Samsung claims primary seat in Apple iPhone
Samsung Electronics has added iPhone to a rapidly expanding list of consumer electronics devices from Apple Inc. for which it supplies the primary SoC, AppleInsider has learned.
People familiar with the new Apple handset say the Korean electronics giant will cater to Apple's request for a unified central processing unit, which will serve a dual role in handling the device's application processing and digital media capabilities.
Meanwhile, those same people say wireless IC maker Infineon has also captured a significant piece of iPhone, asserting itself as the primary supplier of both baseband and RF chips. Specific chip models were not identified.
For Samsung, the iPhone win represents its third for an Apple electronics device in less than six months. Last September, the firm was awarded key roles in the second-generation of both the iPod nano and iPod shuffle, for which it supplies the primary SoC and audio decoder chips, respectively.
Samsung's seat in iPhone is also symbolic of an evolving bond and level of trust with the traditionally tight-lipped Apple, whose concerns over intellectual property and commonality reportedly drove it to seek out the electronics maker as a sole supplier of ICs for its small-form-factor electronics devices.
Word of Samsung's part in iPhone arrives on the heels of several contradictory reports from industry executives, who in interviews over the past two weeks have pegged a variety of different firms as the primary chip supplier for the device.
It all started with a spokesperson from Apple Germany, who on the day after iPhone was unveiled told Reuters that the central processing unit was "from Intel." Within hours, Intel vehemently denied the claim, informing the media outlet it was "not providing the silicon inside the iPhone."
A subsequent identification came from an Intel Italy exec, who pinned the chip as an Xscale -- an integrated circuit developed by Intel and later sold to Marvell. For one reason or another, all appear to have missed the mark.
The identification of Samsung and Infineon in their respective iPhone roles also corroborates a "Likely iPhone supplier list" compiled by Morgan Stanley analyst Rebecca Runkle several weeks before the device's debut. In it, the analyst added that Marvell would supply the WiFi chip, Sharp the display, and Sony the battery.
In the December report, Runkle also singled out Samsung as the primary supplier of 4-gig and 8-gig NAND flash memory with support from secondary supplier Toshiba.
Of all analysts who weighed in on the iPhone ahead of its introduction, the Morgan Stanley analyst was nearly spot-on in all of her predictions, including its various technical specifications.
People familiar with the new Apple handset say the Korean electronics giant will cater to Apple's request for a unified central processing unit, which will serve a dual role in handling the device's application processing and digital media capabilities.
Meanwhile, those same people say wireless IC maker Infineon has also captured a significant piece of iPhone, asserting itself as the primary supplier of both baseband and RF chips. Specific chip models were not identified.
For Samsung, the iPhone win represents its third for an Apple electronics device in less than six months. Last September, the firm was awarded key roles in the second-generation of both the iPod nano and iPod shuffle, for which it supplies the primary SoC and audio decoder chips, respectively.
Samsung's seat in iPhone is also symbolic of an evolving bond and level of trust with the traditionally tight-lipped Apple, whose concerns over intellectual property and commonality reportedly drove it to seek out the electronics maker as a sole supplier of ICs for its small-form-factor electronics devices.
Word of Samsung's part in iPhone arrives on the heels of several contradictory reports from industry executives, who in interviews over the past two weeks have pegged a variety of different firms as the primary chip supplier for the device.
It all started with a spokesperson from Apple Germany, who on the day after iPhone was unveiled told Reuters that the central processing unit was "from Intel." Within hours, Intel vehemently denied the claim, informing the media outlet it was "not providing the silicon inside the iPhone."
A subsequent identification came from an Intel Italy exec, who pinned the chip as an Xscale -- an integrated circuit developed by Intel and later sold to Marvell. For one reason or another, all appear to have missed the mark.
The identification of Samsung and Infineon in their respective iPhone roles also corroborates a "Likely iPhone supplier list" compiled by Morgan Stanley analyst Rebecca Runkle several weeks before the device's debut. In it, the analyst added that Marvell would supply the WiFi chip, Sharp the display, and Sony the battery.
In the December report, Runkle also singled out Samsung as the primary supplier of 4-gig and 8-gig NAND flash memory with support from secondary supplier Toshiba.
Of all analysts who weighed in on the iPhone ahead of its introduction, the Morgan Stanley analyst was nearly spot-on in all of her predictions, including its various technical specifications.
Comments
Well.. so are all the hiss problem with those latest nano part of samsung problem with them not any good at making audio chips?
1.) What hiss problem?
2.) Hiss would not be caused by the SoC. The SoC runs the OS, which amongst other things converts compressed music to PCM. This PCM stream is then converted to analogue by a DAC. iPods use Wolfson DACs. Any hiss is likely to be a PCB layout issue.
Upon investigation, there's no currently commoditized Samsung CPU core that I think would be suitable for the iPhone. The best they seem to have is the aging S3C2440. They do, however, market this video co-processor which may have been bundled in an SoC/SiP with other parts. In this case, I don't see how the iPhone is anything less than a two-chip system.
Intel owned the StrongARM processor design on which the XScale is based before they sold it to Marvell. Samsung is a licensee of the ARM processor also.
