --- which means that they take a "stock" cortex A8 core and add a couple of non-core stuff to it (like an GPU). What these other 80-90 licensees cannot do is to change a single gate on the cortex core itself.
There are only 4 ARM "architecture" licensees --- Marvell, Infineon, Qualcomm and Microsoft. They design their own "compatible" cores. Marvell has the Sheeva core and Qualcomm has the Scorpion core. This is like AMD selling a x86 compatible CPU.
Hummingbird has a stock A8 core vs. Qualcomm has a ARMv7 compatible core.
Another thing that you got it wrong is that you think Samsung is a architecture licensee. There are only 4 architecture licensees --- Marvell, Qualcomm, Infineon and Microsoft.
This was reported by a single blogger stating that McGuire told her this. Not that ARM or McGuire stated this at a conference call unless you have a link that shows otherwise.
In 2008 ARM CEO East stated that a leading handset OEM had signed an architecture license during the ARM 2Q earnings call. Not reported by a ZDNet blogger second hand.
Marvell, Qualcomm and Infineon are not handset OEMs. Samsung and Apple are and some folks believe this to be Samsung and others Apple. Of the two Apple is most likely but either way there's more than the 4 stated. You are correct that Samsung isn't a confirmed holder. My bad. On the other hand it's not likely to be anyone other than Samsung or Apple.
"And looking ahead, we can see further designs out in the future based on the licensing activity of today within this set of results. We are talking about an architecture license with a leading OEM for both current and future ARM technology. So, that means do not get excited about any revenue on this deal because it is all tied up with future technology and the revenue will be recognized over several years, but it is very important as far as we are concerned in terms of securing design-ins with that particular OEM and also in the mobile space altogether"
...
Q&A
Unidentified Audience Member
Yes, good morning. The first question would be on the OEM deals you signed in the quarter. How should we think about those from a revenue perspective and how do they actually work in terms of what the OEMs are getting from you?
Warren East - ARM Holdings plc - CEO
Okay. I mean from a revenue perspective, as I said, think about that as -- because of the nature of the products licensed actually, it's revenue over a period of several years. So, not a material impact on short-term revenue on a quarter-by-quarter basis.
And in terms of what people actually get, well again it varies. The one deal that we talked about this morning is actually an architecture license. The other deals are various, actually, from design-only licenses through to design and ability to manufacture chips.
"Unidentified Audience Member:
Morning, morning. Just a follow-up question on the earlier one about the OEM deals. Particularly, what's the significance that the deal is with the handset OEM rather than the chipmaker, firstly? And then, I guess, related to that, you have been reasonably restrained about Intel this morning. Would you also expect to see them pursue similar deals with kind of handset OEMs, top tier names potentially, any view on that?
Warren East - ARM Holdings plc - CEO:
Yes, two questions. So on that particular transaction, we are not really disclosing any more about that transaction than is in our earnings release and we have already disclosed. Some handset manufacturers want to have more control over the design of their handset, including the components within it, than others. And it's as simple as that. And the ARM business model offers one that level of control if that's what one wants to do and one has the technical resources available to do it. And that's a question for each and every handset manufacturer. So, sorry, I can't really be very forthcoming about that. "
There is at LEAST one more than your list of four. An architecture license deal had been concluded with a handset OEM in 2008. Microsoft isn't a handset OEM either.
Quote:
I never claim that Samsung can't build their own A9 --- I said that they can't do it within the time line given.
The timeline is July of 2011. 7 and a half months from now for the iPhone 5. iPad is closer but given Apple's penchant for secrecy Samsung could be fabbing now.
Quote:
I was talking about time line, time line, time line. I never said that they couldn't design it. I never said that they couldn't manufacture it. I said that they can't design it within the time line given.
Design is done. They're sampling now. Orion might get another spin before full production but the design itself is done for both Orion and Saratoga. ARM made a big deal about 32nm at their conference so given Saratoga sampled in Feb and ARM released a POP I think the 32nm A9 design is ready.
Quote:
I said that even if they can design it, they can't produce it in quantities that would satisfy Samsung's own internal demand, let alone produce in massive quantities that Apple demands.
That Samsung Mobile is using Tegra doesn't imply that Apple isn't getting priority manufacturing for the 2011 iPad and iPhones. Again, Samsung favored production of the A9 for the iPad ahead of production of the Hummingbird for their own phones.
Quote:
Considering the fact that Cortex A15 is design for 32nm, by the time that the fabs are ready to do 32nm --- they would be manufacturing A15 chips then.
