Slashdot Mirror
Apple Switching To Intel Chips In 2006
telstar writes "According to C|Net, Apple has officially decided to drop IBM, and will use Intel processors starting in their '06 line of systems. This change was rumored last month. The announcement is expected Monday at Apple's Worldwide Developer Conference in San Francisco, at which Chief Executive Steve Jobs is giving the keynote speech." From the article: "Apple successfully navigated a switch in the 1990s from Motorola's 680x0 line of processors to the Power line jointly made by Motorola and IBM. That switch also required software to be revamped to take advantage of the new processors' performance, but emulation software permitted older programs to run on the new machines."
Wasn't April a couple months ago?
I suppose C|Net could be right, there's nothing technically stopping a 'switch' to Intel, but I don't see what Intel has in 2006 that IBM can't match, or AMD, or whoever.
GPL Deconstructed
remember though, guys, that this doesn't mean that the rest of the system will be compliant with *ATX or the new BTX... so it might not work on your whitebox computer
Investing forum
How odd, Microsoft uses apple dv kits for the xbox 360 and IBM power pc chips and now apple drops IBM for Intel, how freakin' strange is that?
"Jeremy, you need to get to an internet cafe and cut and paste some appropriate sentiments about me from the world wide
How will they make sure MacOSX doesn't run on cheap X86 machines? Or will they use a different chip family?
Open Source Java Web Forum with LDAP authentication
Emulating 68k stuff is easy, thanks to the cumulative efforts of some very talented individuals working on multiple platforms. But what about decent PPC emulation? Are they going to force recompiles of new software, and completely abandon support for old PPC binaries, or are they going to have really slow support of PPC software?
Microsoft goes for the IBM Power PC chips as Apple signs on for the Intel Chips.
I can hardly wait until tomorrow to see what other strange things are afoot.
Fantasy remains a human right; we make in our measure and in our derivative mode... -- JRR Tolkien
So while this is technically feasible, I doubt that Apple is planning a wholesale switch to Intel chips as there is too much invested in both marketing and developer relations. From a technical perspective, Darwin runs on both platforms and there have been long standing rumors of Apple maintaining dual codebases for current OS X releases, so making things run would not be a problem. Developers however, would require significant resources to recompile their code for compatibility. I suspect that the News.com story is only partially correct. Apple has for some time been using Intel chips in their Xserve, and their may very well be additional products yet to be announced. However, think about this possibility: Apple has significant resources devoted to Altivec just about everywhere in the OS, functions that are not available in any currently shipping Intel chip. But imagine this: What if rather than OS X being run on x86, Intel were to produce a PPC chip with Altivec? I do not know what the current licensing agreements are with Apple, IBM and Motorola, but if the licensing were prohibitive, perhaps Apple certainly could help with the reverse engineering of such a chip.
Even that seems like a bit of a stretch to me as I suspect the reality is more like Apple will be using Intel chips in a potential variety of new areas. Chips for networking and WIMAX for example. Or.....given the performance of Intel mobile chips relative to Motorola chips, perhaps as a warning shot across the bow of IBM, Apple will announce that Apple portable systems like Powerbooks will move to Intel chips. Even though I am quite the Apple aficionado, I have to admit that Intel is doing some pretty impressive portable CPUs. Near future plans for Intel portables include built in WiFi and dual cores. However, I realize that this would introduce more than a little difficulty for developers who have a "portable OS" and a "desktop OS" which would suck.
So....perhaps what is really going to happen is that Intel will produce a "portable" PPC chip for something new? Something like a new Newton? If I recall correctly, my Newton 130 ran an ARM chip, and I believe that Intel has the license rights to develop ARM based CPU cores..... Oh please oh please oh please.....
Visit Jonesblog and say hello.
More likely, knowing Steve Jobs, he couldn't get his way with IBM, so he threatened to go to Intel. IBM decided to call the bluff.
1) IBM has opened up Cell, royalty-free.
2) Apple will never let MacOS run on an open platform/commodity hardware again.
3) AMD has virtually no non-x86 CPU tech.
I predict that Intel will either manufacture a Cell derivative or a big-endian, possibly non-x86 propreitary CPU and chipset.