Samsung has the technology to build SoC's with an ARM core and additional circuitry as required by their partner (Apple) so IMHO they were the favourite all along since they're also used in the iPods to provide one platform and reduce the chips in the iPod.
ARM chips in phones in particular are quite common and recently the Symbian phones have been running of single chip ARM platforms that integrate both the main CPU and the radio chips into one core to reduce power consumption. I'm quite surprised Apple have another chip in there for radio.
They're all kind of right IMHO.
Intel owned the StrongARM processor design on which the XScale is based before they sold it to Marvell. Samsung is a licensee of the ARM processor also.
Samsung has the technology to build SoC's with an ARM core and additional circuitry as required by their partner (Apple) so IMHO they were the favourite all along since they're also used in the iPods to provide one platform and reduce the chips in the iPod.
ARM chips in phones in particular are quite common and recently the Symbian phones have been running of single chip ARM platforms that integrate both the main CPU and the radio chips into one core to reduce power consumption. I'm quite surprised Apple have another chip in there for radio.
I was just going to post pretty much the same thing.
The question is what kind of ARM, and what kind of co-processors. Is it ARMv4 (XScale, ARM7, ARM9), ARMv5 (ARM11), or ARMv6 (Cortex A8 )? Does it have dedicated graphics hardware? Is the cellular baseband capable of being firmware upgraded to 3G standards like HSDPA or EVDO? What's the average dissipated power? These are the questions that can be answered once we identify the specific chip package(s).
Well, we knew it was going to be ARM-based before there was even a rumor about its development. Not only is the iPod ARM-based, but the whole industry of 32 bit embedded, low-power devices is ARM-based.
The question is what kind of ARM, and what kind of co-processors. Is it ARMv4 (XScale, ARM7, ARM9), ARMv5 (ARM11), or ARMv6 (Cortex A8 )? Does it have dedicated graphics hardware? Is the cellular baseband capable of being firmware upgraded to 3G standards like HSDPA or EVDO? What's the average dissipated power? These are the questions that can be answered once we identify the specific chip package(s).
I just read this article, kind of long but interesting, and at the end it talks about OSX on ARM. He mentions his next article will be this very subject. I look forward to finding out how OSX runs on the iPhone.
http://www.roughlydrafted.com/RD/RDM...5E44A3285.html
The company's direction in pushing OS X into new applications raises the question: how is it that the iPhone can run OS X, which demands gigabytes of RAM on the Mac? The next article will take a look.
Will Apple Legal likely pursue sources of product pre-release specifications? Doesn't this kind of information give potential competitors a leg up?
If it is a leak yes. Nobody actually has proven what is in the iPhone yet - just guessed.
I just read this article, kind of long but interesting, and at the end it talks about OSX on ARM. He mentions his next article will be this very subject. I look forward to finding out how OSX runs on the iPhone.
http://www.roughlydrafted.com/RD/RDM...5E44A3285.html
The company's direction in pushing OS X into new applications raises the question: how is it that the iPhone can run OS X, which demands gigabytes of RAM on the Mac? The next article will take a look.
Mac OS X has a hybrid microkernel called XNU. Windows has a bulky monolithic kernel, and one hell of a sloppy one at that. Linux has a monolithic kernel, but since day one it has been easy enough to compile different payloads of the linux kernel, so to speak. I'm pretty sure the Win XP kernel is in the 200MB range. I've never checked to see what the desktop XNU is, but I imagine it's between 1MB and 5MB. I think my Linux 2.6.11 kernel (or thereabouts) on my desktop PC is about 30MB. My Linux 2.6.11 kernel on my XScale boards has been pared down to about 600k.
In other words, OS X's architecture lends itself to being easily configured for other devices. Incidentally, it also lends itself really well to RTOS-like performance. Window's fails on both fronts. That's why Win CE exists.
So.. Marvell _and_ Samsung can be making the CPU.
this Marvell chip?
http://www.eetimes.com/showArticle.j...leID=196600907
Help me out ARM experts.
No speaka da acronym,,, but could the iPhone processor be
this Marvell chip?
http://www.eetimes.com/showArticle.j...leID=196600907
Help me out ARM experts.
Quite possibly. This is a positive step for XScale, which in the PXA270 and 290 didn't really have much to offer. But I would still have to say that I prefer the 65nm OMAP 34xx on paper.
Quite possibly. This is a positive step for XScale, which in the PXA270 and 290 didn't really have much to offer. But I would still have to say that I prefer the 65nm OMAP 34xx on paper.
Oh yeah!
Marvell is a fabless semiconductor designer and might employ Samsung for the manufacturing of the device. Samsung, Chartered, Infineon, Freescale and IBM have a joint foundy collaboration going on syncing their fabrication processses enabling customers to design once and manufacture wherever. Marvell have been collaborating with all these companies in the past.
So.. Marvell _and_ Samsung can be making the CPU.
...and Intel, since the Marvell ARM processor is the Intel Monohans processor, still being made by Intel for Marvell.
The point is, just because there is a problem with one hardware part, doesn't mean they can't fix it and/or create better hardware parts. Call me a huge fanboy that is nerdy to the max, but I stand by Samsung because Apple has the option to go with many many different electronics companies to manufacture its parts and it chooses Samsung for one reason or another.