or Connect
AppleInsider › Forums › Mobile › iPhone › Custom iPhone baseband chips seen as too 'herculean' a task for Apple to tackle in-house
New Posts  All Forums:Forum Nav:

Custom iPhone baseband chips seen as too 'herculean' a task for Apple to tackle in-house

post #1 of 72
Thread Starter 
With recent rumors and hires suggesting Apple may be planning to build its own custom baseband chips for future iPhones, moving away from off-the-shelf silicon from companies like Qualcomm, one analyst thinks it's a highly unlikely route for Apple to take given the time it would have to invest.


iPhone 5s logic board with Qualcomm baseband chipset. | Source: iFixit


Analyst Brian Modoff with Deutsche Bank issued a note to investors on Monday in which he said the issue in developing a multimode baseband from scratch is not dollars, but rather the the time necessary. He estimates it would take at least five years for top-tier talent and about 1,000 total engineers to develop a custom baseband, and he doesn't see Apple taking on that "herculean task."
Brian Modoff with Deutsche Bank believes Apple may instead be looking to better integrate existing baseband chipsets with its own custom A-series chips.
"Instead of an organic baseband development, we believe Apple is rather trying to realize better integration with existing baseband chipsets and their apps processor, or possibly developing their own internal Wi-Fi chipset," Modoff said.

The comments are similar to ones made last week by Rod Hall of JP Morgan, who cited the "notoriously difficult" nature of developing baseband processors. Hall said he didn't believe Apple would be able to utilize custom iPhone baseband chips until 2015 at the earliest --?a timeframe much sooner than Modoff believes is possible.

AppleInsider discovered earlier this month that Apple has hired away senior radio frequency engineers from Broadcom and Qualcomm. The news came after a rumor claimed Apple was planning to move its baseband processor design in-house.

In addition to hiring at least 30 mid- and senior-level baseband software and hardware engineers from existing players like Broadcom and Qualcomm over the last three years, Apple is also advertising more than 50 additional openings related to RF chip design. Still, those numbers are well short of the 1,000-plus engineers that Modoff believes would be necessary to accelerate baseband chip development.

The baseband chip acts as the "brain" of a handset's wireless modem. Working in concert with a transceiver, which controls the sending and receiving of wireless signals through the antenna, the baseband processor is what allows a device to communicate over cellular networks.

Over the last few years, Apple has been selectively bringing certain aspects of product development in-house, making the company less reliant on off-the-shelf parts that competitors can also utilize. Developing its own chips, such as the A-series processors found in the iPhone and iPad, also allows Apple the ability to better keep secrets within the company, without having to share information with third parties.
post #2 of 72
Would it more difficult to develop a CPU? Stoopid analysts.
post #3 of 72
Quote:
Originally Posted by mubaili View Post

Would it more difficult to develop a CPU? Stoopid analysts.

Its not about CPU or tech.

Its about patents. Talent migration. Need-to-innovate-our-own-thing.

post #4 of 72
So Apple will have some of the best minds working on baseband chips with a pretty much unlimited budget and they will fail how?
post #5 of 72
Quote:
Originally Posted by Chandra69 View Post

Its not about CPU or tech.
Its about patents. Talent migration. Need-to-innovate-our-own-thing.

Or license.

"The real haunted empire?  It's the New York Times." ~SockRolid

"There is no rule that says the best phones must have the largest screen." ~RoundaboutNow

Reply

"The real haunted empire?  It's the New York Times." ~SockRolid

"There is no rule that says the best phones must have the largest screen." ~RoundaboutNow

Reply
post #6 of 72

The only thing I am confident about is that analysts have no idea what Apple is doing.

post #7 of 72
undefined
post #8 of 72
Quote:
Originally Posted by mubaili View Post

Would it more difficult to develop a CPU? Stoopid analysts.

Whether it's more difficult than a CPU doesn't matter - Apple already have the talent for a CPU, but they don't have a lot of people on the baseband side of things, and they might not have the patents or licensing needed either. But as it turns out, a baseband chip probably would be more difficult than a CPU. The baseband chip often has its own memory and firmware, and the firmware has to respond to hardware / network events much quicker than a normal OS. If you've got the people it's possible, but it'll still take a long time and it's a specialist area.

 

The analysts are arguing that Apple would need both a lot of time and a bigger team to design their own baseband chip - this seems fair enough to me. The alternatives they've suggested also seem pretty sensible - both would reduce Apple's reliance on third parties, and will give them a good foothold should they want to manufacture the entire baseband chip.

post #9 of 72
Why have been seeing so many posts recently that just write undefined as the message?