Dual core PowerPC G5 on the way, not Intel.
s p
... I'd say it's [Apple power management] software."
... and it's a limited-size market. A redesigned core might be attractive for future multicore processors" for desktops and servers as well, he said.
... and it's designed for the network world," he said. "It would require some modifications. But it's doable."
http://www.eweek.com/article2/0,1759,1823282,00.a
Analysts: Dual-Core PowerPC G5s Due for Apple
Building a G5 PowerBook could be an aesthetic challenge for Apple. The G5 chip tends to consume more power and produce more heat than the G4. Hotter, more power-hungry chips tend to require a thicker, more spacious chassis and larger, higher-capacity batteries--all of which might lead to a more portly PowerBook.
But, analysts say, versions of the 970FX technically already fit into the power envelope needed for Apple to offer a mid- to full-size laptop in the 5-7 pound range. At the moment, two of its three PowerBook G4 portables weigh in at over 5 pounds.
Aiding portability, IBM has also added a power-management feature to the PowerPC 970FX. Called PowerTune, it can cut the chip's clock speed, therefore lowering its voltage, in order to save on power.
Therefore, a 1.8GHz PowerPC 970FX would be a good choice--it would top the current G4 processor--but power management might still be an issue in some other ways.
The 1.8GHz chip "might be 35 watts or something like that. There are plenty of 35-watt [notebook] processors out there. The big problem is you want to get average power [consumption] to be a lot lower. That relies to a large degree on software management," Glaskowsky said. "If I had to pick a reason why it hasn't shown up yet
Still, not everyone believes that the Power PC 970FX makes a great notebook chip.
"Right now, from IBM's perspective, the [PowerPC] 970 is a pretty competitive part, but they definitely lack a low-power version," said Kevin Krewell, editor-in-chief of the Microprocessor Report, in San Jose, Calif. "The question is, can you get it low enough--25 watts to 35 watts--in order to get it into something sleek enough for Apple?"
To arrive at the right mix of frequency and performance, Krewell suggests that IBM and Apple might need to consider creating a new G4-G5 hybrid instead of delivering a low-power 970.
"The best route would be to develop a new [processor] core that's somewhere between the G5 and the G4," Krewell said, "But that's a significant design undertaking
Apple could also adopt a multicore G4 derivative from Freescale Semiconductor Inc., once the chip arm of Motorola Inc., for its portables, Krewell said.
"That's still a 2006 thing
Representatives from Apple and IBM declined to comment for this story. A Freescale spokesman did not return a call.
Editor's Note: This story was updated to reflect the fact that an Apple representative returned a phone call to eWEEK.com but declined to comment.
You forget one thing: Intel has a tonne of cash. My pet theory (if the rumor is true): The whole thing was started by Intel, not Apple. IBM have been having some wins with the PPC arch, and been getting good press with the PPC/Cell chips. Same goes with AMD and the x64 arch. MS has been getting a little too cosy with both IBM and AMD. So Intel hedges themselves. While they may not have the technical "mindshare" lead right now, they still do have a SHEDLOAD of cash. So throw a crapload, and I mean a crapload, of cash at Apple to grease the skids. It puts MS on notice, gives them another exclusive tier one client and is a blow to IBM's chip division (cutting off a big revenue stream)
How it makes more sense?
This is of course only my opinion but here's how I'm thinking.
1) AMD has the performance edge for x86
2) I _think_ they are more 64 bit centric than the Intel afterthought 64 bit processors (but don't quote me on this).
3) AMD has 64 bit laptop chips (I think).
4) AMD are more energy efficient, at least on the desktop.
5) AMD is not Intel and therefor has less "Wintel" over itself. AMD is "different". Apple likes being different.
The Internet is full. Go Away!!!
I have heard rumors that Apple has been talking to Transitive Technologies about Quick Transit which is a code translation system that can re-map system calls on the fly as well as do very fast optimized recompilation of native code. Think of it as a JIT for processor emulation.
If the claims about Quick Transit are true, and there is no reason to believe that they are false as evidenced by the product's success runix MIPS code on Itaniums (see here), then we should actually see a performance increase for PPC applications (not recompiled) running on OS X x86.
