Really switching to Intel!!??

[iJed]

I've never seen so many comments that show a lack of understanding as on this thread.

1) Porting Windows apps to Mac OS X on x86 will have little difference than porting Windows apps to Mac OS X on PPC. The only exception is code that contains assembler (which is maybe 0.00001% of all code).

2) Porting Mac OS X apps from PPC will be relatively straightforward. I'm not sure if porting is even the correct term to use here. Recompilation probably better describes the process.

3) Carbon apps do not pose any greater problem to recompile than Cocoa apps.

4) Many apps will simply recompile without any code changes whatsoever.

5) Porting Mac OS X to x86 does not imply that it will run on any generic x86 hardware. Although, with the Darwin source available and x86 native, it would probably not be that difficult to get an Apple-x86 version of Mac OS X running on generic hardware.

6) x86 would bring massive performance improvements across the Mac range.

7) x86 Macs would not be different in any noticeable way to users.

8) SSE2/3 would not be able to emulate Altivec/VMX. These are very different technologies. It may however be possible to translate some of these VMX instructions to SSE2/3 instructions. Much of the job will end up on the main CPU core though. I doubt Altivec emulation is worthwhile but we'll see if it happens.

9) Emulation of PPC is most certainly possible in the same way that PPC emulated 68K back in the day. Fortunately without the horrible mixed mode manager and performance penalties from emulating parts of the OS.

10) The Mac you just bought has not become obsolete overnight. 68040 Macs were supported for a long time after the PPC transition. Hopefully this time PPC Mac OS X will also be able to emulate x86 software. This may not be cost effective for Apple however.

 

[SiliconAddict]

i can feel it in my gut......

That is what is known as denial. Wait for the cognitive dissonance to kick in and you will be fine.

 

[VanNess]

The cnet story is pure baloney. Here's why:

Substantively, the cnet story is nothing more than a retelling of the month-old WSJ article which simply offered "two industry executives with knowledge of recent discussions." to jump to the conclusion that Apple was going to switch to Intel. The cnet story manages to be even more vague, citing only "sources familiar with the situation." In fact, their version of this story is so vague that, even with all that writing talent at cnet's disposal, it can only muster a two-sentence paragraph to report the "story":

Apple has used IBM's PowerPC processors since 1994, but will begin a phased transition to Intel's chips, sources familiar with the situation said. Apple plans to move lower-end computers such as the Mac Mini to Intel chips in mid-2006 and higher-end models such as the Power Mac in mid-2007, sources said.​

A vague, detail-challenged, two-sentence paragraph that nonetheless has spawned 700+ posts and counting in this thread (and elsewhere on the net) most of which seem to have bitten - hook line and sinker - the conclusion that Apple is going to use Intel CPU's in future Macs, replacing IBM. All of this for the want of 2 sentences that can't even shed a ray of light regarding say, the family of Intel processors Apple will allegedly use (x86 or something else). It's pretty flimsy stuff.

And one massively significant detail has been forgotten: Apple has just released Tiger. Right now, Apple needs to get the word out about Tiger in shinning, glowing irresistible terms without any distractions or side shows. They need the kind of buzz that articles like this generate, not endless speculation into 2006 and beyond regarding the "new Apple hardware platform" and what it may or may not mean. Tiger, Spotlight, Dashboard, Quicktime H.264 HD, iLife and so one will become orphans in the time leading up to the alleged switch to the yet unknown, invisible Intel processor.

Apple spent millions of dollars developing OS X Tiger, and they certainly would like to see some return on that investment. Announcing a processor switch on Monday puts that investment, and the important technologies that were developed for it (Spotlight, etc,), at serious risk by announcing - out of the blue - a future hardware realignment. It makes no sense whatsoever, and Apple is a smart company. I just don't see this happening. Particularly at this crucial time for OS X Tiger's successful rollout and Apple's interest in expanding OS X mindshare and marketshare.

700 posts and counting...and not a single one about Apple's shinny new OS on the eve of the WWDC. What's wrong with this picture?

 

[pudding]

Freescale Shares

Or maybe the B shares are for Intel investment? I mean, the sticker is only one week old. I find it strange to see new shares of Freescale at the same time this (old) rumor surfaces.

