Apple Computer will likely shift to using Intel chips, while circumstances exist that could well push Dell Computer and Sun Microsystems into a friendly embrace, predicted Bear Stearns analyst Andrew Neff. More and more analysts, editors and even users are getting smoother to the idea that Apple might have to make the jump to x86, simply because IBM/Motorola are not interested anymore in fullfilling Apple’s CPU needs. I wrote an editorial about it two weeks ago (a pretty successful article I might add, judging from the outrageous number of hits received). That article seemed to have re-ignite a number of similar articles on the web since then to several tech news sites.
Analysts Says Apple to Lie Down with x86
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Why make the shift to x86 just as the PC market is beginning to rid of it?
Because:
1. The 32-bit market is here, ready to purchase, and for the next 8-10 years it will still be the No1 platform.
2. Because porting a 32bit OS to a pure 64-bit platform will take more time than trying to port it on another 32bit platform. P3/P4/Athlons are RISC too internally these days (except for the memory), so it wouldn’t be extremely impossible to port.
64-bit sounds nice, but if you are the one who is selling, you have to think where you should sell to, to make more money possible.
Intel are the only one’s really moving away from it in any fashion, and even then the itanium is aimed (for now, at least) at servers, not desktops. furthermore, AMD’s 64-bit offering is x86 compatible. the x86 is too entrenched to be going away any time soon.
of course, this move coul piss off M$ a lot, so you have to wonder if the risk is worth it.
> of course, this move could piss off M$ a lot
One more reason for Ms to cancel out IE and OfficeX for Mac. Food for thought for the latest Ms-Apple happenings… 😉
From the article: “There’s a better than 80 percent chance Apple will make the jump in two to four years, he said.”
Apple has no real choice if they want to survive. With their installed base not embracing OS X, they need to reach out and embrace new markets.
Building their company on x86 gives them entry into every market at better price points. And a wealth of new software. And a whole reboot of the Apple mindset, something that is sorely needed.
x86 will be an enormous success for Apple. Once Apple has made the transition to a mainstream processor, there is no reason why they cannot become the #1 workstation vendor in the world and one of the top 5 PC manufacturers.
#m
I hope the writer doesn’t imply that there will be a Sun/Dell/EMC merger. I can’t let my favourite company to die
If M$ would drop their Mac support Apple would have tobring out a x86 Version, and I’m sure M$ doesn’t like that to happen…. that would be the start of M$’ death
At one of the last press conferences Steve answered the question if they are going to make a x86 Version that they’ll first finish the move to OS X by the end of the year and then they have options.. “and we like to have options”….
Me personally – as a PC user – I would be very happy to see Mac OS X86, and I would immediately move
Eugenia : care to enlighten us on what RISC Memory is?
Didn’t someone years ago said the same thing about BeOS and it’s intel version? Apple *can’t* compete with Microsoft. This Apple embracing the Pentium thingamajig must *never* happen. This will kill Apple for sure. Just like OS/2, BeOS, Novell and pretty soon Linux.
But then again I never listen to these so called ‘analyst’!
I am not talking about RISC being a memory.
I am reffering to the way the CPU is interacting (via translation) with the memory bus which is the only part of x86 that it is not RISC, for legacy reasons.
I think it would be an awful idea to switch to X86. Don’t get me wrong I would love a copy of OS-X for my PC. But Mac would compete directly with COTS PC’s price for price. They would also have tons of problems with hackers reverse engineering Apple roms to run OS-X on PC’s. I do believe it is time for Apple to go achitecture shopping, OS-X should be easy to port to anything under the sun. Motorola has fucked Apple royally, and doesn’t have the R/D talent (they layed everyone off) to keep Powerpc’s state of the art. IBM has been f’d by motorola and Apple, IBM developed some real fast G3’s that Apple refused to buy. I think going 64 bit is Apple’s best bet, and they need to do it before the PC industry. IBM’s power series or Sun’s sparc processors are probly the only game in town, if you want companies that will continue to put develop dollars into the chip. The other option is for Apple to contract or maybe even buy Transmeta. Transmeta’s code morphing technology would allow a painless transition without recompiles. Of course Transmeta would have to make some cranking fast chips to please Apple.
it would only be a problem if Apple tried to make OS X an OS that works on all PCs….
besides…..the boot strap instructions for the OS right now are in the ROM chip…that would be sufficient enough to keep rouge installs from happening.
get rid of IE for all I care…..chimera is a way better browser than IE anyway.
Maybe, but Opera and iCab is faster than all the other browsers under OSX. And I am not even an Opera or an iCab fan… 😉
Apple could go to that, but X 86 is the fastest around.
so it has the same look and feel of the OS…it feels like part of the OS and uses gecko so it is the best rendering pages (IMHO)
when I use my mac, I like consistency
when I use my PC I don’t much mind if I have consistency or not, but that is only casue there is none on that platform 😉
Let me count the reasons that x86 is bad for Apple:
1. Much more realworld performance beyond Photoshop & RC5.
2. A wealth of performance/price points available.
3. Access to all the latest PC hardware, including graphics.
4. Access to all the high performance PC drivers.
5. Cheap 64 bit processors available (AMD).
6. Access to much better compilers and tools.
7. Support from a world-class CPU company.
8. Many more potential markets.
9. Improved time to market for important technology.
10. Access to a magnitude more software developers.
#m
Beside your (again) funny discussion about Apple switching to X86 you all forget one litte thing: Only porting the OS does not do the whole job – where are the OS X/86 applications? Apple is pushing hard nowadays to get their developer community from OS 9 to OS X/PPC. I think the transformation will take about 2-3 more years to complete. Do you really think Apple than says: “Hey that all was a joke – come on developers, let’s move to OS X/86!” I think most of them will disappoint.
Ralf.
Developers would be insanely happy if Apple dumped that fucked Motorola processor and moved to x86.
Why?
Because Apple would sell more machines, all of the OS X as OS 9 wouldn’t run on x86, and the developers would all stand to make more money.
By moving to x86, developers could write much better software using the better tools that exist on x86.
It would be a huge win for all involved if Apple let go of their fear and moved over to a mainstream processor.
#m
well I was writing a counter to your post then I realised you were makeing a joke
:-p
I honestly forsee Apple sticking out PPC a good deal longer. Asking developers to suddenly start making two versions of all their products just so all the users can buy them is too much honestly. And not every developer will have both machines. A lot of things were developed for just x86/PPC BeOS and not the other, and the same problem would arise here. PPC inovation my not be moving along at a good rate, but software inovation (MacOS X continues to improve..) can more than keep them alive and well. I for one have gone Mac simply because I like the look and feel of everything in the machine, not just the processor.
Nope.
if the developers wrote poratble programs and stuck to the API only, they would not have to worry at all…..even if 10% of the code is assembly to speed up a bottle neck, it is not a total rewrite, and 3 programmers (who will bitch probably) will have that 10% prototyped in a few months and by the time the next release comes out, the company can have a fully working program……and alll apple needs to do is anounce there intentions and that will get the companies moving.
what better dev-tools exist on x86? Haven’t seen any yet!
By the way – the post earlier says: “Moto/IBM has no interest to fulfill Apples needs…” How do you know this? Are you in bed with Lou Gerstner? Last time Jobs said: “he likes to have options… …at the moment there is no need to think about x86… …Moto’s roadmap for the PPC fits very well for Apples need”.
Be has such small market share that most people that developed for it did not care…
The developers just went though a transition of moving apps over already, some are still in the process (Quark). Not they might be in for another? Ouch… That’s gonna hurt.
Honestly, if Apple moved to x86, they could potentially sell as many computers in ONE YEAR as they have sold Macs in TWENTY YEARS.
It makes no sense for Apple to limit themselves by staying on the Motorola processor.
#m
Jeremy,
let’s put this into an equation:
windows + x86 = PCs
BeOS + x86 = PCs
OS/2 + x86 = PCs
Apple + PPC = Macs
Apple + x86 = ???
try to guess what people are gonna say when they see a pc on their desk with Apple? “Hey look everyone! My PC is runnin’ Apple’s o.s.!!!” And guess what’s gonna happen after that? They are gonna be overrun with calls to support this and that. Apple can’t afford this and that. Be couldn’t. And IBM also (everybody was dropping OS/2 support cuz Windows 3.1 was good enough YUCK!)