If you were Steve and your apps (as well as everyone elses) ran unmodified on intel hardware faster than it ran on your own, you would probably build some boxes based on intel as well.
There may actually be no need for developers to recompile anything. With Quick Transit built into the OS (let's assume it becomes part of OS X), it would be possible to target x86, PPC or even other architectures and yet run at essentially full speed on any deployment architecture. I know this sounds a bit wicked. It did to me as well. I am sure there will be a bit of a performance and memory hit when your applications are not native, but those hits may be completely overwhelmed by silicon horsepower.
If done properly, this could be a very good move for Apple.
The reason that it can be true that 1+1 > 2 is that very peculiar nonzero value of the + operator
I posted this somewhere else recently, but it applies here today.
== On having two architectures to support (x86 and PPC) ==
So a developer has to compile for two targets, optimise for two targets and ship fat binaries? What will that do to development costs, and ease of development for the platform? What if a developer like Adobe decides to only support PPC development, and tells the x86 customers that they support Windows on x86 and that's it?
And then there are Apple's own issues. They've got to effectively double their Mac hardware R&D to support the two completely different architectures.
And what do they gain?
== On Everyone and Their Dog Switching en masse ==
When I talk to people, they buy a PC because they don't know about Macs. When they do know, they often go PC anyway because they're used to it. There's also pricing. Apple's machines do come at a premium, although I maintain that the iBook line is nicely priced (well, it was when I bought this one a year ago).
Suppose Apple sells a box with an Intel inside. For starters, why would the cost drop? A powerful Pentium is not so much cheaper than a G5. The other components are similar enough. Maybe Apple would use PC motherboards? So why would anyone buy their hardware from Apple?
I don't see any evidence that hordes of PC users are going to drop Windows just to get OS X on x86. I see lots of people on hobbyist sites say that they'd buy it, but they're a tiny minority of a large market. Would that translate to actual sales, or would there be a reason why many of them still wouldn't buy it? What about piracy? How many people would 'try it out' for an extended period of time and never get around to purchasing the boxed copy?
And make no mistake: to make up for the lost hardware sales, Apple would need hordes of switchers to buy those shrinkwrapped boxes. If sales aren't what they hope, there goes the business.
That's a point to remember too - if a CPU switch goes wrong, that could pretty much screw the company. A few billion in red ink, combined with potentially facing near-zero sales when you give away your flagship OS (I'm thinking software piracy and commodity x86 hardware here) and we'll see Apple closing its doors. Some mistakes can only be made once.
==On Software==
I mentioned earlier that developers would likely have to support two completely different architectures, even in the 'best case' of Apple going entirely to x86. There's a legacy of PPC Macs out there that you have to sell to, after all.
That means that initially, there would be zero third party applications for OS X on x86. Not a single one. Maybe iLife really *is* all you need, perhaps with iWork. Over time, new apps would come out, but who would buy the new OS in the first year? That would be a hard sell to Herb and Judy Customer. "Sure, there's nothing you can do with it now, but give it a year or so and... Wow!" (I'm exaggerating of course, but you can go only so far with the iXxxx software before you itch to run something else.)
And what of the developers? I mentioned in an earlier post (not well stated though) that this would be the last straw for them. I'll modify that to "last straw for some of them." They have to learn new optimisation techniques, recompile all of their existing code for the new platform and re-release it - and that's the best case for consumers! Realistically we'll see more developers follow the Adobe and Quark path of holding off for 12-24 months for no apparent reason, and only when the market is safe, releasing their product as a new version with new features for the new platform.
The cost for any app being developed will increase. Not by double, as this would encourage more platform-independant code (well, CPU-independant at any rate), but there would be a definite increase. Who will pay for that? We will! Hooray! Software price rises!
And what if the developers simply say that the OS X platform is too unstable? After all, in five years we've seen a lot
Read carefully. Do the research. It sounds nuts, but this might just be the key to this craziness.
The reason that it can be true that 1+1 > 2 is that very peculiar nonzero value of the + operator
Unlike ASOT, I cannot post actual information without fear of losing my job. Because I'm not sure if the exact specs on what I was given is some type of loyalty test to track down leaks. Since Think-Secret doesn't have it yet.