From Freescale website (in investor section):
FSL.B: Class B shares distributed to Motorola shareholders on December 2, 2004

 

[alandail]

And if developers know that they can create just one version of their software - Windows - and it'll run in a window on the Mac OS, where's the incentive for them to provide Mac versions. Really, this is a dangerous, multi-edged sword that's being played with here.

They will have the same incentive that classic developers had to move to OS X native stuff. People running OS X will want to spend their money on OS X native versions. There are also the millions of Mac users with PPC hardware that aren't going to just go away.

If Apple can significantly expand their OS market share, more innovative companies will leverage Cocoa to rapidly develop apps OS X only that reach users unaddressed by their windows only competitors.

It'll take a bold move executed right for Apple to leverage their OS advantage into a significantly larger market share.

 

[J-Ray1000]

Or maybe the B shares are for Intel investment? I mean, the sticker is only one week old. I find it strange to see new shares of Freescale at the same time this (old) rumor surfaces.

Eh, if Intel were to build PPCs, they wouldn't invest in a competitor, unless they were trying a hostile take-over. And if that were the case, Freescale wouldn't open up shares to get murdered with.

No, stock splits, class B, all of that stuff happens regularly with growing companies in diverse industries. The news of Mac on Intel is just a coincidence as Freescale's business is now mostly in imbedded chips and gaming, etc.

 

[J-Ray1000]

They will have the same incentive that classic developers had to move to OS X native stuff. People running OS X will want to spend their money on OS X native versions. There are also the millions of Mac users with PPC hardware that aren't going to just go away.

If Apple can significantly expand their OS market share, more innovative companies will leverage Cocoa to rapidly develop apps OS X only that reach users unaddressed by their windows only competitors.

It'll take a bold move executed right for Apple to leverage their OS advantage into a significantly larger market share.

Speak on, brotha.

 

[Stewdy]

Here's my take...

I think that Steve just wants to be made an honorary member of Blue Man Group. Apple using Centrino in the laptops would no doubt make that happen. I think that's why he always wears black. The apples on the powerbooks and ibooks will glow blue instead of white after the switch. Mark my words.

 

[cybermat]

I thought the G5 FAB Lab is a joint venture between IBM and Apple. Apple pumped a lot of money in setting up the lab. If Apple is going to switch to Intel in couple of years, all the money that Apple pumped is just sunk cost. Not much ROA if you ask me. Then again, maybe the opportunity cost is much higher with Intel.

 

[J-Ray1000]

700 posts and counting...and not a single one about Apple's shinny new OS on the eve of the WWDC. What's wrong with this picture?

Hey, got you posting, didn't it? Besides, what else do you have to do today? Speculation is a rumor-whore's bread and butter.

 

[sord]

I think that Steve just wants to be made an honorary member of Blue Man Group. Apple using Centrino in the laptops would no doubt make that happen. I think that's why he always wears black. The apples on the powerbooks and ibooks will glow blue instead of white after the switch. Mark my words.

If I see a comercial with a blue man advertising Mac I will hurt someone...

 

[G.Kirby]

"Mommy! say it ain't so."

AAAAAAAAAAAARRRRRRRRRGGGGGGGGG!!!!!!!!!!!!!!!

 

[bryanc]

I think this is an intentional distraction

The pressure from rumourmongers (like us) on Apple in the days leading up to these sorts of events is enormous. And Steve hates it when his surprises are ruined. Still, it's good to have the rumour mill buzzing before the Keynote.

So I think this has been intentionally leaked to 'stir the pot' and keep us all distracted for a couple more days. That way, when The Steve unveils the new PowerBook G5, or whatever it is he's got up his sleeve, we'll all be surprised and in awe of the RDF.

He's not only a damn fine CEO, he's a showman, and this is how he creates drama.

I'm not saying we should forget about the AppleTel rumour, or stop discussing it, but try to keep some perspective.

Cheers

 

[Nemesis]

This is all fabricated by Jobs himself!

The cnet story is pure baloney ... It makes no sense whatsoever, and Apple is a smart company.