Hardware is abstracted by the OS….as long as most of a program (90% or so) is programed to the APIs of the OS, then a company will have little problem porting….
an API change is a lot harder to reprogram than is a hardware change.
Steve Jobs is already thumbing his nose at Microsoft.
I can see Apple doing this, assuming that Motorola/PPC is really running out of gas. But, it couldn’t be an immediate thing. As pointed out, Jobs almost (seemingly) casually said that all options are open once OS 9 is dead and the move to OS X is pretty much complete. That is an amazing thing for someone like him to say. He never says anything in public without a purpose.
I still am anxious to see what the Power Mac upgrades will be. If it’s only 1.2 GHz, Apple should speed up this scenario <g>.
1. Much more realworld performance beyond Photoshop & RC5.
My iMac tends to a lot of things pretty well, where do you actually see performance problems on a Mac?
2. A wealth of performance/price points available.
3. Access to all the latest PC hardware, including graphics.
4. Access to all the high performance PC drivers.
Even if Apple switched to x86, why on Earth would they give up their hardware. Just because they use the same processor doesn’t mean you could just install it on any PC. One of the reasons Macs have always performed well is because you don’t run as much a risk of having many, many hardware combinations.
5. Cheap 64 bit processors available (AMD).
Tied down to a legacy architecture, if they wanted to go 64-bit, might as well go fresh.
6. Access to much better compilers and tools.
There are some pretty good compilers already on Macs, why worry?
7. Support from a world-class CPU company.
AMD or Intel? And again, why not go for something new?
8. Many more potential markets.
Why would the markets improve? I doubt the machines would really drop in price that much.
9. Improved time to market for important technology.
Huh?
10. Access to a magnitude more software developers.
That buy these new machines, I’m sure there will be millions… oh wait, they already have programmers on PPC, who knew.
Most people tell me they love to get Windows XP onto their G4s. Face it, Mac OS X is slow and still needs a lot of work. Let’s wait after the 24th to see if this situation has change for the better.
the only reason that all those were equal to PCs is becasue you had an open platform and all OSs were interchangable.
OS X will not be on an open platform (a CPU does not make it closed)
if the ROM chip has all the boot strap instructions for the OS in it, then it will take seriouse hacking to get it to work on a normal PC…and apple has the lawyers to sue anyone who sells such a hacked chip.
OS X is just missing a couple numbers… Steve is waiting until everything’s just right and then he will reveal the truth!
OS X86 😉
Going from NextStep/68k to NextStep/i386 was not a hard port. It was mostly a recompile.
Once OS X is ported over to Intel/AMD, moving apps over will be simple.
#m
1. They do not build standard PC Architecture Boxes.
2. They keep making better designs, and drop prices a bit.
3. They become as serious about servers as they are about desktops.
Apple hardware is superior to most PC manufacturers (perhaps with the exception of Compaq Computers) If they create their own architecture, they will not lose any of their hardware business. They don’t need to license their architecture to anyone else but they can do so for a small hardware fee and OS licensing fee!
As long as they build better looking, better performing, proprietary Intel Boxes (Workstations and Servers). They will beat the hell out of Microsoft.
I think Steve Jobs has yet an other opportunity to really shine again!
Time will tell.
ciao
yc
“try to guess what people are gonna say when they see a pc on their desk with Apple? “Hey look everyone! My PC is runnin’ Apple’s o.s.!!!” And guess what’s gonna happen after that? They are gonna be overrun with calls to support this and that. Apple can’t afford this and that. Be couldn’t. And IBM also (everybody was dropping OS/2 support cuz Windows 3.1 was good enough YUCK!)”
#1: Why would someone look at a Mac and think its a PC? Unless they opened it up and saw the CPU, why would they say such a thing?
#2: Why would they get overrun with support calls? I have news for you: they already have a tech support department and they haven’t gone bankrupt yet.
#3: Besides, they charge extra for support beyond a year. Those other companies probably had much higher costs for that reason alone.
The original Mac was a conscious decision not to use x86, as were the NeXT workstations.
There seem to be two questions here: What Apple *should* do, and what Apple *will* do. Apple *should* do so and so, but is that what they *will* do? As hard as I try, I cannot envision an x86 Mac in any form. I look at Steve Jobs and that’s just not what he’s about – here is a man who is perfectly content selling high-priced computers to a relatively tiny market that loves them. It’s what he’s been doing his entire career (except the parts of it at NeXT which were not spent in the hardware business). He has no lust to dominate the world or put an Apple on every desktop. He favors vast margins over vast Dell-like economies of scale. His products fill a niche.
Nobody knows exactly how Steve thinks but the man himself, but if he thinks the way I think he thinks, x86 is little more than a plan C to him. And really, when you think about it, there is no dearth of options besides x86, and what follows are only a few of them:
– Switching to a different non-x86 platform; MIPS and POWER-derivative being two popular suggestions
– Increased reliance on dedicated hardware coprocessors
– Multiprocessor systems across the board, with quad on the high end
– Continuing to go about PowerPC-as-usual, enhancing the compiler and going about some major optimization of the OS to wring every last drop out of the PPC while doing whatever it takes to move PPC design & manufacturing to more capable hands
– Repositioning the Mac for new markets which don’t rely on strong CPU performance
I could see an x86 Mac as a last-ditch bail-out move, but nothing else.
I say there is a zero chance of Apple running x86!
What the heck is wrong with the PowerX platform?
IBM/Apple/Motorola have way too much invested in that platform to just dump it. Look at the specs of the power series. That is real computing there pardner, not a toy x86 box. Apple problem is to make a PowerX box that no other OS can run on.
>What the heck is wrong with the PowerX platform?
Insanely high prices. Even for the “cheap” models.
And just to clarify somthing:
Apples currently are very difficult to distinguish from PCs on the inside. The only things that seperate Apples from PCs are the design/erganomics(classy look
, the software(OS, Apps, etc…), the limited but usually high quality Apple compatable hardware world, and the PPC based CPU.
Now, we can all agree that limited hardware, despite ususally being good, is still a limitation, and now the PPC looks like it may be as well. If this is the case, then a switch to x86 could go very far to remidy BOTH of those limitations, so that the best parts of an apple, the design and the software, are more afordable and accessable then ever before!
Apple wants to change the world. To give people the power to be their best.
Staying on an antiquated CPU architecture made by an embedded processor company is not going to do anything good for Apple.
Even vs. a Mac with dual 1000Mhz processors, a single 2.53Ghz P4 still beats Apple in almost all performance tests.
Intel will have 3Ghz P4 this year. 3-4Ghz next year, in both single and dual models. And as Intel is a volume manufacturer, many people will be able to afford them.
So we find Apple’s secret strategy… they are moving the entire OS to the graphics processor!
OS Xtreme.
#m
“So we find Apple’s secret strategy… they are moving the entire OS to the graphics processor!”
Maybe not just the greaphics prossesor, but having a deticated co-prossesor to handle just the OS or large parts of it is actually a rather neat idea, it frees up the main CPU for actual user tasks. If anyone can reach that level of hardware-software intigration, its deffinatly Apple.
#1: Macs are easy to spot. They are the ones with attractive cases and usually carrying the nifty new features. Ex.: USB, Firewire, Desktop LCD monitors, etc…
#2: Huh? Macs don’t have the vast quantities of peripherals to support like Windows has to.
#3: Who does most of the drivers on the Mac platform?
Power4 and SPARC are in now way competitors to x86 in desktop space. Not only is there the cost inherent in these chips, but performance would be terrible. If take a look at the SPEC benchmarks, you’ll notice that a 2.5 GHz P4 beats the fastest UltraSparc IIIs quite soundly. And it reachs a good 2/3s the performance of a Power4. Now, take away the mad bus architecture in the Sun, and trim the 128MB of cache on the power4 to something sane, and what do you have left? A chip that really can’t compete with the P4. No, the strength of the USparc and Power4 chips lies in their ability to be put in machines by the fist-full. Either chip, caught alone, is just prey for the x86 chips.
Remember Be Inc.???
They had their own hardware aka BeBox ( w/ PowerPC).