Now, I have no evidence if we have Intel based Macs hiding anywhere. But, I do have evidence of the next PowerMac (yah, yah we just speed bumped them). But, it means at least one more generation of PowerMacs that are 970 based.
Now it could be we are switching to Intel chips and when I walk in Monday, I will learn all my work has been for naught. But, I think since I have access to a PowerMac unlike any other, I should also be allowed to know about a platform switch, but who knows.
Except that the things are just about melting...
(see the 'non-existant' stories about the capacitors on those things, and the temperature readings)
Oh, and apparently the new G5 towers have even more temperature sensors - those chips run *hot*.
And to think that I was tossed between potentially buying an iBook or a Thinkpad (or whatever Lenovo will call it).
Dan Rather was really, really sure too. And he was the Most Trusted Man in America(tm). I'm sure Stephen Glass and Jason Blair we're really, really sure as well.
Don't blame me, I didn't vote for either of them!
If Apple move to Intel, then they are just another clone maker. If you could then get OS X on any x86 PC, how much more would you pay for the pretty box? Their hardware margins would go in the toilet.
This would be the beginning of the end for Apple as a *hardware* company. They could then focus on iPods, software and the like.
-Charles
Learning HOW to think is more important than learning WHAT to think.
The mathematical number 9533.24 cannot be represented exactly as a double-precision number, because 9533.24 expressed in binary has a repeating string that goes on forever. It is 10010100111101.00111101011100001010001111010111000 01010001111...
When you round it after 53 bits, you have 10010100111101.00111101011100001010001111010111000 0101, or 81889908046875/8589934592 or about 9533.2399999999997817.
Similarly, 215.10 is 11010111.00011001100110011001100110011001100110011 00110011001...
Rounded to 53 digits, that is 11010111.00011001100110011001100110011001100110011 0011 or 7568158436307763/35184372088832 or about 215.09999999999999432.
The difference is exactly 327852904935829005/35184372088832 or 10010001100110.00100011110101110000101000111101011 1000001101 or about 9318.1399999999997874.
However, you cannot represent the difference in double-precision, because it requires too many bits.
The result of a subtraction instruction is rounded, and you get 640337704952791/68719476736 or 10010001100110.00100011110101110000101000111101011 1 or about 9318.1399999999994179.
(Caveat: I produced the above numbers with some quick Maple commands. They could be off a bit, but the concepts are correct.)
It might be nice if calculators intended for the general public used decimal arithmetic internally. (But it still would not be able to exactly calculate 1/3 * 3. There will always be limits to mathematical correctness.) But that is an issue of application design; it has nothing to do with correct floating-point results, as mentioned in the post you responded to. The floating-point arithmetic here is correct.
In addition, CNET's main answer to the insane technical issues that this would involve is, "Steve Jobs said it would work."
The technical issues really aren't that insane. Most high-level code PPC is only a recompile away from working on an x86. Development tools could easily support compilation to some sort of fat-binary (see: 68K-->PPC transition) or dual-binary scehe,.
The only technical sticking point would be Altivec code. Lot of manual work to translate it to SSE1/2/3. However, if Intel could support the Altivec instruction set in hardware... it could be a fairly seamless transition for developers. The technical aspects of supporting it in hardware wouldn't be too hard; I think the only challenges would be legal (patents, etc).
If
OtakuBooty.com: Smart, funny, sexy nerds.
it would be easy to trace Jobs' path through this latest episode:
- IBM bends over and hands them the PPC970.
- Jobs publicly trumpets that the chip will hit 3GHz in a year's time.
- Jobs gets humiliated by the fact they didn't even come close and still aren't there after 2 years.
- Jobs throws constant fits and IBM suddenly considers using their tech for more gracious customers... say, game console manufacturers
- Jobs get jealous of the attention paid to said console manufacturers
- Jobs delivers an ultimatum, IBM calls his bluff, and Jobs is suddenly looking elsewhere for the future of his platform.
Sound crazy? Consider the Altivec debacle and how IBM backed away from the AIM alliance after that. Consider Moto's redirection to embedded processors (hello game consoles) and Jobs' resultant fishing for new tech. Jobs may be running out of corps that will put up with his, uh... particularly demanding negotiation skills. Then again, maybe Monday will bring a nice surprise... a choice of chips for the Mac platform. Here's hoping.
