Apple nabs senior RF engineers from Broadcom amid rumors of new in-house baseband team
At least two high-level baseband hardware engineers have left wireless semiconductor company Broadcom for new positions at Apple this year, adding weight to speculation that the iPhone maker is planning to design and build its own baseband processors.
![](http://photos.appleinsidercdn.com/gallery/8901-302-Screen-Shot-2014-04-09-at-101952-PM-l.png)
Former Broadcom principal engineer Paul Chang joined Apple as a senior program manager in February after more than 11 years at the chipmaker, AppleInsider has learned. While at Broadcom, Chang was RF hardware lead for the development of baseband transceivers that made their way into mobile devices from both Nokia and Samsung.
Chang has extensive experience "working as the RF chip lead on complex mobile communications product development projects that generated multi-billion dollar revenue and involved hundreds of people from cross functional teams." He played a large role in spinning up Broadcom's mobile communications CMOS business unit, driving product development "from the inception to delivery to customers."
He holds a bachelor's degree in electrical engineering from Carnegie Mellon, a master's in the same field from UCLA, and an MBA from UCLA's Anderson School of Management, while his name appears on at least three Broadcom patents covering integrated circuit manufacturing methods. An update to Chang's LinkedIn profile confirmed the new position at Apple.
Another longtime Broadcom engineer, Xiping Wang, joined Apple in January after serving over 10 years as a design engineer and hardware development manager with Broadcom. Wang began his career as an RF engineer at Motorola after attending UC Davis.
![Broadcom WICED module](http://photos.appleinsidercdn.com/0106-broadcom-1.jpg)
All together, Apple has assembled at least 30 mid- and senior-level baseband software and hardware engineers from Broadcom and current iPhone baseband vendor Qualcomm over the past three years. Apple is also advertising more than 50 additional openings related to RF chip design, an indication that the build-up is not yet complete.
Discovery of the hires comes just two days after rumors began circulating that Apple was planning to move its baseband processor design in-house.
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.
Building baseband components is an exceptionally complex process that is notoriously difficult to perfect. Until Intel's introduction of its XMM 7160 last October, only one company -- Qualcomm, whose MDM9615M is inside Apple's iPhone and iPad -- had managed to create a baseband processor that integrated support for LTE, 3G, and EDGE networks inside a single package.
Unlike Apple's recent success with its A-series application processors, there has been no known large-scale acquisition to bolster a possible in-house baseband team. The company picked up semiconductor firms P.A. Semi and Intrinsity for a combined $399 million to support the A-series effort.
The move would, however, fit in with Apple's recent push to exert more control over the components that power its most important products. Apple inked a $578 million deal to open a dedicated sapphire factory in Arizona late last year, and has reportedly bid at least $479 million for controlling interest in Renesas SP Drivers, the exclusive provider of LCD drivers for the iPhone.
![](http://photos.appleinsidercdn.com/gallery/8901-302-Screen-Shot-2014-04-09-at-101952-PM-l.png)
Former Broadcom principal engineer Paul Chang joined Apple as a senior program manager in February after more than 11 years at the chipmaker, AppleInsider has learned. While at Broadcom, Chang was RF hardware lead for the development of baseband transceivers that made their way into mobile devices from both Nokia and Samsung.
Chang has extensive experience "working as the RF chip lead on complex mobile communications product development projects that generated multi-billion dollar revenue and involved hundreds of people from cross functional teams." He played a large role in spinning up Broadcom's mobile communications CMOS business unit, driving product development "from the inception to delivery to customers."
He holds a bachelor's degree in electrical engineering from Carnegie Mellon, a master's in the same field from UCLA, and an MBA from UCLA's Anderson School of Management, while his name appears on at least three Broadcom patents covering integrated circuit manufacturing methods. An update to Chang's LinkedIn profile confirmed the new position at Apple.