Then they killed it.
they then ported to macs.
Then they ported to x86.
But why they (be Inc.) still did not succeed ??
Perhaps because the PC OEMs ( Dell, Compaq/HP, Sony,etc..)
just won’t preinstall BeOS nor any other non-M$
operating systsm on their systems.
What make you think that if Apple does ship their
OS X to x86 PCs, the PC OEMs will ship the OS X
on their PCs. Do you really think M$ will let PC OEMS
ship OS X preinstalled???
If I were Apple, I would definitely NOT follow
the path that the Be Inc. has gone…
If apple does x86, only the nerds will even notice. The general users will notice a slight price drop and performance increase.
The worst thing that apple could do is start competing with Dell. What has been keeping Apple alive is their innovation and uniqueness, not rolling out tons of beige boxes. The last time they had clones they nearly went out of business.
Yes, it’s class-less to reply to oneself. I’m a bohemian, then. I must mention that the Alpha was the only potential competitor to x86 in the one-to-two CPU machine market. A 1GHz Alpha still can compete well with the newest x86 chips, and doesn’t need the huge supporting architecture that the SPARC and Power chips are coupled to. The Alpha, manufactured in bulk (perhaps with a 0.13 micron process to kick up the clock speeds) could be a reasonably-priced competitor to x86, perfect for the high-end prices commanded by Macs. And, unlike PPCs, they’d actually deliver high-end performance. Sad, how the good ones are always the first to go…
would programs be able to run straight on osX/x86 and osX/ppc or would they need a simple recompile, complete rewrite, small rewrite? what would it take? thanks.
I guess there will be a transition stage where apps will be compiled as ‘fat’ binaries for both ppc and x86.
What percentage of the pc market does Dell have? What percentage do other manufactuerers have? Because that’s the point.
BeOS showed that porting, even if you stick to the API isn’t always going ot be easy. Different processors will even handle some “standard” things differently so most code would atleast need be tweaked for both platforms.
also..
Staying on an antiquated CPU architecture made by an embedded processor company is not going to do anything good for Apple.
And x86 is young and full of life? The only real future I see in x86 is AMD’s Hammer. They are pulling out some incredible clock cycles, they are getting faster and faster each year, but the rate of the speed benefits only continues to decline. When you compare a 500mhz PIII to a 1GHZ, the difference is phenomenal. But when you compare just the 1GHZ to the 2, the boost is hardly as impressive.
What are it’s costs like? If it is at a fairly reasonable price (I say reasonable because macs already command a premium) then Itanium looks promising to me. You have a 64 bit chip with promising SPEC’s and without real major support yet. Intel wouldn’t mind because it would shift attention away from the future of x86 which intel may have problems with. AMD is going 64-bit while Intel is just ramping speeds (which will reach it’s limits at some point) and yamhill is looking spotty right now. If apple could set up significant parts of the os up with floating point processing then OS X could become a good deal faster.
Let’s take a look at the roadmap. Motorola has nothing that can compare with the x86 roadmap over the next few years.
4Q2002
AMD Athlon 2800+ (Barton) is expected to be released in Q4.
Intel Pentium 4 3.0Ghz (Northwood A) CPU is expected to be released in Q4.
Intel Pentium 4 3.06Ghz (Northwood B) CPU is expected to be released in Q4.
Intel Xeon (Prestonia B) 2.8Ghz is expected to be released in Q4. This CPU will have a 533Mhz FSB.
Intel Xeon MP (Gallatin) 1.5, 1.9 & 2Ghz are expected to be released in Q4. Gallatin is the 0.13micron successor to Foster MP, and is expected to be available with either 1Mb or 2Mb of L3 cache and feature Intel’s HyperThreading SMT technology.
1Q2003
Intel Xeon 3.0 & 3.06Ghz (Prestonia) are expected to be released in Q1.
AMD Athlon (Clawhammer) 3400+, the 64-bit successor to the Athlon, is expected to start shipping in Late Q4 2002 or Early 2003, with the first iteration being single processor only.
2Q2003
Intel Pentium 4 3.2Ghz is expected to be released in Q2.
AMD Opteron 4000+ (Sledgehammer), the large cache (1Mb L2), 8-way multiprocessor version of Clawhammer, is expected to start shipping in H1 2003. The on-die memory controller of Sledgehammer will support a 128-bit DDR interface (compared to the 64-bit DDR interface of Clawhammer) for quicker memory access to a larger quantity of RAM (up to 8 DIMMs per CPU).
AMD Opteron (Clawhammer DP) is expected tp be released in H1 2003. This is the Dual Processor capable version of Clawhammer. Unlike its predecessor this CPU is expected to join the Opteron rather than the Athlon family.
AMD Athlon (Clawhammer) 4400+ is expected to be released in Q2
AMD Athlon MP (Barton), the follow-up to the Thoroughbred core Athlon MP, is expected to be released in H1 2003. Barton MP will contain 512Kb of L2 cache and potentially a 333Mhz DDR FSB and some Hammer optimizations.
3Q2003
Intel Pentium 5 (Prescott) IA32 CPU is the 0.09micron successor to the Northwood core of the Pentium 4 due between Q2 and late Q3, depending on the state of the CPU industry. Prescott is expected to launch at 3.2Ghz and higher and will introduce Intel’s HyperThreading technology to the desktop (first introduced in Prestonia). Prescott will also feature a number of architectural improvements including a larger 1Mb L2 cache and potentially an updated instruction set. Prescott is expected to feature an 667Mhz (166Mhz Quad Pumped) Front Side Bus.
4Q2003
AMD Opteron 4400+ (Sledgehammer) is expected to be released in H2 2003. This processor is expected to be the first to be manufactured on AMD’s 0.09 micron process.
2004
Intel Tejas IA-32 processor is expected to be release in H1 2004. Tejas is the successor to Prescott and may incorporate some kind of IA-64 support within the core and a faster front side bus speed (1200Mhz?).
2005
Intel 0.07um process is expected to be released in 2005. The clock speed of Intel’s NetBurst architecture processors is expected to be around 8-10Ghz on Intel’s 0.07 micron process by 2005. The core voltage of these processors is expected to be around 0.85v, and these processors are expected to have around 400M transistors (around 10X greater than the P4).
PC5400 DDR II SDRAM is expected to become available in 2005. PC5400 modules will operate at a clock speed of 667Mhz, giving a memory bandwidth of 5.4Gb/s.
DDR III SDRAM is expected to become available in 2005.
#m
Ty: IBM’s power series or Sun’s sparc processors are probly the only game in town, if you want companies that will continue to put develop dollars into the chip.
The question of Sun’s SPARC processors replacing the PowerPC is sort of interesting. Apple and Sun both use OpenFirmware. The boot process and basic driver functionality are controlled by FORTH. Sun users see this as the “OK” prompt. I think there’s some key combination that can bring it up on a Macintosh. This common firmware heritage is what allowed the occasional Apple engineer to run MacOS < 10 on a SPARCStation. At least according to rumor; no MacOS/SPARC was ever released into the wild (that emulator doesn’t count).
However, moving to SPARC is extremely unlikely. Neither Sun’s nor Fujitsu’s UltraSPARCs are known for their raw uniprocessor performance. I’ll admit that the I*e (embedded and very low workstation) and I*i (medium workstation) are cheaper and draw less power than the full blown UltraSPARC III models, and might be suitable for Apple, but their price-performance doesn’t really approach X86.
Switching from Motorola to IBM for the high-end Apple processors is a bit more likely, though you certainly wouldn’t see a dual-core POWER4 in your laptop. One very large potential stumbling block for Apple’s switch to IBM is Altivec. Altivec is what busted the original IBM-Motorola PowerPC alliance — while it’s useful for Motorola, IBM didn’t find it worthwhile. Given the amount of emphasis Apple’s put on Altivec, I can’t believe that they would be able to dump it. In order to switch, not only would Apple need Motorola’s permission for IBM to use Altivec, IBM would have to be willing to use it in the first place.
Someone mentioned MIPS. A few small high-end companies use them in very expensive servers (SGI, HP’s Tandem division, Fujitsu-Siemens’ RELIANT). HP and Fujitsu-Siemens are transitioning away, to Itanium and SPARC. Everything else is embedded. There are no advantages I could see for Apple’s MIPS adoption.