Don't forget that this transition will likely be taking place from mid-2006 (low end) to mid-2007 (high end). From the article:
That likely means (for the high end) some next-generation Intel CPUs like Merom (notebooks) and Conroe (desktops). On the other hand, the PowerBook seems to be the Mac most in need of a modern CPU, so I can't see Apple waiting for Merom (due H1 2006) or a low-heat G5. If these crazy "Intel inside Mac" rumors are true, I think Apple would want Yonah (dual-core 65nm Pentium M) inside the PowerBook.TO START
PRESS ANY KEY
Where's the 'ANY' key? I see Esk, Kitarl, and Pig-Up...
Assume for the moment that this is true (as implausible as it may seem, there is good reason to believe Cnet on this); What should Apple developer do? 1) Applaud it because it could open up Apple's platform to new possibilities. 2) Boo it because now they will have to transition another time after just completing the OS-X transition and support another version. 3) Walk out of WWDC because Apple has now destroyed the Mac platform and "Osborned" the company 4)Wait 2 years to develop any new Mac software until after the Mac/Intel platform is released (and to see if the Mac is still around)
1. Apple has been losing the MHz war for years, and looking worse each year for it (even though Intel has even begun to back away from MHz claims).
Two years ago Steve Jobs said, "We'll be at 3GHz next year!". Next year came, and they were 2.5GHz with excuses. Now a year later they're at a whopping 2.7GHz. It's embarassing, even if it isn't a real issue.
2. Apple doesn't need more difficulting in getting products built and delivered on time.
IBM has thus far not been as reliable as Intel in getting processors out the door quickly.
3. We already knew that OSX worked on Intel, as it had been reported many times that Apple had a working port in-house.
4. Adding "Intel" to the Apple ads will help legitimize Apple computers to some of the people who assumed Macs were just second-rate computers.
Average consumers don't know PPC. They do know Blue Man Group and "Intel Inside". Incidentally, that's why I believe Apple didn't go to AMD, even though AMD has a better product (in the Opteron).
5. Lastly, it's a stock-boosting move.
Every few years tech companies must reinvent themselves, replace their CEO, or otherwise make a radical change. Just look at AAPL closing price, and then compare it to tomorrow's (which is actually today's now!) price.
.sigs are for post^Hers.
This rumor is false, and the proof is in the news cycle this week:
1. Intel pre-announces, e.g. 18 months away, a dual core laptop chip. The last time they made a move like and pre-announced a product days before a "SteveNote" was right before the G5 announcement.
2. Intel demo'ing the Mac mini knock-off over the past couple of weeks.
If Intel had a deal with Apple, they would never be pulling this type of PR which would anger Jobs to no end. This, for me, is the smoking gun.
I suspect Apple has some very big things up their sleeve, and they are still burned about the leak over the Mac mini, so they are doing some misdirection at the expense of Cnet. (Cnet has never been very pro-Apple.)
Jobs at D3 said they have a huge year coming with breakthrough products, and since most leaks seem to happen a few days before the "SteveNote" what better way to keep people pre-occupied than with an outrageous rumor like this to distract the media and rumor mongers.
I think this is misdirection, pure and simple.
Apple sold a lot of G4s to nervous switchers on the understanding thatthey could run all their x86 software under Connectix's Virtual PC software. Microsoft bought Connectix and Virtual PC for OSX hasn't exactly been top of Microsoft's priorities (I assume the Connectix staff are probably working on making the Xbox360 run Xbox software).
This leaves Apple with a hole in it's marketing. If Apple does launch a Mac with an x86 in it, I'm betting it's there as an addition to the G5s, and being effectively a hardware accellerator for an own-brand 'Virtual PC'. It wouldn;t be the first time Apple has done this.
A cheap, headless x86 coprocessor in a Mac Mini sized box that lives on the other end of a firewire cable could be a very interesting proposition.
A pizza of radius z and thickness a has a volume of pi z z a
April fools joke, eh? Would it be even more of a joke if I said that Apple was going to raise the Itanic? The only fools here are people who think Steve is going to port MacOS to the x86.