"The real haunted empire?  It's the New York Times." ~SockRolid

"There is no rule that says the best phones must have the largest screen." ~RoundaboutNow

Reply

"The real haunted empire?  It's the New York Times." ~SockRolid

"There is no rule that says the best phones must have the largest screen." ~RoundaboutNow

Reply
post #10 of 72

What is the status on the Qualcomm RF360? If that chip is finalized for the iPhone 6 it would be very beneficial to Apple. One chip that can handle pretty much all bands on GSM, CDMA, and LTE. It would work on pretty much every carrier in the world. I wonder how far they are from production.

 

http://www.qualcomm.com/chipsets/gobi/rf-solutions/qualcomm-rf360-front-end

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply
post #11 of 72
This is interesting, in some ways this makes Apple become more like what Motorola Solutions does... and more of a telecommunications company, but not just. MOT was never good at technology transitions, from analog to digital and from feature phones to smart phones, but Apple is a leader. Time will tell if it can stay a leader...
post #12 of 72
Quote:
Originally Posted by SolipsismX View Post

Why have been seeing so many posts recently that just write undefined as the message?

 

If you happen to use Facebook for forum log in on other sites for some reason it will log you off of your account here and log you in with your Facebook account if you ever created one. The post are then posted as "undefined" Not sure if it is a glitch with safari, Facebook, or AI. But for some reason it always logs me off this account and tries to sign me in on Facebook if I visit a site where I actually use FB to log in for comments and come back. 

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply
post #13 of 72
Quote:
Originally Posted by ceek74 View Post

So Apple will have some of the best minds working on baseband chips with a pretty much unlimited budget and they will fail how?

Who said anything about an unlimited budget?

post #14 of 72
I wonder if Apple is planning on using custom technology to supplement existing baseband tech (like Qualcom's.

In one of his presentations of the pCell technology, Steve Perlman said they had installed the pCell chips into (on the back of??) iPhones in addition to the standard LTE chips. That way when a user entered into a metropolis supported by pCells, they would get the advantages of pCell. Conversely, when they left the pCell area, the cell communication would be handed off to the regular [non-pCell] cell network.
"Swift generally gets you to the right way much quicker." - auxio -
"He who laughs, lasts!" - Mary Pettibone Poole -
Reply
"Swift generally gets you to the right way much quicker." - auxio -
"He who laughs, lasts!" - Mary Pettibone Poole -
Reply
post #15 of 72

 these "...guys are not going to just figure this out. They’re not going to just walk in."

post #16 of 72
"Apple could never build a mobile 64-bit chip on their own." - some idiot analyst last year
post #17 of 72
Quote:
Originally Posted by macinthe408 View Post

"Apple could never build a mobile 64-bit chip on their own." - some idiot analyst last year

 

And piling on...

 

Remember all the analysts predicting the M7 chip?   Yeah, they're clued into what Apple is or isn't doing.

post #18 of 72
Quote:
Originally Posted by jsewell View Post

 these "...guys are not going to just figure this out. They’re not going to just walk in."

Yeah 1biggrin.gif
Quote:
"We've learned and struggled for a few years here figuring out how to make a decent phone. PC guys are not going to just figure this out. They're not going to just walk in."
Palm CEO Ed Colligan, commenting on then-rumored Apple iPhone, 16 Nov 2006
"Swift generally gets you to the right way much quicker." - auxio -
"He who laughs, lasts!" - Mary Pettibone Poole -
Reply
"Swift generally gets you to the right way much quicker." - auxio -
"He who laughs, lasts!" - Mary Pettibone Poole -
Reply
post #19 of 72
...they're not just going to walk in here....oh wait.....
post #20 of 72

Did a little more research on the Qualcomm RF360 and it appears to be ready for mass production now. I will be very surprised if this is not what Apple uses for the iPhone 6. This would be the holy grail.

 

From Qualcomm.

The WTR1625L, a product of Qualcomm Technologies, Inc., is the first in the industry to support carrier aggregation with a significant expansion in the number of active RF bands. The WTR1625L can accommodate all cellular modes and 2G, 3G and 4G/LTE frequency bands and band combinations that are either deployed or in commercial planning globally. Additionally, it has an integrated, high-performance GPS core that also supports GLONASS and Beidou systems. The WTR1625L is tightly integrated in a wafer scale package and optimized for efficiency, offering 20 percent power savings compared to previous generations. The new transceiver, along with the Qualcomm RF360 Front End chips, is integral to Qualcomm’s single-SKU World Mode LTE solution for mobile devices.