Personally? I’d bet on Apple going either with the multi-processor route and/or with IBM for some time to come. An architecture transition would be very painful. The performance gap is not yet wide enough. Several years from now? Maybe.
Yours truly,
Jeffrey Boulier
Perhaps the new Prescott instructions are the next version of SSE which has been co-designed with Apple and even more capable than Altivec?
That would give Apple a year to make the transition over to x86 which sounds about right.
Apple could make quite a splash being the first/best manufacturer to have a Pentium 5 machine available and running the next generation of Mac OS X. Would be very cool.
#m
I don’t see x86 as possible. If Apple is going to migrate off of the Motorola chip they are going to get have get real benefit and that means a 64 bit chip. As I see it:
1) Stick with Motorla and go for lots of subprocessors: This is the easiest and since SGI has already done most of the hard work on getting high end graphical workstations with so-so processors to run excellent 3D graphics…. Having used the RS/6000s in 1990 (7 processors: CPU, dedicated vector processor, dedicated memory subsystem, dedicated graphics subsystem… these things were 10x as fast as if they depended on a good CPU alone).
2) Partner with IBM and bring the costs down on the Power4 chip. This might be a great way to seperate the PowerMac from the iMac. A year from now the iMac’s running G4/G5 the PowerMac runs the Power4 (partially stripped).
3) Migrate to a full 64 bit chip of some sort. Lets face it they are running BSD and using GCC. The Unix crowd has already done the hard work here. So again the PowerMacs go 64 bits the next year the iMac and the eMac. A year or two later they stop releasing updates for their 32 bit models, 32 bit development software… and force a hardware upgrade to everyone left.
Apple could make an amazing 64 bit x86 workstation using the various “Hammer” AMD processors.
(and)
They’d have compatible and affordable processors for use in their low-end machines, iMac to eMac, and their laptops, iBook to tiBook.
IBM Power4 (more likely the upcoming Power5) is not going to be suitable for a range of machines from low-end consumer domes to laptops to high-end workstations.
Only x86 offers that range today.
Itanium2/Itanium3 are interesting chips that would be perfect for a stellar 64 bit workstation. It may be that Itanium3 has a supercharged x86 processor in it so it does offer a migration path to 64 bit world. Provided Intel does not come out with its own x86-64-like set of instructions.
In Itanium world —
Intel Madison 1.2 – 1.6Ghz IA64 CPU is expected to be released in 2Q 2003. Madison is the 0.13 micron successor to the McKinley (‘Itanium 2’) IA64 CPU. Madison is expected to based around the McKinley core, but will contain a larger 6Mb L3 cache.
Intel Montecito IA-64 CPU is expected to be released in 2004. Montecito is the follow up to Intel’s previous Madison IA-64 CPU core, and is expected to incorporate some technology from the Alpha CPU line.
Either AMD or Intel offers Apple much more than any other processor vendor.
#m
“#1: Macs are easy to spot. They are the ones with attractive cases and usually carrying the nifty new features. Ex.: USB, Firewire, Desktop LCD monitors, etc…”
Exactly.
“#2: Huh? Macs don’t have the vast quantities of peripherals to support like Windows has to.”
Exactly.
“#3: Who does most of the drivers on the Mac platform?”
Err… Apple I would assume, though many third party hardware makers do their own.
Im afraid I probably missed it. What was it you were trying to say with this post?
“What make you think that if Apple does ship their
OS X to x86 PCs, the PC OEMs will ship the OS X
on their PCs. Do you really think M$ will let PC OEMS
ship OS X preinstalled???”
Why would Apple want to ship their OS on PCs? Thats not even the subject of this discussion! Apple would never do such a silly thing. Even if they did shift to x86 achitecture as well, they’d still kill their own hardware business if they licensed to other manufacturers!
What made you think anyone here thought that??? Well, except for Ronald, he seemed to think the same thing as you.
you are a fool if you think Apple would ship only the OS.
try thinking that they will ship there own propriatary X86 computer and will not allow OS X to boot on any other machine.
… That is, until somebody hacks it.
so…no problem for them.
When you’ve got Carracho and Limewire.
Think about this; if Apple went to Intel and said we’d like to buy this many processors per year, can you discount for us, I would think Intel would jump at it. I know that Intel likes to a) tweak msft’s nose when possible, and b) gain customers it doesn’t already have. By selling apple the Itaniums (or whatever the version is), they’d also effectively lock AMD out of future sales in regards to Apple.
In effect, I think it would be a major win for Intel in the workstation market.
Apple could not manage to eek out anything more than a pathetic battery life with Itanium in their notebooks without it causing third-degree lap burns. Apple’s notebooks are known for their good battery life. Barring tremendous advances in battery & cooling technology, I don’t think it could happen.
Remember, Apple needs an architecture that can scale from a small notebook to a high-end workstation. That narrows it down to PPC, MIPS, and x86. It’s just that x86 is so dirty.
There is also the price issue… Even with volume discounts, Itanium would be very expensive relative to PPC and x86.
“The original Mac was a conscious decision not to use x86, as were the
NeXT workstations.
”
In the early 1980s, the 68000 was clearly superior to the x86 line.
That stopped being true when Motorola ceased developmement of 68k for
desktops.
In the early 1990s, the PPC was clearly superior to x86. That is no
longer true. Motorola seems to have little interest now in developing
the PPC for desktops.
IMO the best way out of this for Apple is to go the Amiga/Game machine
route – farm out the CPU-intensive stuff to custom chips. Jaguar is
the beginning of this.
I will read these posts after seybold (and see how time changes everything).
Pro mac users will be taken care of.
Why x86 is bad (sarcasm, i belive) for Apple (according to #m), and an analysis of his points:
1. Much more realworld performance beyond Photoshop & RC5.
Okay. Who knows, so I will let that stand. Assuming Apple rewrites a sh*t load of ppc asm, and altivec optomised code. Otherwise you could see a speed drop in alot of places. Also, apple will probably have to write an altivec emulator, and ship libraries with the (now slow a hell) altivec calls they implemented to ease transition.
2. A wealth of performance/price points available.
Yeah. Knowing Apple this is probably rather bad. People are going to want their P3/P4/Duron/Athlon/Transmeta based systems, but Apple is going to want to ship exactly two types of chips: one slow, one fast. Apple will also want no more than three speeds in any one chip family.
But from a users point of view this might be good (so that pressure falls on Apple).
3. Access to all the latest PC hardware, including graphics.
Okay. Here is were you get dumb. I don’t see any modern graphics hardware that doesn’t fit in my Mac. The problem is drivers. Changing to x86 isn’t going to magically write drivers…
4. Access to all the high performance PC drivers.
And to continue on from the last answer: wrong! Obviously windows drivers will not be drop in replacements. Now think for a moment: What would happen if Linux added a simple kernel interface to windows video drivers–Just pop in the .DLL, change a few setting, and you are ready to rock. A) speed hit. B) no new linux specific drivers from any companies (the OS/2 vs. WIN3.1 argument). C) one huge mess as you have newbies trying to get a new driver out of a SETUP.EXE, into the propper directory, and configure Linux, and XFree86 to use it. WINE would sure fall over on an installer. Using windows gfx drivers would be poison to any OS (possibly even to a microsoft OS).
5. Cheap 64 bit processors available (AMD).
I may be one of few Mac users who is looking forward to getting my hands on a 64 bit CPU (and I think Intel and AMD both have respectable designs), but even I think that is one nightmare Apple does not want to step into. They would sit still until either Hammer or Itanium emerged the clear victor. With PPC they could make the jump much sooner.
6. Access to much better compilers and tools.
Well, I know the x86 version of gcc gets more than its share of love, but I think the PPC version is quite spiffy as well. They may even *lose* a compiler in the process. How likely do you think Moto would be to port CodeWarrior to OS X/x86 after Apple severs relations with them? The best you could hope for would be icc after Apple (inevitably) goes to Intel.
7. Support from a world-class CPU company.
Well, aside from a sarcastic comment to the effect of “We’ve seen what kind of suport Apple gets from CPU companies with bigger markets” I will let this one fly.