Wrong. The 32nm fab qual'd in June and Samsung's Lin stated at the November TechCon it's ready for full rate production. ARM has specific A9 enhancements (POP) for Samsung's 32nm process ready for license and they showed working silicon at TechCon. The A15 is still a ways off.
If Samsung is currently in full rate production in their 32nm fab Apple could have enough for an iPad 2 launch in March/April. Supplies will be constrained but they can still sell the current iPad as the lowest end model with the two higher end models being the iPad 2 like they do with the iPhone 3GS and iPhone 4.
I'm starting to think that 32nm is actually the most likely scenario. Apple didn't buy Intrinsity, PA Semi and trot out the A4 to let 45nm Tegra 2 based Android tablets and phones have an advantage until 2H 2011 and miss their desired iPad and iPhone refreshes.
Comments
Given that stock cortex a8 chips were available one year before both Samsung and Apple released their version, you are saying that Samsung and Apple have caught up with the big silicon companies in one year's time.
Wow, way to miss the point and say something else unrelated and nonsensical. You must be channelling the BB CEO.
I'm saying that despite what you wrote Samsung already gave Apple priority for the A4 OVER their own Hummingbird product. Therefore what you wrote was stupid and therefore you completely evaded this by writing another stupid statement to attempt to divert attention.
Samsung is a "big silicon company". More so than Ti. So what you wrote was once again factually incorrect and stupid. Samsung didn't "catch up" to anyone. Intrinisty and Samsung enhanced their Samsung's A8 product used in the 3GS to be one of the very best SoC of the A8 generation: the Hummingbird.
And before you embarrass yourself further, the reason why Samsung > Ti is because Samsung is sampling 32nm A9s where Ti isn't going to move past 45nm themselves but probably end up using TSMC. Half of Ti's production was already outsourced in 2007. It's probably higher now and even more so when they go fabless for 32nm and beyond. So much for "big silicon".
Meanwhile:
"South Korea's Samsung Electronics Co. Ltd. said its foundry business has qualified a 32-nm low-power process with high-k metal gate technology. The company lays claim to being the first foundry to qualify a low-power process using high-k at 32-nm."
http://www.eetimes.com/electronics-n...oundry-process
Absolutely clueless about Samsung aren't you?
Given that stock cortex a8 chips were available one year before both Samsung and Apple released their version, you are saying that Samsung and Apple have caught up with the big silicon companies in one year's time.
You don't even know what the ARM Cortex line is do you. (yes that was a statement not a question)
ARM licenses IP, not chips. Companies contract to use the IP to make the chips with whatever customizations they want.
There is no such thing as a stock Cortex chip!
They don't exist!
Because the Cortex IP itself isn't the entire package of IP needed to generate a functional silicon CPU!
The IP was available in 2008, the first complete shipping implementations (not counting pre production tape-outs) didn't show until Feb/Mar 2009.
Anywhere.
You.
Still.
Fail...
As for big silicon. I only consider Intel and AMD in that company. If you haven't shipped 25nm or smaller memory runs you don't qualify as big silicon. 32nm is a generation old tech. Cheap to build the fab for that, only about a $ Billion and a quarter for the machinery. You want big silicon you don't start until your fab runs $3 Billion for less than 25 nm.
You don't even know what the ARM Cortex line is do you. (yes that was a statement not a question)
ARM licenses IP, not chips. Companies contract to use the IP to make the chips with whatever customizations they want.
There is no such thing as a stock Cortex chip!
They don't exist!
Because the Cortex IP itself isn't the entire package of IP needed to generate a functional silicon CPU!
The IP was available in 2008, the first complete shipping implementations (not counting pre production tape-outs) didn't show until Feb/Mar 2009.
Anywhere.
You.
Still.
Fail...
As for big silicon. I only consider Intel and AMD in that company. If you haven't shipped 25nm or smaller memory runs you don't qualify as big silicon. 32nm is a generation old tech. Cheap to build the fab for that, only about a $ Billion and a quarter for the machinery. You want big silicon you don't start until your fab runs $3 Billion for less than 25 nm.
I know that only 4 companies to-date has signed ARM architecture license --- Qualcomm, Infineon, Marvell and Microsoft. They can design their own cores.
I know that all the other players license the cores and put their own extra stuff around it.
I know that TI --- which has a long history of designing application processors --- will barely able to produce low 6 figure quantities of OMAP4430 dual core cortex A9 application processors for the RIM Playbook.
