Apple on chip engineer hiring spree [why I'm not thrilled]!
By Ronald O Carlson
From its purchase of PA Semiconductor to the hiring of over 100 people with expertise in chip design or manufacturing, it is becoming increasingly clear that Apple is serious about creating chips in-house. However, given the company’s past record, I’m not all that excited about the prospect of filling my drawer with even prematurely useless junk from Cupertino.
The Wall Street Journal reports that Apple has hired scores of people with previous chip related expertise. Notables include Mark Papermaster of IBM fame and a slew of people recently hired away from troubled chipmaker AMD, such as Raja Koduri and Bob Drebin. The Journal reports:
Apple’s hiring spree in semiconductors started well before the acquisition [of PA Semiconductor] and has continued through the past few months, according to postings on the networking site LinkedIn. The site contains more than 100 people listing current Apple job titles and past expertise in chips, including veterans of Intel, Samsung and Qualcomm.
On the one hand, yes, Apple does need chip expertise in order to create its groundbreaking products—widgets with features others don’t have or haven’t done nearly as well. From the company’s groundbreaking 17 inch notebook that offers 8 hours of useable battery (you mileage probably varies) to the iPod touch’s excellent gaming performance, Cupertino’s care and attention to detail really show through, and the company is to be commended for this.
On the other hand, there is wide (deluxe) debris trail that follows in Apple’s wake. For example, the company’s brilliant Apple Display Connector combined monitor, power and USB wiring into a single neat cable. However, the catch is that no one else in the industry picked up on the idea and Apple itself dropped ADC just a couple years after introducing it.
Another piece of junk for the drawer?
For tens if not hundreds of thousands of customers who owned an ADC equipped a PowerMac and Cinema Display, there’s a very strong likelihood that one or the other product (probably not both) died, leaving the user with little recourse but to buy new (ie not ADC) hardware.
This chain of events can be found repeatedly throughout Apple’s history with the G4 Cube, Newton, NuBus, LC slot, eMate, OpenDoc and ADB all coming to mind. The next obvious candidate for deluxe debris status is the company’s new iPod shuffle, which requires a custom manufactured headset.
Granted time and technology do move on, and some detritus is to be expected, but Cupertino seems to have more than its share of these failures. Further, Apple’s greatest successes of recent years can be tracked back to technologies borrowed, stolen or shared, like switching from PPC to Intel processors, ADB to USB and AppleTalk to Ethernet, NuBus to PCI, etc.
That said, things have on the whole been better since Steve came back. Will this time buck that trend or will chips designed in Cupertino pay off for the company, users and industry alike for years to come? I’m reservedly hopeful, but not hardly enthusiastic…
What’s your take?
Related:
Entry was posted
on Thursday, April 30th, 2009 at 1:41 pm
under 
Print This Post
Read more entries by Ronald O Carlson.
Stumble It!
April 30th, 2009
You have to distinguish between internally used silicon and the interconnect specifications. Connectors are required to share specifications or two devices won’t cooperate even if each works fine on its own.
Internal devices don’t have that burden. They only need to work well on their own. Apple has designed custom ASICs for use in their computers for many years. This is not a totally new step for them.
The potential for Apple if they execute well is very large. iPhones, iPod touch and maybe the new media gadget will all benefit from this by becoming smaller, thinner, lighter, less power hungry and easier to manufacture.
The downside is that they could get stuck with an outdated design. This is the reason they are hiring the pros, the best of the best.
April 30th, 2009
Hi Ron,
I’m nowhere near as concerned as you.
While Apple has some history of failures when they go it alone more and more of their success is now tied in to hardware that uses more and more custom chip design and manufacture.
The iPods and iPhones are almost totally reliant on custom chips. Even the processors used in them are incredibly short run designs by the standard of AMD and Intel.
We now have OS X running on these designs. When Apple bring out a device that will compete with the netbooks (and it won’t be a netbook as we now know it) you can be sure that it will have a large number of custom chips and probably run on an ARM processor. Remember that Apple also own a chunk of ARM an have done for many years. There are already Apple branded ARM chips in the current iPhone.
So chip design and manufacture is not a new area for Apple and is of growing importance. While the Macintosh currently relies on more standard chip designs our iPods and iPhones need the hardware design skills Apple is recruiting.
May 1st, 2009
Gosh! I agree! We should stop all innovation! Because some things may not work out!
May 1st, 2009
Ron,
your argument is nonsensical.
The disadvantages of incompatible external connectors like ADC, ADB, Nubus or the LC slot bears no relation to enhanced and customised CPUs and VLSI chips.
There is no incompatibility to worry about as PA Semi is an expert in customising the very same ARM processor architecture that Apple already uses in the iPhone/Touch. Likewise Apple has designed their own ASICs and other custom chips since the days of the Apple II.
-Mart
May 1st, 2009
Hmm, my comment hasn’t posted – I’ll try again:
Ron,
your argument is nonsensical.
The disadvantages of incompatible external connectors like ADC, ADB, Nubus or the LC slot bears no relation to enhanced and customised CPUs and VLSI chips.
There is no incompatibility to worry about as PA Semi is an expert in customising the very same ARM processor architecture that Apple already uses in the iPhone/Touch. Likewise Apple has designed their own ASICs and other custom chips since the days of the Apple II.
-Mart
May 1st, 2009
Although I agree with your criticism of ADC and the LC slot, I can’t resist commenting on your beef with some of Apple’s other proprietary connection technologies:
ADB – How would it have been better for Apple not to have developed the Apple Desktop Bus standard back in 1987 and benefitted from USB-like features such as daisy-chaining keyboards, mice, digitising tablets, printers, etc etc off the one cable connector bus, while the rest of the industry struggled along with parallel ports, serial mice and different PS2 connectors for keyboards etc for all those years? When USB finally came along, Apple themselves were the ones who in 1998 adopted USB across the board and dragged the rest of the industry kicking and screaming away from the mish mash of horrible old legacy ports.
NuBus – How would it have been better if Apple had not adopted the Texas Instruments NuBus standard way back in 1987 and benefitted from plug and play ease of use and 32-bit performance while the rest of the PC industry choked along with 8-bit and 16-bit ISA and then EISA buses with all their dip switch, IRQ line, I/O address and DMA channel hell and incompatibilities and configuration nightmares. Again, when PCI came along and finally matched the plug-and-play advantages of NuBus in the mid 90’s, Apple changed all of their systems over.
LocalTalk (AppleTalk was the networking protocol that ran on LocalTalk and also ran on Ethernet) – Apple also led the way with LocalTalk, very cheap networking technology that was built into every Mac at a time when complex expensive Token Ring cards and thick heavy coax cables and transceivers made life hell for PC users. Again, when Ethernet was released, Apple adopted the standard and built it into every Mac. Likewise the AppleTalk protocol itself was a revelation in automatic peer-to-peer discovery of other devices on a network. Imagine not needing to dedicate a PC to sharing a printer to every user on a LAN or being able to file-share between computers without needing a file server. Again, once TCP/IP gained popularity, Apple led the way adopting it while Microsoft dragged it’s feet.
-Mart