8. Many more potential markets.
This make no sense, unless there is a substantial number of people out there who won’t buy any computer without an “Intel Inside” sticker on it.
9. Improved time to market for important technology.
Uh. No. They still have to build it, and only one piece would be different.
10. Access to a magnitude more software developers.
x86 ASM programmers out of work?
And her is (most of) post number two:
Honestly, if Apple moved to x86, they could potentially sell as many computers in ONE YEAR as they have sold Macs in TWENTY YEARS.
Oh sure. I can just imagine the ads now: “Double the speed at the same price“. Sales would drop like a rock since CPU is a small percentage of the cost of a computer, and people certainly can compare a 2Ghz Windows Machine for about $1000 to a 2Ghz Mac for $2500.
[RANT]
With hardware you can pick two: good, fast, cheap. Apple tends to pick the first two. PC manufacturers usually pick the latter two. I find it werid that the same people who are pointing out “You can get a [fast] PC for $600” are also the same people constantly crying over crashed drives, and foobared computers, and they never see the connection. I have never lost a drive. I have never had a (less than a decade old, at which I start really abusing them) Mac die on me. You get what you pay for.
If Apple goes to x86, I will go to Debian for a couple years, then evaluate the market.
[ENDRANT]
Actually sooner than seybold, say in ~2 weeks.
[RANT]
With hardware you can pick two: good, fast, cheap. Apple tends to pick the first two. PC manufacturers usually pick the latter two. […]
If Apple goes to x86, I will go to Debian for a couple years, then evaluate the market.
[ENDRANT]
I think you missed something:
With PC, you can pick three: good, fast, cheap.
With Mac, you can pick three: good, slow, expensive.
It seems the “two” you always get to pick with Mac are:
slow
expensive
With my $600 PC, I can switch to any one of my other two PC’s while I get a super affordable and available replacement part.
When your fancy white iPaidTheBigBucks computer has a problem, you’ll be stuck with nothing while you wait for Apple to use their secret tools to take apart your little white mushroom and repair its tender little fungus insides.
Unless there is a fucking miracle at Motorola or Apple lowers their margins and gives you 4 CPU’s for the price of one… Mac is destined to be SLOW. And it would be a genetic mutation for Apple to give their customer a decent price.
So have fun with Debian.
And for the coprocessor people… how about a $3 sound chip for goodness sake? I know Apple is the greediest damn computer company around… maybe they can offer another .MAC service… “goodsoundchip.mac” for $100/year.
And maybe another $0.25 for another mouse button? Ooops… forgot… everything is on subscription price now, bundled with stuff you don’t give a damn about.
Sp how about “alltheextras.mac” for $250/year that gives you the sound chip, the extra mouse button, a giant RC5 benchmark poster, and a t-shirt?
The t-shirt: “all your dollar are belong to apple”.
#m
I wrote an editorial about it two weeks ago (a pretty successful article I might add, judging from the outrageous number of hits received).
Hits doesn’t mean they support the article, but it does mean advertising revenue.
Why make the shift to x86 just as the PC market is beginning to rid of it?
Really? I didn’t notice that. Since Itanium is only targeting the market Sun, IBM and SGI are in, high end 64-bit workstations and servers, everything else for the PC is, well, x86.
P3/P4/Athlons are RISC too internally these days (except for the memory), so it wouldn’t be extremely impossible to port.
It wouldn’t be impossible to port if it was CISC anyway. Most of OS X is portable, and early versions of OS X in Apple were developed on x86 before moving to PPC. Perhaps QE and some other components may not be purely portable, but porting it would be easier than the port of OpenStep to PPC.
> of course, this move could piss off M$ a lot
One more reason for Ms to cancel out IE and OfficeX for Mac. Food for thought for the latest Ms-Apple happenings… 😉
Apple on PPC and Apple on x86 wouldn’t be different, except that Apple would save money and have better hardware to handle OS X. But do I think Apple hardware would be not overpriced? Nope. If Microsoft could support Apple while they are on PPC, there is no reason not to support it on x86.
Plus, most likely Windows could run on new x86 Macs, so people wanting to use Windows and Mac OS could buy a Mac and buy Windows – Microsoft makes money here.
Building their company on x86 gives them entry into every market at better price points…. there is no reason why they cannot become the #1 workstation vendor in the world and one of the top 5 PC manufacturers.
Trust me, Apple would still charge a premium for x86 Macs. After all if Sony could do it, they could too. They would surely cut cost by the move, but would the Mac get cheap PC zealots like myself? Doubt it.
I hope the writer doesn’t imply that there will be a Sun/Dell/EMC merger. I can’t let my favourite company to die
It would failed up faster than HP/Compaq merger. (Yes, I still think the merger is as stupid as a kid jumping into a tank of H2SO4.
Didn’t someone years ago said the same thing about BeOS and it’s intel version? Apple *can’t* compete with Microsoft. This Apple embracing the Pentium thingamajig must *never* happen.
Apple would die off faster than you could say “Be Inc.” if they abandon its hardware business. What i think is that Apple would move to x86 and still keep its hardware close (e.g. to run OS X, still have to buy a Mac). To Microsoft, this means nothing, because considering Apple’s history, on the price war, Apple couldn’t and wouldn’t even try to win.
But Mac would compete directly with COTS PC’s price for price.
They aren’t now?
They would also have tons of problems with hackers reverse engineering Apple roms to run OS-X on PC’s.
They don’t need to reverse engineer it, they have the specs. Plus, they would most likely build a new type of ROM for its x86 machines.
…OS-X should be easy to port to anything under the sun.
But didn’t you said that porting to OS X would be hard?
I think going 64 bit is Apple’s best bet, and they need to do it before the PC industry.
Then they better do it fast, because Hammer would come out by end of this year. (Yes, PCs won the race before Macs even tried, though the race isn’t that significant because currently most 32-bit users aren’t still being limited by 32-bit’s limitations).
IBM’s power series or Sun’s sparc processors are probly the only game in town
Power series aren’t built for normal workstations and godforbid, laptops and notebooks. The level of heat emision and the amount of energy it needs, will, well, i think you get the picture.
Sparc on the other hand is one of Sun’s only competitive product, I doubt they would let a potential competitor use it. Theye never let anyone else use it. Also, it has the same heat and power problems as the POWER series.
other option is for Apple to contract or maybe even buy Transmeta.
Transmeta may solve Apple’s cost problems, and cause iBook and TiBook’s batery life to sky rocket, but it doesn’t solve any of PPC’s speed problems.
Transmeta’s code morphing technology would allow a painless transition without recompiles.
Their code morphing thing is made for the x86 spec. Counting by the time the x86 one was made, it would take some 2-4 years to even have something usable, then after that can Apple switch to it. I don’t know why if any of this is worth it anyway.
get rid of IE for all I care…..chimera is a way better browser than IE anyway.
And way more unstable 🙂
Maybe, but Opera and iCab is faster than all the other browsers under OSX. And I am not even an Opera or an iCab fan… 😉
Though it seems Opera and iCab’s future on Mac OS seems bleak. iCab, IIRC, never had a stanble release, while Opera still is at 5.0 Beta for OS X….
when I use my PC I don’t much mind if I have consistency or not, but that is only casue there is none on that platform 😉
Don’t blame the platform, blame the third party developers (you could blame MS for Office’s non-standard look, but the UI follow’s Windows).
10. Access to a magnitude more software developers.
I don’t see how this would change from the switch, cause Apple would still use GCC, would still push the same APIs etc.
Only porting the OS does not do the whole job – where are the OS X/86 applications?
Can be easily accomplish by the same solution picked up by the move from 68k to PPC. Have some kind of chip that does PPC emulation, as well as software emulation.
And not every developer will have both machines. A lot of things were developed for just x86/PPC BeOS and not the other, and the same problem would arise here.
GCC solves that problem. You could compile for PPC on a x86 machine. And vice versa, I believe. Plus, this isn’t the same as BeOS which continues to have the PPC version. This is like the move from 68k to PPC.
and alll apple needs to do is anounce there intentions and that will get the companies moving.
They would have to stop production on all current PPC machines and sell them all out before making the announcement, because after the announcement, nobody would be stupid enough to buy a PPC machine.