All this "Intel rubbish" started at "All Things Digital" conference, where Jobs demonstrated iTunes 4.9. Who was the source of that important information? Who else was there from Apple that might know such an information? No one! Steve Jobs only. It MUST have been his plot.

WSJ sent that info into cyberspace straight away, and Apple didn't complain about it at all. No subpoena ... nothing. Hmmm, how strange!

Then, during the last week, not a single speculation coming from world famous Mac rumor sites! Apple bought their loyalty this time, so no important stuff leaks out ...

All is prepared now.

I believe all this was a smoke screen to hide a mighty punch in a stomach to all of us -- Apple's gonna introduce quad machine (2 x dual-core processors) and a mobile G5 laptop!

Delivery date -- August/September.

And I bet quad machine will run at 3 GHz! Steve delivers his promise this time, and doubles it!

 

[Frobozz]

I find it weird that nobody has commented on my post yet (#577).

(topic is Intel buying Freescale, with stock tickers and prices as a "proof").

WTF is "FSL-B"? The dashboard widget doesn't know what it is (Quote.com), but Yahoo! Finance does... And FSL-B is less than 1 week old...

Something really is going on if you ask me!

I think your suggestion matches many other posts-- but you have some evidence. It's very interesting and makes sense if you ask me. Intel doesn't care how they make their money. If they are going to crank out PowerPC chips, they just add additional revenue. Same concept with Microsoft selling Office to Mac users. They make more money than they would have (assuming their cost to produce is low, etc.)

I've always felt thereis some grain of truth to these rumors but I will eat my shorts if the Mac went to Intel. It makes sense that, if anything, Intel will be producing some PowerPC chip, or controller, or PCI-E, for use in Apple products. A buyout of FreeScale by Intel would just be a nice little (?) twist to it all.

 

[-Jeff]

Porting Windows apps to Mac OS X on x86 will have little difference than porting Windows apps to Mac OS X on PPC. The only exception is code that contains assembler (which is maybe 0.00001% of all code).

In the end, all code is assmebler. Not 0.00001%

 

[Yvan256]

From Freescale website (in investor section):
FSL.B: Class B shares distributed to Motorola shareholders on December 2, 2004

I guess I was misled by this page because of the lack of any other options than "1d" and "5d".

December 2004 was 6 months ago. There's something broken with Yahoo! Finance, shouldn't it show "3m" and "6m" view options?

 

[DPazdanISU]

one thing i've noticed is that think secret has not commented on this yet. That usually means it's false, however they might be one of the sites thats getting paid off by apple to keep quiet... If only we could jump to monday to find out that the cell processor (the ones in P3) will be the new PM processor (2 teraflops/sec, if we don't move to that I will be jealous of my little cousin's P3 being 10 times more powerful than our 3000 G5 PMs! ) Someone please reasure me that everything will be ok...

 

[mstecker]

Umm... Not Ignorance

Whew...rarely have I ever read such a concentrated amount of misinformation and ignorance. It's not even possible to try to correct all of the nonsense posted so far, so I'll just confine myself to pointing out that "all you have to do is hit the x86 switch and recompile" is obviously written by someone who has never ported software, or even run a compiler. Do you know anything at all about "big endian" (PPC) and "little endian" (x86)? Clearly not. And that's just one potential showstopper...there are lots of performance "gotchas" that require writing routines in different ways for the best speed. So even if your program does work, it probably isn't going to be as fast on x86 as it was on PPC unless you spend a lot of time optimizing it.

Ideally, "hitting x86 and recompiling" would work, but the world is far from ideal. It works for your basic "hello world" program, and increasingly does not work the more complex your program becomes. Writing a truly architecture-independent program is not as simple as you think, and you can bet that since most developers just assumed (I guess) that PPC was going to be used indefinitely, their code would take work to port.

Back in the NeXT days of Fat Binaries, I helped port three large applications from the original NeXT 68K to run on 68K, PA/RISC, SPARC, and X86. Almost all of them (these were objective-C applications) recompiled and ran with a push of a button.