I also know that Samsung while they are big in silicon, they are heavy on memory chips, not actual application processors. That's why Samsung licensed the Hummerbird.
Now you are telling me that Samsung and Apple --- while they are both big companies with lots of cash ---- both are pretty new to this industry and somehow they are going to do even better TI --- which is to produce millions of dual core cortex A9 application processors for the iPad 2.
I know that only 4 companies to-date has signed ARM architecture license --- Qualcomm, Infineon, Marvell and Microsoft. They can design their own cores.
I know that all the other players license the cores and put their own extra stuff around it.
I know that TI --- which has a long history of designing application processors --- will barely able to produce low 6 figure quantities of OMAP4430 dual core cortex A9 application processors for the RIM Playbook.
I also know that Samsung while they are big in silicon, they are heavy on memory chips, not actual application processors. That's why Samsung licensed the Hummerbird.
Now you are telling me that Samsung and Apple --- while they are both big companies with lots of cash ---- both are pretty new to this industry and somehow they are going to do even better TI --- which is to produce millions of dual core cortex A9 application processors for the iPad 2.
You are showing yourself to be an absolute talk out your ass sideways idiot. Apple is an ARM licensor, as is TI- what do you think the OMAP series are? And Samsung is a licensor too--Hummingbird isn't made or designed by ARM, the CortexA8 core was licensed by Samsung so they could make the Hummingbird. There are also a handful of DoD and European defense contractors that contract for ARM IP directly too.
Your statements only highlight your utter lack of knowledge as to what ARM sells and what licensors get. At least you keep creeping a little closer to reality each time you try to cover your arse, but the rate is abysmally low given the learning opportunities you have been provided.
What made up un-facts will you try next?
You are showing yourself to be an absolute talk out your ass sideways idiot. Apple is an ARM licensor, as is TI- what do you think the OMAP series are? And Samsung is a licensor too--Hummingbird isn't made or designed by ARM, the CortexA8 core was licensed by Samsung so they could make the Hummingbird. There are also a handful of DoD and European defense contractors that contract for ARM IP directly too.
Your statements only highlight your utter lack of knowledge as to what ARM sells and what licensors get. At least you keep creeping a little closer to reality each time you try to cover your arse, but the rate is abysmally low given the learning opportunities you have been provided.
What made up un-facts will you try next?
I never said otherwise.
I said that big merchant established chip companies are barely able to produce low quanities of dual core cortex A9 application processors --- these are stock merchant chips. Now you are telling me that Samsung and Apple who are both relatively new to the scene are going to be virtually certain to provide millions of these same kind of chips at the same time as TI. Remember these are newcomers who didn't have the expertise and had to license the Hummerbird from someone else.
Is it possible? Sure, it is possible. But you are making it sound as if it is a virtual certainty, and I disagree with that assessment.
I know that only 4 companies to-date has signed ARM architecture license --- Qualcomm, Infineon, Marvell and Microsoft.
You are showing yourself to be an absolute talk out your ass sideways idiot. Apple is an ARM licensor, as is TI- what do you think the OMAP series are? And Samsung is a licensor too--Hummingbird isn't made or designed by ARM, the CortexA8 core was licensed by Samsung so they could make the Hummingbird. There are also a handful of DoD and European defense contractors that contract for ARM IP directly too.
I never said otherwise.
wow.
just wow.
Either you have the worlds shortest memory or convenient lying is considered a valid debate tactic in your world. Your post's above are QFT.
I said that big merchant established chip companies are barely able to produce low quanities of dual core cortex A9 application processors --- these are stock merchant chips. Now you are telling me that Samsung and Apple who are both relatively new to the scene are going to be virtually certain to provide millions of these same kind of chips at the same time as TI. Remember these are newcomers who didn't have the expertise and had to license the Hummerbird from someone else.
Is it possible? Sure, it is possible. But you are making it sound as if it is a virtual certainty, and I disagree with that assessment.
You keep saying that, but you have no evidence. You are just making up words hoping something will stick.
There are shipping CortexA9 chips in the wild. Not small feature-size production runs for low power, but there isn't any reason to expect that at less than the 1 year point from IP availability. The implementations are actually going swimmingly well based on both current producers of plug-in device cores shipping well respected units. This means there are no architectural gotchas for the small feature size crowd to deal with, they only have to deal with setting up the fab process.