Last time Jobs said: “he likes to have options… …at the moment there is no need to think about x86
When Jobs doesn’t give a resounding no, it means there is a huge possiblity. Just like PDAs, Jobs say never, and viola, it never was release. For Windows iPod, Jobs said maybe, and viola, what do we have…
And he was also responding to a question about x86. Motorola’s roadmap looks neat, though complete devoid of any information of the processor, except the most basic (like 64 bit, 32-bit support, blah blah).
Honestly, if Apple moved to x86, they could potentially sell as many computers in ONE YEAR as they have sold Macs in TWENTY YEARS.
Sorry, I don’t see how. Please, the move won’t result Apple updating their hardware every week, or pricing it competitively with Dell and home-made PCs…
They are gonna be overrun with calls to support this and that.
Seriously, I wonder why. OS/2 nor Be ever complained about this. Also, a lot of people, my mum especially, regards the Mac as a brand of PCs.
My iMac tends to a lot of things pretty well, where do you actually see performance problems on a Mac?
Well, everything I do on the PC is faster than an iMac 800mhz (top of the line), though I haven’t tried compiling apps. Especially surfing the net. I haven’t tried Linux on Apple’s new machines though…
Tied down to a legacy architecture, if they wanted to go 64-bit, might as well go fresh.
AMD being tied down to a legacy architecture doesn’t mean Mac OS X needs to be tied down to the legacy part of it. The legacy stuff are there for backwards compatiblity. Besides, big deal, PPC is just 2 decades younger than x86. Also, Hammer is bound to be the cheapest 64-bit processor out there, would be well suitable for Apple’s kind of products.
AMD or Intel? And again, why not go for something new?
Well, going for something new sounds stupid for a company like Apple. Apple isn’t close in having the amount of R&D money as Intel nor AMD, so they are better off just sticking to PPC than making something new.
if the ROM chip has all the boot strap instructions for the OS in it, then it will take seriouse hacking to get it to work on a normal PC…
People want to use Windows and Mac OS on the same machine. An x86 machine would do that fine. Apple would never do anything to block Windows, or even Linux and BSD on running on new x86 Macs, but they would block OS X being install on machines other than from Apple.
Once OS X is ported over to Intel/AMD, moving apps over will be simple.
The OS may be portable, but a lot of apps aren’t.
Apple hardware is superior to most PC manufacturers (perhaps with the exception of Compaq Computers)
And IBM. If you want quality PCs, while you don’t mind spending a bit more on it, plus internation support and warranty, well, I would recommend IBM. The most sturdy built PCs out there. There is even a high chance of them beating Apple in terms of quality.
The original Mac was a conscious decision not to use x86, as were the NeXT workstations.
What was the reason NeXT workstations being pulled off the market? Yeah, being extremely expensive.
As hard as I try, I cannot envision an x86 Mac in any form.
Try imagining a Mac. Yup, that’s how it’s gonna be, no different in looks, in style. Most likely in performances and perhaps price. Try not to think of a beige box with an Apple logo, but a Mac with a Intel Inside sticker.
Switching to a different non-x86 platform; MIPS and POWER-derivative being two popular suggestions
I’m not sure about MIPS, but stripping down POWER to a usable state in current Macs would erode any increase in performance over PPC.
Increased reliance on dedicated hardware coprocessors
And instead of cutting cost for the same thing?
Multiprocessor systems across the board, with quad on the high end
Most consumer apps won’t be able to take advantage of it, and most of them would probably never will take advantage of it.
IBM/Apple/Motorola have way too much invested in that platform to just dump it.
Apple had invested a lot, but they also invested a lot on 68k processors, why did they dump them? Besides, IBM and Motorola are getting less and less interested in the market, Motorola is not making much money out of it, IBM is being snubbed by Apple…
#1: Macs are easy to spot. They are the ones with attractive cases and usually carrying the nifty new features. Ex.: USB, Firewire, Desktop LCD monitors, etc…
Hmmm, with the exception of Firewire (Apple invented it for crying out loud!), USB and all-in-one LCD computers came out way before Apple’s stuff.
Perhaps because the PC OEMs ( Dell, Compaq/HP, Sony,etc..)
just won’t preinstall BeOS nor any other non-M$
operating systsm on their systems.
Uhmmm, Dell had Linux desktops for a very long time till they killed the consumer Linux stuff because of lack of demand. My opinon way be alien to you, but for me, bundling two different OSes on the machine is like bundling two different processors from Intel and AMD.
What make you think that if Apple does ship their
OS X to x86 PCs, the PC OEMs will ship the OS X
Most sane people are suggesting the Macs you see now would come with a x86 processor and not a PPC. No change with Apple’s anti-clone stance.
What has been keeping Apple alive is their innovation and uniqueness, not rolling out tons of beige boxes.
Who ever said Apple would have beige boxes when they move to x86?
A 1GHz Alpha still can compete well with the newest x86 chips
But stripping it down to be close in being financially viable to be in any Mac, Alpha’s good performance would be gone.
would programs be able to run straight on osX/x86 and osX/ppc or would they need a simple recompile, complete rewrite, small rewrite?
If the 68k to PPC transition is the same as the PPC to x86 transition, you could run your PPC apps. As for making it native, it depends on the app itself.
What percentage of the pc market does Dell have? What percentage do other manufactuerers have? Because that’s the point.
Dell has about 20%. Together, branded takes close to 50%. The rest are small time companies. I don’t see your point though.
BeOS showed that porting, even if you stick to the API isn’t always going ot be easy.
This is because BeOS’s PPC and x86 version used completely different compilers. But right now, Apple is pushing GCC, which means if the app isn’t using CodeWarrior but GCC, there shouldn’t be a problem.
When you compare….
It actually depends on the brand and model. Some just increase the clock speed like early P4s, but didn’t increase the performance, while others are different…
(I say reasonable because macs already command a premium) then Itanium looks promising to me.
Not in the next 3 years then. Financially, forget it. Technically, no good compilers available as of now.
Apple could make quite a splash being the first/best manufacturer to have a Pentium 5 machine available and running the next generation of Mac OS X. Would be very cool.
G3 to G4 to P5, very good for ads if you ask me. Power Mac P5…..
… That is, until somebody hacks it.
No one would be able to sell it before being bankrupted because of paying for legal fees. Maybe Mac4Linux… but…
I know that Intel likes to a) tweak msft’s nose when possible
They do? So all the rivaly for Microsoft love between Intel and AMD is fake?
Pro mac users will be taken care of.
1.2ghz G4 doesn’t sound that inviting, since 800 to 1000mhz wasn’t that fantastic, besides finally being able to say “GHz” when you tell the processor frequency.
Okay. Here is were you get dumb. I don’t see any modern graphics hardware that doesn’t fit in my Mac. The problem is drivers. Changing to x86 isn’t going to magically write drivers…
Because of OpenFirmware, stuff like Quaddro4 on a Mac isn’t possible. In fact, IIRC, GeForces and Radeons on the Mac differs from the PC counterpart…
They would sit still until either Hammer or Itanium emerged the clear victor.
Why? To see why wins in the server market to find the next processor for Xserve? Puh-leze, they aren’t in the same market.
…people certainly can compare a 2Ghz Windows Machine for about $1000 to a 2Ghz Mac for $2500.
And they don’t compare a 2ghz Windows machine at $1000 and a 1ghz machine at $2,500 now?
With hardware you can pick two: good, fast, cheap. Apple tends to pick the first two.
Perhaps that is true back then when PPCs were better than x86s. But since Apple fucked it, caused IBM and Microsoft to leave the PPC desktop market, times have changed. PPC isn’t as good nor as fast as their PC counterparts.
…are also the same people constantly crying over crashed drives
I wonder why actually when Apple move to x86 they would use cheap hard drives than the hard drives they are using now
I know Apple is the greediest damn computer company around…
I didn’t agree with the way Apple moved to .Mac (e.g. not charging for only the extra tools and let current users keep the free @mac.com address), but really, Apple isn’t the only one doing it. It cost a lot of money making iTools free.
And maybe another $0.25 for another mouse button?
if I buy a branded machine, it would most likely come with a 2 button mouse, I would still go out and buy a better mouse. Come on, after paying big bucks for your mac, you can’t even fork $10-15 for a optical Logitech scroll mouse?