Think about it. There is NO CHANCE IN HELL that Stevie will put an x86 processor at the heart of a Mac because it'd basically open the door to clones and homebrews of white-boxes running MacOS. Apple wouldn't dare give its users that much freedom.
No, we're talking ITANIC, baybe!
It makes logical sense, and it's a match made in heaven. AMD64 killed Intel's Itanium game plan in the short term causing sales of the machines to make a huge belly flop, plus it killed Intel's long term plans of eventually replacing the x86 with Itanium rather than slapping on 64-bit capability to the x86 architecture, which AMD did anyway and Intel was forced to adopt.
Intel sure as hell won't let billions of dollars of R&D for this new processor to go down the crapper. The solution? Move the processor to a different platform!
Intel can simply offer its wares to other architectures like the Mac. Plus, Intel has the fabs to make sure that Apple won't be starving for chips, which is a problem Apple constantly griped about with Motorola and IBM.
It'd also give Microsoft one more reason to fear Intel. If Intel is playing both sides of the PC vs. Mac war and suddenly becomes smitten with Microsoft (as they're reported to be lately), Intel would be in a position to play dirty.
>>>If Apple were ever serious about this, they wouldve done it when first migrating to OSX. That would've been the perfect time, since OSX needed completely new application support and Motorola had hit a wall with the PPC. But come 5 years later, Apple now has more developers and users then ever before
Well, I can assure you that someone who started developing for the Mac with OS X did not develop Carbon applications. (based on the old os 9 technology).
These applications would be Cocoa-based applications and they would be very easily ported to Intel OS X.
- dj
When HP made its (insane) decision to switch to from Alpha to Intel, it didn't happen overnight: There were a couple more generations of Alpha. This could be the same.
IBM owns the patents to most of the SOI and strained silicon processes. That's why AMD and IBM have an IP sharing contract so they can combine process tech. Intel's strained silicon process is very cheap, but has pretty bad heat issues (gotta have the SOI to buttress the leakage caused by strained matrix).
Kinda leaves Apple in a funny spot. AMD worked with IBM to come up with the process for 90nm SOI/SS at their Fishkill plant which is used in G5 and AMD's chips. I really, really don't see Apple switching to Intel for anything. The reason IBM cannot supply enough G5's is that it is a really hard chip to make. Now, IBM could cross license AltiVec to AMD and AMD could design a dual core AMD64 ISA based chip with it. That would be interesting, an AMD64 chip with AltiVec and no SSE2/3...
That would fix two big issues with the G5...memory latency and production capacity (it's a slightly less complicated chip...although much bigger and more expensive die-wise). Also, both the Fishkill plant and Dresden plants can make such a chip...in addition, AMD has a new fab (FAB36) due to open in a year (for 65nm production). Very fun, but I give this near zero probability.
[RIAA] says its concern is artists. That's true, in just the sense that a cattle rancher is concerned about its cattle.
The hardware in my mini is going to be or, as some have already pointed out, may already be outdated. I knew that when I bought it. What I didn't see coming was a massive code fork to x86, so that 2 years down the road, the newest mac software will possible be 'x86 only'. This boils my blood. I had to go through this crap with the first Windows CE devices, some software was compiled for ARM, and some for Intel, and I was always on the wrong side of whatever the program wanted to run on.
I have also paid a lot for the software I have. How long will adobe and Microsoft support the powerPC versions of their software once Apple goes Intel? I now feel like the lifespan of my machine is seriously degrading. I don't like it one bit.
God is real unless declared integer.
His secret? Keeping up to date with the various Mac-oriented sites, and a gullible Slashdot readership. That's all. I don't think he's a "troll" in the classical sense of the word, but I do think it's sad that there's someone out there with nothing better to do than collect admiration and karma on Slashdot by pretending to work at Apple.
I think it's sad that you're so contrarian that you don't believe the bleedingly obvious, which is that he did indeed work for Apple. He's since been fired. What gave him away was mentioning that he wrote the documentation for launchd. That was a very short list.
(Yes, I work for Apple -- no, I don't know anything, so don't bother asking. What I know about ASOTV comes from the grapevine.)