Apple isn't a fab, so your assessment of their experience in producing chips is utterly irrelevant and worthless. Because Apple simply doesn't do that! And the opposite side of that, the diss on Samsung is way off base too, because Samsung is a proven 32nm memory producer for a couple years now, meaning they have the proven technical ability to run the fab at the feature size the low-power CortexA9 based cores will be fabricated on.
I know that only 4 companies to-date has signed ARM architecture license --- Qualcomm, Infineon, Marvell and Microsoft. They can design their own cores.
This is completely wrong as usual. Intel, Apple, and Samsung are also architecture license holders. Apple is unconfirmed given their traditional secrecy but most folks believe so. Especially given ARM's coy statement and Apple's early history in creating ARM for Newton with Acorn.
I know that all the other players license the cores and put their own extra stuff around it.
I know that TI --- which has a long history of designing application processors --- will barely able to produce low 6 figure quantities of OMAP4430 dual core cortex A9 application processors for the RIM Playbook.
This seems highly unlikely. Given your penchant for making "facts" up I'd want a link.
I also know that Samsung while they are big in silicon, they are heavy on memory chips, not actual application processors.
See what I mean? A made up "fact". Samsung is only the second largest semiconductor company behind intel and the largest supplier of application processors.
"Scemama notes that Samsung is the world's largest vendor of application processors."
http://www.eetimes.com/electronics-n...--says-analyst
That's why Samsung licensed the Hummerbird.
Again wrong. They didn't license Hunningbird. They designed it in collaboration with Intrinsity.
Now you are telling me that Samsung and Apple --- while they are both big companies with lots of cash ---- both are pretty new to this industry and somehow they are going to do even better TI --- which is to produce millions of dual core cortex A9 application processors for the iPad 2.
Wrong again. Neither Samsung nor Apple are new. Apple has been a fabless designer for a while (mostly northbridges prior to buying pa semi but they also owned a big chunk of ARM). And samsung didn't become the second largest semi company expected to overtake Intel overnight. Samsung passed Ti a while ago.
I don't think you wrote a single correct statement in this entire post.
Again wrong. They didn't license Hunningbird. They designed it in collaboration with Intrinsity.
That is corporate talk semantics. It's like the xbox CPU's. Microsoft licensed the IBM powerpc architecture and then hire IBM to help them design the xenon cpu. The end result is that the CPU itself is a Microsoft's own IP. Who did most of the work? It's IBM.
That is corporate talk semantics. It's like the xbox CPU's. Microsoft licensed the IBM powerpc architecture and then hire IBM to help them design the xenon cpu. The end result is that the CPU itself is a Microsoft's own IP. Who did most of the work? It's IBM.
You are completely wrong. On all counts, well except that it was IBM that did the work, but sometimes even a blind squirrel finds a nut.
Once again you prove that you can make crap up and spout it off with the best of them, being utterly incorrect on any important point the whole while.
Microsoft did not design the Xenon. IBM did, completely. Microsoft contracted IBM to design and produce the chip, providing requirements and feedback in the process. IBM started with their existing PPC muli-core design and then wrapped three of them together with VMX optimizations and communications crossbars to meet the MS XBox CPU spec requirements. Microsoft didn't license and design anything, they bought finished, custom specified hardware. That's all.
Don't let yourself be confused by MS having a couple employees working with IBM. That doesn't constitute design any more than it did between Apple, Motorola and IBM when the AltiVec/VMX design was created. Apple's engineers had the grand picture and knew what they wanted working with Moto and IBM every day. They were even very capable, participating completely in the architecture process. But all the actual decisions were Moto's/IBM's, so very explicitly AltiVec belongs only to Moto and is a Moto design just as VMX is for IBM.
Your idea of corporate semantics is the thing that's f*%-ed up here. Samsung DID design the Hummingbird (not the Hummerbird as you have repeatedly called it throughout the thread). Samsung licensed the core IP from ARM, then contracted for engineering assistance from Intrinsity to create a complete CPU design. After they were done Samsung fabbed their own proprietary CPU design.
You don't understand any of this. At all.
Get the clue that you will not be allowed to disseminate incorrect information at your every whim. And when you make stuff up with such an obviously deficient background it isn't difficult at all to contradict by those of us who do happen to have appropriate backgrounds.
Get the clue that you will not be allowed to disseminate incorrect information at your every whim. And when you make stuff up with such an obviously deficient background it isn't difficult at all to contradict by those of us who do happen to have appropriate backgrounds.
http://www.insidedsp.com/tabid/64/ar...0/Default.aspx
Quote:
"Hummingbird represents an interesting business model. Samsung has a license for the Cortex-A8 from ARM, and a license for Hummingbird from Intrinsity. Intrinsity received a fee to develop Hummingbird, and both ARM and Intrinsity will collect royalties when the core is used in SoCs. Hummingbird belongs to Samsung; Intrinsity can?t license it, re-use it, or port it to another process."