Another longtime Broadcom engineer, Xiping Wang, joined Apple in January after serving over 10 years as a design engineer and hardware development manager with Broadcom. Wang began his career as an RF engineer at Motorola after attending UC Davis.
![Broadcom WICED module](http://photos.appleinsidercdn.com/0106-broadcom-1.jpg)
All together, Apple has assembled at least 30 mid- and senior-level baseband software and hardware engineers from Broadcom and current iPhone baseband vendor Qualcomm over the past three years. Apple is also advertising more than 50 additional openings related to RF chip design, an indication that the build-up is not yet complete.
Discovery of the hires comes just two days after rumors began circulating that Apple was planning to move its baseband processor design in-house.
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.
Building baseband components is an exceptionally complex process that is notoriously difficult to perfect. Until Intel's introduction of its XMM 7160 last October, only one company -- Qualcomm, whose MDM9615M is inside Apple's iPhone and iPad -- had managed to create a baseband processor that integrated support for LTE, 3G, and EDGE networks inside a single package.
Unlike Apple's recent success with its A-series application processors, there has been no known large-scale acquisition to bolster a possible in-house baseband team. The company picked up semiconductor firms P.A. Semi and Intrinsity for a combined $399 million to support the A-series effort.
The move would, however, fit in with Apple's recent push to exert more control over the components that power its most important products. Apple inked a $578 million deal to open a dedicated sapphire factory in Arizona late last year, and has reportedly bid at least $479 million for controlling interest in Renesas SP Drivers, the exclusive provider of LCD drivers for the iPhone.
Comments
I'd assume there's some pretty big challenges with managing RF noise though.
In a few years time Apples competition will find it difficult to combat the latest iDevice because they are relying on mass produced parts from other companies.
Seems slow in the game!
If nothing else this could improve Apple's ability to negotiate with Qualcomm.
One thing that Apple will not be able to escape is paying patent royalties since a lot of this tech is still patented by Qualcomm and Broadcom, although some of the Rockstar patent portfolio will help them.
With Apple now getting into its own glassmaking, chip design and baseband modem design, the last holy grail is doing their own screen technology. Watch this space...
Yes, at the volumes Apple is selling iOS devices, small savings would quickly add up.
Apple's high-volume, high-end devices really do put them in a unique position that no one else can match. Samsung would be the next vendor in the list but since their high-volume, high-end devices certainly don't have the same profit-margin, price point, or volume they are a distant second to Apple. That said, how much of a lead does Apple really have over Samsung here since economics of scale does have a ceiling so that making 20 million in a year v 60 million in a year, for example, isn't going to change the cost for design as the savings get exponentially smaller. If one fabricator can do 20 million the same as 60 million then Apple could save there but there are still per unit or batch costs, and if Apple needs additional plants to to go beyond 20 million it may even cost more to get those up and running to meet their needs.
On screen technology, even Japanese companies are hard pressed to compete with Korean fabs. I don't see how any American company would have any resources to draw on.
Buy Sharp you say? Japanese would never go for it. What Apple has been doing is investing in Sharp's production development, and even that has seemed problematic. If we get IGZO phones or iPad minis this year, or 4K displays, we will know some more progress has been made.
The point is, if even Sharp has trouble doing something Apple-like in displays, how could Apple do anything with far fewer resources?
Apple's volume gives advantages in manufacturing costs and negotiating power. They should be able to contract custom SoCs at minimum cost and license/cross-license IP at minimal costs.
If Apple needs to pay premiums for additional manufacturing -- those are good problems to have.
Apple is certainly big enough to start to engineer their own proprietary phone technology that could potentially put them in the drivers seat and eliminate their need to license costly existing patents.
This would make a lot of sense. One of the advantages Qualcomm has with the SnapDragon is that the baseband modem is now being integrated on the SoC package for the CPU. It's a big deal since Apple still has to buy a separate chip from Qualcomm to do the same thing.
The fact that I saw ARM emblazoned on the broadcom baseband module rang this bell for me.