Yes, there were a couple of places where things broke becuase of the little-endian/big-endian disconnect - but you know what? Those were ALL - and are BY DEFINITION places where programmers wrote sloppy code. If you have code that assumes ANYTHING about byte length, variable size or at worst byte ordering, you're writing sloppy code, and you get what you get.
None of these problems were "fixed" by having separate versions of the code for different architectures - they were fixed by changing the code so it wouldn't make any assumptions about all of the above.

Also, it's worth pointing out that when NeXT went to multiple architectures, most applicaions immediately appeared on the new architectures the day they shipped - and many offered their customers recompiled versions without charge.

Matthew

 

[Snowy_River]

700 posts and counting...and not a single one about Apple's shinny new OS on the eve of the WWDC. What's wrong with this picture?

Hmmm... could this actually be a ploy by someone interested in distracting the public from Tiger?

 

[the_mole1314]

I wonder if they will have to have paramdeics at the Apple Stores this monday.

I shure hope this is false....

 

[Warbrain]

The pressure from rumourmongers (like us) on Apple in the days leading up to these sorts of events is enormous. And Steve hates it when his surprises are ruined. Still, it's good to have the rumour mill buzzing before the Keynote.

So I think this has been intentionally leaked to 'stir the pot' and keep us all distracted for a couple more days. That way, when The Steve unveils the new PowerBook G5, or whatever it is he's got up his sleeve, we'll all be surprised and in awe of the RDF.

He's not only a damn fine CEO, he's a showman, and this is how he creates drama.

I'm not saying we should forget about the AppleTel rumour, or stop discussing it, but try to keep some perspective.

Cheers

I was just going to say something along these lines. You beat me to it.

I was just thinking about what other people have said about this. This whole rumor would have a whole lot more strength and weight to it if it were to have originally appeared on Appleinsider or Think Secret. I would be more of a believer if it came on either of those two sites as I know a lot of other people would, too. This whole speculation by C|Net, which is a total ass when it comes to Mac news or anything, and the Wall Street Journal is stupid. It's from two people in the industry who may know something but not everything. It's not going to happen, Jobs is going to announce something different than this and we'll all be happy in the end. Hopefully.

 

[darkwing]

Steve Jobs hates the reds (communists) thats why he made the video 1984. IBM is in bed with them. They are partners with Lenovo which is owned partly by the chinese govt. Steve will be damned if he is going to let the new Russia (China) make chips for his systems. He is going all american here and going with intel.

Now Jobs can kick out systems with Dothan chips that run for 9 hrs and cool as ice in the powerbooks and with dual core too. On top of that if IT staff could put osx on their existing server and workstation boxes they would do it in a heartbeat. The world is about to change here folks, and it will be overnight.

This will be exciting.....

Considering my Powerbook is built in Taiwan and my ipod was shipped straight from China, I'd hardly call Apple "all-american." So much for hating the communists. Besides, Jobs is a democrat anyway.

 

[Snowy_River]

They will have the same incentive that classic developers had to move to OS X native stuff. People running OS X will want to spend their money on OS X native versions. There are also the millions of Mac users with PPC hardware that aren't going to just go away.

If Apple can significantly expand their OS market share, more innovative companies will leverage Cocoa to rapidly develop apps OS X only that reach users unaddressed by their windows only competitors.

It'll take a bold move executed right for Apple to leverage their OS advantage into a significantly larger market share.

That's a wrong analogy. With classic, if the developer didn't move to OS X native, they'd loose all their users as users updated and wanted native apps (and potentially removed classic altogether). For Windows, they'd have the majority of their base - the Windows users - and the ability to say that it runs on Mac OS. Granted there would be some incentive, but not the level of the Classic -> OS X, and what incentive there is could easily be offset by the cost savings of only developing one version.

I'm not saying that there wouldn't be developers who would do it, I'm just saying there would likely be a lot of developers who wouldn't, likely including some that currently provide OS X software.

 

[alandail]

700 posts and counting...and not a single one about Apple's shinny new OS on the eve of the WWDC. What's wrong with this picture?

wait - the whole thread is about Tiger - Tiger running on multile architectures. It's Tiger combined with the declining relevance of Classic that makes this discussion even viable.