It is the EXACT business model as the console chips --- IBM designed them, and Microsoft owns the IP.
http://www.insidedsp.com/tabid/64/ar...0/Default.aspx
Quote:
"Hummingbird represents an interesting business model. Samsung has a license for the Cortex-A8 from ARM, and a license for Hummingbird from Intrinsity. Intrinsity received a fee to develop Hummingbird, and both ARM and Intrinsity will collect royalties when the core is used in SoCs. Hummingbird belongs to Samsung; Intrinsity can’t license it, re-use it, or port it to another process."
It is the EXACT business model as the console chips --- IBM designed them, and Microsoft owns the IP.
Yes, Samsung licensed ARM IP from ARM (because you have to in order to make ARM chips) and custom IP from Intrinsity. They then hired Intrinsity to help apply that IP to the A8 that Samsung ALREADY BUILT AND USED IN OTHER PRODUCTS. Which part of this is hard to understand?
Of course Intrinsity had IP to use...otherwise what is the value add to Samsung to hire Intrinsity?
This in no way supports your idiotic contention that Samsung cannot build the A9 on its own, especially given it has already done so and is sampling the chip. Just like it did with it's original A8 product. Which part of Samsung is the 2nd largest semi company and largest app processor company confuses you? Moreover Samsung is a architecture license holder and Ti is not.
It is also not like the business model of console chips given that Samsung makes chips unlike Microsoft (or Sony for that matter). This is like saying that Intel followed the console business model because it had to get a license from ARM and buy the StrongARM IP from DEC and therefore didn't have the chops to design chips.
Considering that the latest rumors floating around is Samsung going to use the Tegra 2 on its Galaxy Tab 2 --- which means that Samsung's own dual core A9 chip is not ready yet. And if Samsung is not ready, then probably Apple's own dual core A9 chip is not ready as well.
The Tegra 2 is first to market so Samsung might go that route, especially for the gpu performance. On the other hand its the same rumor that has been circulating since earlier in the year.
Of course, if true, that completely demolishes your other argument of how Korean and Japanese companies act since this is Samsung Mobile dumping Samsung Semi in favor of nVidia who traditionally uses TSMC. Google has been pushing Tegra 2 for Honeycomb.
What does that mean for Apple? It depends. As an ARM license holder it can design the A5 alone and get any decent fab to build it for them. After all, it owns all of Intrinisty's IP.
It would be interesting if the A5 is being fabbed by Samsung Semi's new 32nm HKMG fab. Samsung announced their 32nm fab was ready for volume production and Samsung developed 32nm HKMG ARMs as a part of the CPA and was highlighted at TechCon 2010 by ARM.
So Samsung Mobile going Tegra 2 might really mean that 45nm Orion isn't ready and likely never will be ready because it'll go 32nm. Apple being Apple they probably pre-bought a huge amount of 32nm fab capacity from Samsung and Samsung sure as hell isn't going to fab its own processors at Global Foundries.
Yes, Samsung licensed ARM IP from ARM (because you have to in order to make ARM chips) and custom IP from Intrinsity. They then hired Intrinsity to help apply that IP to the A8 that Samsung ALREADY BUILT AND USED IN OTHER PRODUCTS. Which part of this is hard to understand?
Of course Intrinsity had IP to use...otherwise what is the value add to Samsung to hire Intrinsity?
This in no way supports your idiotic contention that Samsung cannot build the A9 on its own, especially given it has already done so and is sampling the chip. Just like it did with it's original A8 product. Which part of Samsung is the 2nd largest semi company and largest app processor company confuses you? Moreover Samsung is a architecture license holder and Ti is not.
It is also not like the business model of console chips given that Samsung makes chips unlike Microsoft (or Sony for that matter). This is like saying that Intel followed the console business model because it had to get a license from ARM and buy the StrongARM IP from DEC and therefore didn't have the chops to design chips.
All your talk about me getting it wrong that Samsung licensing the Hummingbird --- has been proven that you are wrong.
Another thing that you got it wrong is that you think Samsung is a architecture licensee. There are only 4 architecture licensees --- Marvell, Qualcomm, Infineon and Microsoft.