A7 M7.... R7... N7
As I hypothesized yesterday, there may be things that Apple can do at the very low levels of the IP stack and the baseband encoding that could be done better on, say (highly speculative), the GPUs on the A*.
and you may want to build special comm capabilities into the guts of the radio unique to the Apple space... I keep on thinking Pcell and/or mesh network technology (why not build 'routing' into the radio, and not let it surface it all the way up to the CPU.... heck it may even make tethering a lot less power hungry... 'got a packet in on wifi'... it's just passing through to the the LTE radio... don't even interrupt the CPU... and route it upstream.
And finally, make 'find my friends' a hardware thing... the radio can hear everything now, just that the signalling encrypts it so it's justs spectrum noise (yeah, I'm oversimplifying it). What if at a much lower power level across all radios apple has the ability to beacon to other Apple Radios without sending information to the cloud (how find my friend works now... again overly simplified), just start recognizing 'Hey... I see Pedro's phone is in the room (BT).... Frank's iPad is down the hall(WiFi), Jerry's is within a cell tower's distance(LTE).' done at the radio... not at the CPU. [caveats: I speak nothing about privacy here... just capability... I begrudgingly trust Apple to do the right thing, if they do anything]. even things like wearables would be enhanced by building apple specific radio signaling...
As an aside, fascinating how the 5S decimates the to-be-released S5 in so many tests- while being a 6 month old device with 1/2 the CPU speed, cores, and RAM. A testament to Apple's incredible in-house skill with the A7 chip, and bodes well for whats next. What's also fascinating is that I have literally not heard a peep online about this from the "LOL IPHONES ARE SO UNDERPOWERED" contingent, and now it's normal that a flagship Samsung phone gets smoked by a 6 month old iPhone.
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.
I'd assume there's some pretty big challenges with managing RF noise though.
I believe the RF is confined to the transceiver, which controls the sending and receiving of wireless signals through the antenna, the baseband chip acts as the "brain".
Don't poach us, we'll poach you.
Cute.
Quote:
Apple is also advertising more than 50 additional openings related to RF chip design, an indication that the build-up is not yet complete.
Although it's not specified if Apple actively pursued them, I would think that they probably came over following the advertising of openings related to RF chip design.
The difference of course is as great as hunting and poaching, one is legal, the other not.
In fact, it's figuratively night and day or rather day and night respectively.
You won't hear to much about the performance of the A series but you will hear over and over that, Apple shouldn't do anything (desktop level Arm cpu, Apple search), and on the other hand you will continue to hear that Apple doesn't innovate, but Apple has the resources to do it, it's just the question of does Apple have the will to do it.
If it does, you can bet Apple will replace the former with Appletooth. They’re not exactly a fan of things that are blu…![;)](http://forums-files.appleinsider.com/images/smilies/1wink.gif)
I believe the RF is confined to the transceiver, which controls the sending and receiving of wireless signals through the antenna, the baseband chip acts as the "brain".
Yes, that's true, but it also serves as a further filter between the transceiver and the CPU. The silicon for a baseband processor will have all manner of noise filtering, and will also be built using larger geometries than the main CPU which will further help in dealing with RF noise.
For that reason, I don't know if you'd be able to embed the baseband and the CPU onto the same silicon, or even different silicon in the same package, but I may be wrong.
I never said that Apple should buy Sharp or manufacture their own screens. Apple's strategy is not to own factories, pioneered by Tim Cook in the late 90's. What I was implying was that Apple begin developing their own screen technology and not rely other partners for this. Who knows, they may already be doing this according to their patent filings. But I would think that Apple developing their own screen tech and then having Sharp or whomever manufacture it may be appealing to Apple since only Samsung has the capability to match Apple for engineering....CPU design, memory, and screen tech are all things Samsung does anyway.
The other thing I could see Apple doing is camera tech and not relying on Sony for it.