 

http://gigaom.com/2014/02/06/the-first-smartphone-powered-end-to-end-by-qualcomm-will-debut-this-year/

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply
post #21 of 72
If you understand the details of Apple's license with ARM, you would know that they use a lot of Arm design work to make the task of designing a better chip easier. They also have a higher level license for making their own chips that follow the ARM standards. It is certainly possible that Apple has the deep pockets to pay for a separate type of license from Qualcom or Broadcom. Apple has used Qualcom for some time now. Perhaps they will license many of the frand patents from Qualcom and Broadcom to make their own baseband chips in house. Apple has used the current system to it's own advantage for some time because they have the money to do what ever they desire to do strategically. I can imagine that Qualcom would not want Apple's A9 or A10 to greatly outdistance their own in house Cpu's with integrated basebands. Their ability to compete with Apple as providers for Android phone makers is not something they would like to give up.

The eventual advantage Apple could leverage here is a software system that controls what the baseband is doing based upon the task the phone is being used for. It is also possible to add bandwidth from a next generation wireless point to the phone functions to avoid call dropping and other annoyances on calls made from one iPhone to another (like iMessage). This would directly attack the wireless providers source of revenue and ability to upcharge their customers. Even having the ability to do this in house would be a deal changer as far as negotiations with Cellular companies goes.
post #22 of 72

Since the baseband chip works alongside the processor, integrating it into one package will lead to improvements in power efficiency.

post #23 of 72
Brian Modoff, what if Apple started this project several years ago? Your premise that they haven't started yet is shortsighted.
post #24 of 72
Quote:
Originally Posted by gwmac View Post
 

Did a little more research on the Qualcomm RF360 and it appears to be ready for mass production now. I will be very surprised if this is not what Apple uses for the iPhone 6. This would be the holy grail.

 

From Qualcomm.

The WTR1625L, a product of Qualcomm Technologies, Inc., is the first in the industry to support carrier aggregation with a significant expansion in the number of active RF bands. The WTR1625L can accommodate all cellular modes and 2G, 3G and 4G/LTE frequency bands and band combinations that are either deployed or in commercial planning globally. Additionally, it has an integrated, high-performance GPS core that also supports GLONASS and Beidou systems. The WTR1625L is tightly integrated in a wafer scale package and optimized for efficiency, offering 20 percent power savings compared to previous generations. The new transceiver, along with the Qualcomm RF360 Front End chips, is integral to Qualcomm’s single-SKU World Mode LTE solution for mobile devices.

 

http://gigaom.com/2014/02/06/the-first-smartphone-powered-end-to-end-by-qualcomm-will-debut-this-year/


Indeed the RF360 is ready for mass production, both Qualcomm's upcoming Snapdragon 808 and 810 support the chip.

post #25 of 72

"Analyst Brian Modoff"

 

EOM.

Proud AAPL stock owner.

 

GOA

Reply

Proud AAPL stock owner.

 

GOA

Reply
post #26 of 72
Quote:
Originally Posted by Chipsy View Post
 


Indeed the RF360 is ready for mass production, both Qualcomm's upcoming Snapdragon 808 and 810 support the chip.

Any reason why the RF360 could not be made to work with the A8 in an iPhone 6? Found this about the iPhone 5

 

"Right next to the Qualcomm MDM9615M is the Qualcomm RTR8600 multi-band/mode RF transceiver. The RTR8600 is paired alongside the MDM9615 to support various bands, including 5 UMTS bands, and over 5 LTE and 4 EDGE bands."

 

What would be so hard or different about getting RF360 support in an iPhone versus all the other Qualcomm baseband chips Apple has been using in past models? 

 

http://www.chipworks.com/en/technical-competitive-analysis/resources/blog/inside-the-iphone-5s/


Edited by gwmac - 4/14/14 at 1:10pm

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply
post #27 of 72
How about a non herculean task of purchasing Qualcomm and instantly owning the space? Better than letting that cash sit around eroding to inflation after the meager investment returns it gets.
post #28 of 72
Quote:
Originally Posted by mvigod View Post

How about a non herculean task of purchasing Qualcomm and instantly owning the space? Better than letting that cash sit around eroding to inflation after the meager investment returns it gets.

Not in a million years will that ever happen. Just the fact that Qualcomm is used by so many companies for products alone is enough to keep Apple away. But Broadcom might not be a bad fit for Apple. 

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply
post #29 of 72
Who says Apple isn't working on this for a few years. Fact is that we nor analyst gave it a thought before. That doesn't say Apple couldn't be in line or ahead of the described development path. Commonly the public and investors are signaling trends at the moment, the obvious is 'bout to happen.
post #30 of 72
Anyone doubting the difficulty of doing baseband (for technical and legal reasons) need only look to Intel to get a head-check. Even after acquiring Infineon's baseband group, Intel's baseband business is still 1/10th the size of Qualcomm's. The ratio is far worse if you narrow the focus to LTE.

http://www.fiercewireless.com/story/intel-mediatek-broadcom-and-nvidia-try-catch-qualcomm-lte/2014-01-10
post #31 of 72