I never claim that Samsung can't build their own A9 --- I said that they can't do it within the time line given. And the recent rumors about Samsung going to use the tegra 2 on their own Galaxy Tab 2 tablet reinforces my arguments.
I never said that Samsung has a business model like Microsoft. I said that the arrangement between Intrinsity and Samsung is like IBM and Microsoft. Intrinsity/IBM did all the designing but Samsung/Microsoft owns the IP.
The Tegra 2 is first to market so Samsung might go that route, especially for the gpu performance. On the other hand its the same rumor that has been circulating since earlier in the year.
Of course, if true, that completely demolishes your other argument of how Korean and Japanese companies act since this is Samsung Mobile dumping Samsung Semi in favor of nVidia who traditionally uses TSMC. Google has been pushing Tegra 2 for Honeycomb.
What does that mean for Apple? It depends. As an ARM license holder it can design the A5 alone and get any decent fab to build it for them. After all, it owns all of Intrinisty's IP.
It would be interesting if the A5 is being fabbed by Samsung Semi's new 32nm HKMG fab. Samsung announced their 32nm fab was ready for volume production and Samsung developed 32nm HKMG ARMs as a part of the CPA and was highlighted at TechCon 2010 by ARM.
So Samsung Mobile going Tegra 2 might really mean that 45nm Orion isn't ready and likely never will be ready because it'll go 32nm. Apple being Apple they probably pre-bought a huge amount of 32nm fab capacity from Samsung and Samsung sure as hell isn't going to fab its own processors at Global Foundries.
I was talking about time line, time line, time line. I never said that they couldn't design it. I never said that they couldn't manufacture it. I said that they can't design it within the time line given. I said that even if they can design it, they can't produce it in quantities that would satisfy Samsung's own internal demand, let alone produce in massive quantities that Apple demands.
Considering the fact that Cortex A15 is design for 32nm, by the time that the fabs are ready to do 32nm --- they would be manufacturing A15 chips then.
http://www.insidedsp.com/tabid/64/ar...0/Default.aspx
Quote:
"Hummingbird represents an interesting business model. Samsung has a license for the Cortex-A8 from ARM, and a license for Hummingbird from Intrinsity. Intrinsity received a fee to develop Hummingbird, and both ARM and Intrinsity will collect royalties when the core is used in SoCs. Hummingbird belongs to Samsung; Intrinsity can?t license it, re-use it, or port it to another process."
It is the EXACT business model as the console chips --- IBM designed them, and Microsoft owns the IP.
Ummm. No. I find it very interesting how you cherry-pick which sentence to use, esopecially when the very next sentence of YOUR SOURCE says this :
Hummingbird belongs to Samsung; Intrinsity can?t license it, re-use it, or port it to another process.
Taking statements out of context and gratuitous information hiding allows all kinds of whacky things to be said.It's know as being dishonest.
The model of core development in that article is exactly what I said it was, Samsung contracted Intrinsity to work on Hummingbird. The method of payment includes some royaly-like mechanism, but that didn't change the ownership of the Hummingbird core itself.
Ummm. No. I find it very interesting how you cherry-pick which sentence to use, esopecially when the very next sentence of YOUR SOURCE says this :
Taking statements out of context and gratuitous information hiding allows all kinds of whacky things to be said.It's know as being dishonest.
The model of core development in that article is exactly what I said it was, Samsung contracted Intrinsity to work on Hummingbird. The method of payment includes some royaly-like mechanism, but that didn't change the ownership of the Hummingbird core itself.
I didn't cherry-pick --- I specifically stated that it is the same as IBM designing the Xenon chip and Microsoft owning the IP. Microsoft bought an architecture license for the PowerPC chip, then hire IBM to design the Xenon. Microsoft keeps the intellectual property for the Xenon --- IBM can't sell the Xenon to another customer and can't license the Xenon to another customer.
All your talk about me getting it wrong that Samsung licensing the Hummingbird --- has been proven that you are wrong.
No this statement only proves you either A) cannot read for context or b) are so embroiled in your fantasy that you cannot admit you are still 100% wrong, so you delude yourself, or C) realize the core of B but are so egotistical that you intentionally attempt to keep up subterfuge to win a lost point by volume of falsehood and opportunistic out-of-context interpretation..
Another thing that you got it wrong is that you think Samsung is a architecture licensee. There are only 4 architecture licensees --- Marvell, Qualcomm, Infineon and Microsoft.