(PS: Sorry for the 3 post long reply, but come on, this thread errupted into a 70 post thread before I could even come online… I tried to trim down the post)
Who does Apple sell computers to? The people who read OSNews and worry about absolute speed are probably not a big part of Apples Market, if they were don’t you think they would have switched years ago? No I think Apples major market is elsewhere.
I can’t see any major advantage for switching, it would only make Macs look bad compared to much cheaper PCs but now they would be directly comparable and this if anything will cut sales not improve them.
Some of their markets do need performance (Audio, Video etc) and this is where Altivec can do a good job – they really do need a faster bus interface here but this is coming along with an improved compiler.
I think Apples speed problems are at most temporary and this is the only reason people are speculating on any change. Motorola and IBM have some interesting stuff in the works (or shipping) which if utilised in a Mac could imporve performance – things like hardware accelleration of network stack.
If anything I think a move to PowerX might be on the cards but even this would just be at the top end with G4s continuing lower down.
Cheap hardware, great OS (but not exceptional), nice design but very expensive. They can continue that way if their customer target are rich blinded people…
If they have decided to swich from a cpu to an other, they need to something different from the others. This is very important for them and for us. They need a completely new architecture and i suggest the Elbrus E2K arch wich is the most advanced ever. Elbrus is eeking for inversors and apple is dreaming about brute force CPU. Apple should buyout Elbrus MCST.
http://www.elbrus.ru
OS X on the x86 would first of all be a marketing disaster.
Apple has bashed pentiums for years, and you would then have no reason to buy apple hardware if they switched to x86. Moreover, apple hardware would be easily comparable to PC hardware so you’d have no fudge factor, as you do with the powerpc, to claim a superior technology and justify higher costs. But you know, apple does not have to use an x86 to buy a chip from intel. Intel needs volumes to pay for those new fabs. They might be willing to fab the power pc for apple.
2) Apple would loose support of its development community. What you want us to do more work and now port OS X apps to x86? No way. There will be no applications on x86.
3) OS X’s adoption will coincide with sales of new computers as is the case with XP. New computer sales are a lot slower than they once were and apple, having a higher price point, is particularly exposed.
4) Andrew Neff makes plenty of other outlandish forecasts that never come true. the fact that he got one thing right does not mean he has some magical crystal ball.
5)Intel’s #1 customer for x86 and hence for everything is effecitively microsoft. Microsoft will not sit around and do nothing if apple is on an x86. Beleive me, MS will retaliate against apple and against intel.
6) There are other things that apple can do. Apple does not need to jump to an x86 to satisfy its performance requirements. That is a complete lie and a monumental marketing myth.
7) 99.00 % of all computer users stoped caring about processor speed when processors speeds hit 700 or 800 Mhz.
The MHz gap is important to people in this group and not necessarily to the mainstream community.
8) apple’s biggest problem is not its MHz gap. They are failing to gain share because of their high price points. Lower the price point and i’d bet half of the people that complain about their MHz rating in this group would run out to buy one.
Nonetheless four years is a relatively long time. By then the adoption of OS X should be complete. The itanium might even be a good choice for desktops by then so it is understandable why jobs would say “we like to have options.” But i wouldn’t hold my breathe for a switch in the next 24 months. won’t happen.
I doubt x86 is unbeatable at it’s price/performance point. That would make it PERFECT (and i don’t mean just for apple). But it’s not, it’s got baggage like all hell and i’m sure that it’s performance can be improved other than by just ramping the clock. The really cool thing here is that we can do a clean architecture break. The windows world can not do that, so it tries to the best with what it currently has. Frankly we can do almost anything now (barring previously mentioned limitations like power consumpion and such) and I think apple should do a little reasearch and look at something that is not “dirty”, is faster than PPC right now, and has good prospects for future speed enhancements. That is the key.
Um laurent, most of us can’t read russian so would you mind giving us a rundown of the achitecture? If it really is wonderfully done, then it’s an option for apple. And like Steve said, options are good.
Apple is a hardware AND software company. If they were to release OS X for generic PCs, they would instantly become primarily a software company. MS would kill them.
What is more likely to happen is that Apple will go to a proprietary solution using the X86-64 architecture. Use a proprietary motherboard, bootROM, and possibly a special version of the Hammer chip and you have a version that has a lot of the advantages of the PC platform that Apple can still control completely. Furthermore, I would imagine that they would still want to control how peripherals and intenral cards interact with the machine as they do now.
Having control over the OS and hardware is something Apple does. It has advantages and disadvantages, but it’s part of the culture, I don’t see that ending even if they do go X86 or some variation.
http://www.elbrus.ru/mcst_e/proect/e2k_arch.htm
i remain stunned at the success of intel’s marketing campaign.
they made a very conscious decision to ramp up the MHz/GHz at the cost of all else. In the process they have created pavlovian response amongst PC users to an increase in processor speed.
But the year is not 1995. It is 2002. I wonder why all of the apple critics that bash apple’s performance constantly never turn their microscopes on themselves. Maybe a look at Intel’s little GHz marketing campaign might be suitable because it is full of BS as well. Remember moores second law. The economic benefits of proceeding with moore’s first law diminish with time. Apple’s super computer on a chip claim is by no means the only BS in the computing world.
they’ve always been a hardware company and they wouldnt put themselves in a position of supporting so much with their software. at the moment they support G3, G4 processors, dont tell me they’ll support x86-64, pentium, itanium or anything else.
if apple switch to a different processor its likely to be a pumped XScale…
For sure a Mac is a PC with different GLUE chipsets and CPUS.
Anyway we have to remember that Apple is deeply involved in the CPU design so they know it very very well.
Apple is indeed the designer of the chipset.
Going to X86 … using third party’s chipsets maybe? What do you think? wonna make apple a regular PC brand.
what keeps apple going if the stand-alone approach to the market: I use a really good CPU, I build my chipsets, I put good technology first (firewire DVD burners).
Apple is hardware and software… remember please. if they switch to x86 it will be a PeeCee .. and maybe you want also the MacOS X working on standard PCs ???
Apple will die in 3 years, they have no power to go against dell .. they produce wordefull designs .. custom desgin .. they build the turnkey solution down including the firmwares too .. they use no cheap components …
They spent ages optimizing for altivec and developers too .. what do you want a PeeCee with the name apple on top?
They need to upgrade technology …. go Apple for DDR, go for quad PowerPCs .. a new PowerPC reaching top speeds .. it’s two years we are waiting for it .. if they miss it this time too .. probably you are all right .. they will go for AMD (no intel, no reason to do it).
To eugenia:
if they move to x86 going to x86 32 bits will be a suicide!!! if they have to do it they need to make developers and users lotta lotta happy := 64 bits
…why i need a 3 GHz Processor to make the same things I can make today with 1 GHz?
… the same reason that you needed the 1Ghz to make the same things that you made yesterday with 400Mhz.
#m
I really hope Apple does port to x86. This will open the market up to a really large selection of processors with some insane speeds at some cheap prices… I have a few points that I think people are missing here :
Hardware
You have to remember that they won’t port to just plain old x86 that will run on the latest Abit etc. motherboard. They will create their own workstation that uses a x86 processor – thats all. You will still have to buy a computer to run OS X from Apple, it doesn’t matter if its a PPC or an x86 – it will still look the same (well I am sure they will have some “twist” to make them noticable), still come with the same bundle of manuals/software/cables/CD’s.
You won’t be able to buy OS X and install it on your overclocked Celeron’s at home either – Apple would loose far too much money if they let that happen. (Although I personally would love this)
Software
A very large section of the graphics and sound software market has just been bought by Apple (Shake etc.), why not offer all their apps optimized for x86 and see who follows…. People will buy the x86 workstations just for faster speeds on a single application in this market.
I am also sure that Apple will hastle most software houses to offer their applications on x86 too, if not help with programmers, help by paying for some of the major ones such as Photoshop.
Why bother?
Well, we all know about how the G4 chips are starting to lag behind the x86 ones. It doesn’t matter about benchmarks, you have a top of the range Apple running the latest OS X and a top of the range PC running WindowsXP and guess which one feels more responsive – the PC. OS X *needs* more CPU speeds, it will get some from the GPU’s soon, but it still needs more IMO.
