Going for Broke [Will Apple and Intel merge?]

PBS ^ | June 9, 2005 | By Robert X. Cringely

[WOSG]

Sorry, pure guess is wrong.

IBM has a large chip business, and with Cell processor is putting more behind it. They sold PC hardware because it wasnt making good enough margins, and they are now in the business of being a 'technology' supplier to companies liek DELL.

Moreover, Freescale wants to sell chips and would be happy to supply 100% of Apple's needs if IBM didnt want to. (Currently Apple takes Freescale's G4 and IBMs G5). It was an Apple decision, not an IBM/Freescale decision.

The way I see it, it only makes sense in this perspective: If Apple made this decision 20 years ago, we'd have gotten a viable competitor to Microsoft.

Apple must be thinking that their hardware margins are effectively 0%, and they can compete on their software basis. If that is the case, you should be able to but an "OS X" software package and put it on your own PC box.

That *or* they feel the desktop is becoming less and less relevent.

IMHO, this decision is not right - they are missing the boat on the OS wars, and they are also missing the key value differentiator they have in a closed box solution. 20 years ago, this would have been the right decision. Now, when by their own admission the CPU / OS layer is more flexible *and* you have "OS X"-like OS available on x86 ... LINUX ... this is throwing away some real Mac advantages and putting them in direct competition with more low-cost alternatives.

Mac OS X on Intel sounds a lot like Solaris/x86, which ended up going nowhere.

[WOSG]

"Because IBM couldn't push the PowerPC in the ways Apple wanted (no 3Ghz G5 or PowerBook G5 yet);"

That sounds right, although the Intel chips tend to run hotter and were running into their own performance wall. Also, low-power G6 was on the way.

"because Apple apparently thinks Intel is a more reliable long-term partner than AMD and IBM;"

Apple's other partner has been Freescale ...

" and because IBM was probably getting less sensitive to Apple's demands,"

Apple will find out that they are a smaller %age revenue demand on Intel than they were to IBM's and Freescale's chip business.

The reasons, while valid, seem to be mis-timed, in that they were *more* valid 5 to 10 years ago than now.

Intel's about to go dual core because their single core has run out of steam. Meanwhile we had dual core Apples on tap.

[WOSG]

Thanks for your insight. ... the most credible reasons I can see are:
" a direct challenge to Microsoft, and the fact IBM was dilatory in getting a laptop usable version of the G5. "

[Proud_texan]

Thanks for the update, haven't paid much attention to them in the last couple of years after they closed half their chip plants and yet still managed to lose $389 million in that operation for the year as well as sell off their HDD division. What were their profits from chips last year?

I would find the task of supplying drivers for the diverse number of add-ons for the PC to be daunting, if they go that direction it will be interesting to see how they survive the "your OS doesn't work with my off-brand/legacy video card" crowd.

[frgoff]

Simple. IBM is focusing on the game console market, Motorola is focusing on the embedded market. Intel is focusing on the PC market. Apple makes PCs.

[frgoff]

Why? Would you just chuck a $6 BILLION a year business? And don't give me this software only model. Dell makes more money than Microsoft.

[krb]

Not a chance. it was really weak when intel got DEC semiconductor division (alpha).

Compaq got that part of DEC.

Intel got the DEC StrongARM (now Intel XScale).

[ProudVet77]

Actually the guys that went to Compaq are now working back in Hudson for Intel. When Compaq got bought out by HP they sold off the Shrewsbury semicondutor that had been bought by Compaq back to Intel. Have a friend whose resume reads like this:
Digital Equipment
Compaq
HP
Intel
and he actually hasn't left the same job he's had all along.

[ThinkDifferent]

That sounds right, although the Intel chips tend to run hotter and were running into their own performance wall. Also, low-power G6 was on the way.

What low-power G6? Apple's main problem was that neither IBM nor Freescale could produce a comeptitive laptop CPU. The G5 is too hot and the G4 doesn't compete well with the Pentium-M.

Apple will find out that they are a smaller %age revenue demand on Intel than they were to IBM's and Freescale's chip business.

Yeah, but that doesn't matter. They don't need Intel to do anything they don't normally do, unlike IBM and Freescale. Apple is pretty much the only customer for the G5, and Freescale sells CPUs mostly for embedded use where low power is more important that high MHz or a decent bus.

[ThinkDifferent]

My theory is that Apple will be getting out of the hardware business altogether. Which is something they should have done a long time ago.

No, and no. Directly competing against Microsoft on commodity hardware is not a winning strategy; see OS/2, pre-Apple NeXT, and Be.

[The_Reader_David]

Thanks for the confirmation.

I'm now even more certain of my analysis.

[jpsb]

Good questions (great questions really) and his answer makes more sense then anything else I've read.

[Leonard210]

...and the fact IBM was dilatory in getting a laptop usable version of the G5.

Good analysis The_Reader_David. The fact that they've been running OS X on Intel chips in Cupertino for 5 years indicates that this has always been a priority and probably more than a "just-in-case" scenario. So, what if SJ heard from IBM that a realistic timetable for a G5 that would be laptop capable was 5 years? If I have a working alternative in a back room and my supplier tells me they essentially can not keep up, guess what I'm gonna do.

[WOSG]

"G4 doesn't compete well with the Pentium-M."

What are the spec-marks?

[WOSG]

See this note ...

http://www.themacobserver.com/article/2005/01/12.14.shtml

"Heck, I'll bet a lot of people would snatch up a lot of single-core G4 Powerbooks at 1.8GHz to 2.0GHz. That'd be a 20-33% increase in speed, which isn't too shabby.

I wouldn't be surprised if we never do see a G5 Powerbook. Apple may stick with souped-up G4s from Freescale until the "G6" is ready. And that might not be such a bad thing. The G4 still has legs and is a rather power efficient processor. " - posted early 2005..


and this posted in 2002 ...
" Anonymous wrote:
I have to wonder what life would be like if Apple used x86 chips.


Apple would be dead, just another software company trampled by the MS/Dell rhino and left for dead by the side of the road."

Now we get to find out.

[Ernest_at_the_Beach]

That *or* they feel the desktop is becoming less and less relevent.

I think the market is changing again....with all of the new game stations coming.....where does the enthusiasts requirement for hot new chips for their game machines go?

plus Intel is very interested in getting itself into the wiring of the home,.,,which is the same market Sony is going after, ...with the Cell processor....MS has the Multimedia stuff, going after it also....but Sony owns the HD TV business,,,so they may be the kingmaker.,...

[Leonard210]

Apple must be thinking that their hardware margins are effectively 0%...

I don't know where you'd get that idea. There is no indication that Apple is hurting in hardware or software sales. Quite the opposite. 400 + Apple design awards, 500,000 Apple developers selling over $500,000,000 in product through Apple retail and online 109 Apple retail stores serving 1,000,000 visitors per week, 2,000,000 copies of Tiger shipped in 6 weeks, fiscal 2005 second quarter ended March 26, 2005 posted a net profit of $290,000,000, shipped 1,070,000 Macintosh units, shipped 5,311,000 iPod units, for a 70 percent revenue growth and a 530 percent increase in net income.

There is also no indication that Apple on Intel means Apple on eMachines. If they took that approach, you are correct, they will have kissed their hardware sales goodbye as well as their 109 retail outlets. Again, however, there is no indication that they are about to do that.

[ikka]

Mac OS X on Intel sounds a lot like Solaris/x86, which ended up going nowhere.

Actually I am dumping most/all of my Linux servers and going with Solaris/x86 and a smattering of older SPARC machines also running Solaris.