You are clearly just making shit up here. And you don't even know what the flavors of ARM licenses are. As for contradiction of 4 licensees--from the ARM Ltd Annual report:
In 2009, we signed 87 licences, the highest number of licences signed in a year. This takes the licensing base to more than 660 licences.
http://www.arm.com/annualreport09/
I never claim that Samsung can't build their own A9 --- I said that they can't do it within the time line given. And the recent rumors about Samsung going to use the tegra 2 on their own Galaxy Tab 2 tablet reinforces my arguments.
Hey, buying the farm on the advice of a rumor. Great way to lose your ass.
I never said that Samsung has a business model like Microsoft. I said that the arrangement between Intrinsity and Samsung is like IBM and Microsoft. Intrinsity/IBM did all the designing but Samsung/Microsoft owns the IP.
Actually you still have it wrong, I said the Microsoft part on the 12th in this post:
Microsoft did not design the Xenon. IBM did, completely. Microsoft contracted IBM to design and produce the chip, providing requirements and feedback in the process. IBM started with their existing PPC muli-core design and then wrapped three of them together with VMX optimizations and communications crossbars to meet the MS XBox CPU spec requirements. Microsoft didn't license and design anything, they bought finished, custom specified hardware. That's all.
Are you so vacuous that you can't remember what arguments were used against you and suddenly think they sprung from your own brain? You don't get credit for this, But I think its absolutely hilarious that you now think the stick that beat you is your savior.
The rest of that post dealt with associated issues. IBM still owns the Xenon and its IP, by license they just cannot sell the Xenon product to anyone else. In contrast Samsung OWNS the Hummingbird product, it isn't Intrinsity's to sell.
I was talking about time line, time line, time line. I never said that they couldn't design it. I never said that they couldn't manufacture it. I said that they can't design it within the time line given. I said that even if they can design it, they can't produce it in quantities that would satisfy Samsung's own internal demand, let alone produce in massive quantities that Apple demands.
Shifting goalposts. Everything else you have said is so broken or malappropriated this is inconsequential.
Considering the fact that Cortex A15 is design for 32nm, by the time that the fabs are ready to do 32nm --- they would be manufacturing A15 chips then.
Coretx A15 is a desktop/plug-in laptop class chip. Not a mobile device centric chip. Your lack of ability to realize that impeaches all you have to say regarding the relationships between desktop chip manufacturing and mobile CPU manufacturing.
I cannot I say this any more clearly. You are entirely out of your element here.
You are exhibiting all the weaknesses of a typical internet blowhard: 1) Find an article in the internet via Google; 2) Quote a sentence out of context to use as support for some assertion you really have no background in; 3) Screw it up by not having understood the quoted precisely because you are lacking the background necessary to actually understand the article, and so missing important, relevant, information; 3a) or maybe you do understand it, but hope others are sloppy and just believe you because you say so; 4) Repeat as often as necessary hoping the masses will get tired and let you "win".
Are you really gunning for a #1 here: http://www.cracked.com/funny-3809-in...nt-techniques/
No this statement only proves you either A) cannot read for context or b) are so embroiled in your fantasy that you cannot admit you are still 100% wrong, so you delude yourself, or C) realize the core of B but are so egotistical that you intentionally attempt to keep up subterfuge to win a lost point by volume of falsehood and opportunistic out-of-context interpretation..
I am right in this case. It has nothing to do with being egotistical --- it has to do with you two trying to slam me for not understanding the business arrangement. But I was right all along.
You are clearly just making shit up here. And you don't even know what the flavors of ARM licenses are. As for contradiction of 4 licensees--from the ARM Ltd Annual report:
http://www.arm.com/annualreport09/
You don't even know what I am talking about. All the other licensees are "core" licensees --- which means that they take a "stock" cortex A8 core and add a couple of non-core stuff to it (like an GPU). What these other 80-90 licensees cannot do is to change a single gate on the cortex core itself.
http://www.theregister.co.uk/2010/07/23/microsoft_arm/
There are only 4 ARM "architecture" licensees --- Marvell, Infineon, Qualcomm and Microsoft. They design their own "compatible" cores. Marvell has the Sheeva core and Qualcomm has the Scorpion core. This is like AMD selling a x86 compatible CPU.
Hummingbird has a stock A8 core vs. Qualcomm has a ARMv7 compatible core.
Another thing that you got it wrong is that you think Samsung is a architecture licensee. There are only 4 architecture licensees --- Marvell, Qualcomm, Infineon and Microsoft.
This was reported by a single blogger stating that McGuire told her this. Not that ARM or McGuire stated this at a conference call unless you have a link that shows otherwise.