It could drop the prices of Apple hardware a little if they used a *real* mass-market CPU such as AMD or Intel.
It will make it easier for people like Adobe to make CPU specific optimisations for their cross platform applications.
It will make it easier to hack your Apple box to run OpenBeOS (or whatever its called in a few years time), Linux, FreeBSD, OpenBSD, all the commercial software for Linux is normally on x86, it will open your Apple-workstation up to running this also. Yes, I know that you can get YelloDog etc. for PPC but lots of the software isn’t as upto date as it could be, and if your not up to compiling your own, then the distro is what you use.
Just my 2 euro-cents.
Andrew Edward McCall
Your analogy comparing Mac to BeOS is flawed:
1) Be never got any traction. It never gathered more than an extremely small cult following, even though it was a superior OS.
2) Apple, with a ~5% market share, is a major player, and has massive brand recognition. Anywhere it goes, the faithful will follow and…
3) other people will try a new OS if they’ve heard of it before.
4) Microsoft no longer has an iron grip on its hardware vendors anymore, thanks to the Department of Justice. If Microsoft tries to coerce the hardware vendors, a few well placed phone calls will have SWAT teams of lawyers and regulators up their butt faster than you can say ‘unlawful competition’.
Where Apple should (and most likely would) do is to make the x86 box proprietary; having grey boxes run MacOS would be suicide. Imagine thousands of newbies buying grey boxes, and finding out that half of their peripherals aren’t supported.
I would absolutely love to run the MacOS on my PC! I’ve got OSX on my 500Mhz iMac and I would love to get it on my regular desktop!!!
Bring it on Mr. Jobs!
…a new PowerPC CPU from IBM based around the POWER4 architecture, to replace the current POWER2-based chips Apple’s stuck with from Motorola. Apple has too much vested marketing interest in the PowerPC line, and they’ve always taken backward binary compatibility into account with their updates. I’m not convinced that software emulation of a G4 on even current Intel/AMD chips would be sufficient; for a technical discussion of why, see http://www.emaculation.com/ppc.shtml — I agree with the author that actually using a separate, real PPC for emulation is probably the only way to do this acceptably. And let’s face it, if people are complaining about the cost of Apples now, this isn’t the way to go if it can be avoided.
they would just change out the proc in there box, they would not let Dell et al install OS X on a normal PC.
I don’t see how developers would be pissed off at Apple for going x86. Unless commercial apps were written in Assembly or used some hand coded Assembly, it really doesn’t matter what CPU the Mac platform moves to. Apple will be facing the big task of porting OS code and libraries to x86 instructions and if they’re written in C/C++/Obj-C Apple just needs to recompile and debug and modify some code. The 3rd party developers just need to get a development computer and recompile and test their apps. Most of the code in an application is UI code or makes calls into the OS services, and that is up to Apple to work out, so 80% of your app’s code is already “ported”.
I have said that multiple times on this board, but no one seems to read.
ckristian
Anyway we have to remember that Apple is deeply involved in the CPU design so they know it very very well.
No. Apple just takes it from Motorola and do their marketing and advertising for Motorola. What Mot calls AltiVec, Apple renames “The Velocity Engine.” Do you really think Apple employs CPU engineers?
Apple is indeed the designer of the chipset.
Again no Apple is not. Perhaps they designed the motherboard schematics so that it can fit in that circular base of the flat panel iMac, but Mot makes the CPUs and mobo chipsets.
The only thing that makes Apple a hardware company is that they design and make their own cases, invented firewire, oh of course they sell and distribute Mac systems.
Apple is hardware and software… remember please. if they switch to x86 it will be a PeeCee .. and maybe you want also the MacOS X working on standard PCs ???
How is the Mac not like a “PeeCee” right now? A Mac and PC both have CPUs (although different) both use SDRAM, both have PCI and AGP ports – although the Mac has fewer legacy ports. Both have operating systems and software. From a generalized point of view, a Mac and PC are one and the same – they’re both just computers, they do the same exact function just differently. Just because Apple puts it in a pretty gift wrap and use a proprietary hardware design, does not make it fundamentally different from PCs. The components are the same just arranged differently.
Apple will die in 3 years, they have no power to go against dell .. they produce wordefull designs .. custom desgin .. they build the turnkey solution down including the firmwares too .. they use no cheap components …
No cheap components? TheRegister ran an article a while back saying how lousy the internal modems are on Macs. The writer could not get a decent connection via dial-up and maintain it. I am a witness to this because my iMac flat panel can not hold a connection longer than 5 minutes since the day I bought it 4 months ago. How about the LCD monitors that get dead pixels and Apple basically tells people T.S.
Oh yeah, it’s great to have a computer that looks pretty and attractive. Gives me something to do while I wait for the window to resize.
They spent ages optimizing for altivec and developers too .. what do you want a PeeCee with the name apple on top?
AltiVec only came out a few years ago, I don’t see how that is considered “ages.” And AltiVec really isn’t as great as Apple wants people to think. AltiVec can’t accelerate double precision floating point computations, the Pentium 4’s SSE2 can. How is AltiVec useful to the scientific community where accuracy is more important than performance?
To eugenia:
if they move to x86 going to x86 32 bits will be a suicide!!! if they have to do it they need to make developers and users lotta lotta happy := 64 bits
There is no concrete evidence that 32 bit x86 is bad for Apple. Many people before me have said so: 32 bit x86 is so deeply entrenched that it has many years of life left in it. The Mac – in my eyes – is more of a consumer computer for the simple minded people, who just want to surf the web and check email. They have no need for 64 bit computing Geeks will be thrilled with having a 64 bit platform, but there are so many other areas that need improving and if they do get improved upon, Apple would have satisfied those geeks just the same.
Developers aim at what is here and now. We do make an effort to be prepared for the future, but it would be wrong to bank on something that’s not out yet. 64 bit computing isn’t going to make developers happy. How about better FRIGGIN’ documentation on the CarbonLib??? The API documentation is horrible – “This is function XYZ. Here are the parameters, but no we’re not going to tell you what they mean. Oh but it is supported in CarbonLib version 1.whatever.”
Just adding to the earlier posts on using the graphic co-processor.
The July issue of Wired had an interesting article that talks about Nvidia’s effort to compete with Intel. Nvidia CEO Jen-Hsun Huang says “The Xbox is how the computer will be built in the next 20 years. More semiconductor capacity will go to the user experience, The microprocessor will be dedicated to other things like artificial intelligence. That trend is helpful to us. It’s a trend that’s inevitable.”
more at
http://www.wired.com/wired/archive/10.07/Nvidia.html
TLy, I am certanly not opposed to Apple ultimately changing to x86 or 64 bit processors. And, I agree, developers would not be in some terrible bind. And, I also *really* agree that OS X, as it is now, is simply too slow.
However, a couple of things. There are, in fact, Apple people working at/with Motorola. I don’t know the extent of their work – it may be just as you suggested – working on the specs for motherboards, etc., I don’t know. Also, I really think, although it is hard to generalize about this, that it is the Dell/GateWay PC buyers that are the web surfers and emailers. Even consumer Mac buyers are much more likely to be making home movies, digital photo albums and that type of thing.
Anyway, I want to see 10.2 and the new Power Mac upgrades. That will tell all of us a lot, to be sure.
BTW this article sounds like, well, is a troll.
Nice job.
Speaking of …
Insanely great:
Runs on,
PPC (IBM, Motorola, whatever)
x86 (Intel, AMD, whatever)
Toaster ovens (embeded, whatever)
Developers would have the “Option”, therefor a labor of love, profit, whatever.
Options, Kids
Jobs is not going to shoot himself in the foot. Duh.
See: BeOS
New Keword: Whatever.
🙂
what you just posted made no sence to me.
Eugenia:
Maybe, but Opera and iCab is faster than all the other browsers under OSX. And I am not even an Opera or an iCab fan… 😉
Opera is one of the slowest on the Mac, Mac World did a test on it and it came out dead last, and with javascript it took 50sec. longer to load than Mozilla (which is slow)
More than 50 seconds for a browser to load? Whoa.