One possibility which occurred to me that I haven't seen anyone mention is whether this is a reactive move on Apple's part due to Qualcomm behavior. I think the shift to 64 bit processing caught Qualcomm seriously off guard; they recently came out and said they're going with an ARM A57 design for their next (64-bit) mobile CPU instead of their previously custom Krait cores. IIRC, a custom-core 64-bit mobile CPU is coming _after_ that, but not until later 2015, which is like, forever away (about the time of Apple's A10, I'd say)

 

Seeing that Qualcomm has more non-Apple sales than Apple sales, I wouldn't be totally surprised if Qualcomm decided to try to make things a little difficult for Apple in integrating their broadband chips into iPhones. Not sure how, but I'll bet they could do it, just to even out the playing field a bit to please their other non-Apple customers, and save face a bit after the 64-bit shocker. And maybe Tim Cook's rational response was "fine- you know, we can bring this in-house too, if that's really what you want us to do".

 

Whether that would actually happen, and what's involved in that, I dunno. But Cook is a supply-chain guy, and if he sees suppliers (coughSamsungcough) messing with Apple, he's already shown he's ready and willing to bring that business elsewhere.

post #32 of 72

Edit above: "baseband" not "broadband"...

post #33 of 72
Quote:
Originally Posted by mvigod View Post

How about a non herculean task of purchasing Qualcomm and instantly owning the space? Better than letting that cash sit around eroding to inflation after the meager investment returns it gets.

 

The purchase price of Qualcomm would be somewhere above about $150 Billion.  While Apple has a ton of cash, that cash is spread around the world, and repatriating it for a purchase would result in a reduction due to taxes.

 

Further, if Apple were to purchase Qualcomm, they'd need to decide what they wanted to do with it.  Either it makes money (meaning they continue to do business with Apple's competition), or they only serve Apple, in which case $150 Billion is a hell of a lot of money (plus annual costs) to buy the chips they'd use.

post #34 of 72

I fully expect Apple to release an LTE/LTE Advanced ONLY device in the near future.  You heard it here first.  

 

Watch the carriers squirm.  ;)

post #35 of 72
Quote:
Originally Posted by pdq2 View Post
 

Edit above: "baseband" not "broadband"...

Look at your post above. See the little pencil icon next to the red flag, that is the edit button. 

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply

 

Never argue with an idiot. They will only bring you down to their level and beat you with experience. 

Reply
post #36 of 72
Quote:
Originally Posted by gwmac View Post
 

Look at your post above. See the little pencil icon next to the red flag, that is the edit button. 

 

Thanks- never noticed that before ; )

post #37 of 72
Quote:
Originally Posted by mubaili View Post

Would it more difficult to develop a CPU? Stoopid analysts.

 

 

Quote:
Originally Posted by Chandra69 View Post
 

Its not about CPU or tech.

Its about patents. Talent migration. Need-to-innovate-our-own-thing.

 

 

Quote:
Originally Posted by Phone-UI-Guy View Post
 

The only thing I am confident about is that analysts have no idea what Apple is doing.

 

As it was pointed out, the same thing could have be said about a CPU especially a multi-core CPU, it loaded with technologies and patent landmines. If we look back in time, people were probably saying apple would never make their own process with a long list of reasons why.

 

It completely obvious, it has all the so call experts scratching their heads wondering what this all could mean. With a $170B in the bank do not second guess what Apple is up to, they are the only company who has the $ and talent to pull things off that others would deem impossible.

 

post #38 of 72
Just like in the case of Nvidia and Icera, I'm sure Apple will end up (or already has?) buying a fabless semiconductor firm with the necessary experience and IP and supplement it with these other hires.
At the time when they purchased PA semi, no one expected them to have the best performing, custom core ARM chip on the market in only a few years. I'd imagine the baseband challenge will follow in the same line....
post #39 of 72

Apple couldn't do this from scratch and certainly not quickly. They'd need to find the PA Semi of the baseband world to buy. I don't really know who that is. Qualcomm are too big to be bought, most of the others are too incompetent.

post #40 of 72
Quote:
Originally Posted by libertyforall View Post
 

I fully expect Apple to release an LTE/LTE Advanced ONLY device in the near future.  You heard it here first.  

 

Watch the carriers squirm.  ;)

 

GSM will probably be phased out after LTE. There's already talk of carriers alrady switching off their 3G networks but none would even dream of getting rid of GSM. GSM is too reliable and it's dirt cheap to deploy. It's a great fail safe.  

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: iPhone
  • Custom iPhone baseband chips seen as too 'herculean' a task for Apple to tackle in-house
AppleInsider › Forums › Mobile › iPhone › Custom iPhone baseband chips seen as too 'herculean' a task for Apple to tackle in-house