In 2008 ARM CEO East stated that a leading handset OEM had signed an architecture license during the ARM 2Q earnings call. Not reported by a ZDNet blogger second hand.
Marvell, Qualcomm and Infineon are not handset OEMs. Samsung and Apple are and some folks believe this to be Samsung and others Apple. Of the two Apple is most likely but either way there's more than the 4 stated. You are correct that Samsung isn't a confirmed holder. My bad. On the other hand it's not likely to be anyone other than Samsung or Apple.
"And looking ahead, we can see further designs out in the future based on the licensing activity of today within this set of results. We are talking about an architecture license with a leading OEM for both current and future ARM technology. So, that means do not get excited about any revenue on this deal because it is all tied up with future technology and the revenue will be recognized over several years, but it is very important as far as we are concerned in terms of securing design-ins with that particular OEM and also in the mobile space altogether"
...
Q&A
Unidentified Audience Member
Yes, good morning. The first question would be on the OEM deals you signed in the quarter. How should we think about those from a revenue perspective and how do they actually work in terms of what the OEMs are getting from you?
Warren East - ARM Holdings plc - CEO
Okay. I mean from a revenue perspective, as I said, think about that as -- because of the nature of the products licensed actually, it's revenue over a period of several years. So, not a material impact on short-term revenue on a quarter-by-quarter basis.
And in terms of what people actually get, well again it varies. The one deal that we talked about this morning is actually an architecture license. The other deals are various, actually, from design-only licenses through to design and ability to manufacture chips.
"Unidentified Audience Member:
Morning, morning. Just a follow-up question on the earlier one about the OEM deals. Particularly, what's the significance that the deal is with the handset OEM rather than the chipmaker, firstly? And then, I guess, related to that, you have been reasonably restrained about Intel this morning. Would you also expect to see them pursue similar deals with kind of handset OEMs, top tier names potentially, any view on that?
Warren East - ARM Holdings plc - CEO:
Yes, two questions. So on that particular transaction, we are not really disclosing any more about that transaction than is in our earnings release and we have already disclosed. Some handset manufacturers want to have more control over the design of their handset, including the components within it, than others. And it's as simple as that. And the ARM business model offers one that level of control if that's what one wants to do and one has the technical resources available to do it. And that's a question for each and every handset manufacturer. So, sorry, I can't really be very forthcoming about that. "
http://media.corporate-ir.net/media_...transcript.pdf
There is at LEAST one more than your list of four. An architecture license deal had been concluded with a handset OEM in 2008. Microsoft isn't a handset OEM either.
I never claim that Samsung can't build their own A9 --- I said that they can't do it within the time line given.
The timeline is July of 2011. 7 and a half months from now for the iPhone 5. iPad is closer but given Apple's penchant for secrecy Samsung could be fabbing now.
I was talking about time line, time line, time line. I never said that they couldn't design it. I never said that they couldn't manufacture it. I said that they can't design it within the time line given.
Design is done. They're sampling now. Orion might get another spin before full production but the design itself is done for both Orion and Saratoga. ARM made a big deal about 32nm at their conference so given Saratoga sampled in Feb and ARM released a POP I think the 32nm A9 design is ready.
I said that even if they can design it, they can't produce it in quantities that would satisfy Samsung's own internal demand, let alone produce in massive quantities that Apple demands.
That Samsung Mobile is using Tegra doesn't imply that Apple isn't getting priority manufacturing for the 2011 iPad and iPhones. Again, Samsung favored production of the A9 for the iPad ahead of production of the Hummingbird for their own phones.
Considering the fact that Cortex A15 is design for 32nm, by the time that the fabs are ready to do 32nm --- they would be manufacturing A15 chips then.
Wrong. The 32nm fab qual'd in June and Samsung's Lin stated at the November TechCon it's ready for full rate production. ARM has specific A9 enhancements (POP) for Samsung's 32nm process ready for license and they showed working silicon at TechCon. The A15 is still a ways off.
If Samsung is currently in full rate production in their 32nm fab Apple could have enough for an iPad 2 launch in March/April. Supplies will be constrained but they can still sell the current iPad as the lowest end model with the two higher end models being the iPad 2 like they do with the iPhone 3GS and iPhone 4.
I'm starting to think that 32nm is actually the most likely scenario. Apple didn't buy Intrinsity, PA Semi and trot out the A4 to let 45nm Tegra 2 based Android tablets and phones have an advantage until 2H 2011 and miss their desired iPad and iPhone refreshes.