Why hasn't Apple switched to x86?

[Rivergater]

The answer to the question is so simple:

Apples hasn't switched to x86 because they don't want to compete directly with Microsoft. Which company in their right mind would?
If you take a look at a basic economic text, you'll read about how in monopolistic competition (which is basically the form this industry takes), firms will try to compete by differentiating their product so they don't have to compete through price.

By sticking with the technology they've been using, its a form of product differentiation. Remember also when Apple licensed its technology out to other companies, they almost died. The commoditization of their hardware was a terrible strategy for them at that time (maybe if they tried it at a different stage of the business, it would've worked). It was only after Jobs came in and bought back those licenses did Apple reemerge as a strong company (there were other facts in this as well but I won't list).

Apple doesn't want to be in the business of producing commodities. They don't want to make a computer that you can compare to PCs. They are doing well with this, because even in this thread, there is a consensus (or debate I should say) that its an apples to oranges comparison. They want to make an electronic product that cannot be compared to PCs and that have other elements of value besides just being able to crunch data, calculate, etc. They want their products to be more like a home appliance or a personal appliance. Its a great strategy for them since they are a niche player.



EDIT: grammar.

 

[ZaneNBK]

Originally posted by: Beau6183
Then why don't you through some numbers and specs out to put some proof behind your claims?

Sure, lots of comparisons on the web:

CPU Scorecard - G4 1ghz 13th down on that list. Tied with Xeon 2ghz and Athlon XP 1800+.

 

[FeathersMcGraw]

Originally posted by: ZaneNBK
What's the purpose of sticking with PowerPC? Most of the hardware in Apples is the same hardware used in x86 machine. Their stubborn refusal to switch platforms is costing them SOOO much money, both in increased costs and decreased sales.

Why would Apple abandon their niche market and compete with a bunch of PC manufacturers who are coming to the realization that desktop PCs are commoditized and that the margins on systems are growing thinner every year, even if the current economic conditions hadn't cut IT spending pretty much across the board?

 

[Random Variable]

The PowerPC may be a "good" chip

Don't forget that the IBM Power4 is fundamentally a PowerPC chip.

 

[kgraeme]

Originally posted by: Lucky
edit: anyone know if a USB keyboard designed for PC's will work on a mac? They shop said they use different connections for a mac.

Yes, you can use a PC USB keyboard on a Mac. I Have used both a Dell USB keyboard and a MS Natural Keyboard Pro.

 

[ElFenix]

Originally posted by: hoihtah
i think you're missing the point here.

do some search on pro's and con's of risc and cisc processors.

and tell me that if apple had started with cisc... they'd be where they are right now.



risc vs cisc
scsi vs ide
firewire vs usb
lcd vs crt

do you see a pattern here?

apple did start with cisc for the macintosh. the motorola 68000 was in the first mac, and they used up to the '040. which was slightly faster than a 486, IIRC.

 

[Lucky]

Originally posted by: kgraeme

Originally posted by: Lucky
edit: anyone know if a USB keyboard designed for PC's will work on a mac? They shop said they use different connections for a mac.

Yes, you can use a PC USB keyboard on a Mac. I Have used both a Dell USB keyboard and a MS Natural Keyboard Pro.

wtf, why would they tell me differently. they had a PC USB keyboard for $26 and they wanted $65 for the mac.

 

[bluemax]

apple did start with cisc for the macintosh. the motorola 68000 was in the first mac, and they used up to the '040. which was slightly faster than a 486, IIRC.

No way! The 68040 was a smeggin' awesome processor that would handsomely smoke a 486 and most Pentiums! (I think the line is right around P2 or so...)

The 68000 archetecture is just inherently better overall. The 80x86 archetecture from the ground up is flawed. I forget a lot of the details, but things like data paths, memory limitations, a whole whack of stuff really holds back the x86 processor platform.
All the flaws of the 8086/8088 are still present today for "compatibility" reasons.

I wish Intel and Microsoft would just make a Quantum Leap(tm) and DUMP all the old Windows code and x86 chip "compatibility" and start over from scratch. But most people would whine they can't use their 10 year-old software if they did.
I've been saying it for a few years now - a total "do-over" with a dual-boot option right out of the box would do Windows wonders!


Maybe what we need is a new platform. New chip/archetecture, new OS, but it'd have to backed by a HUGE company to have the push to make it popular enough to have people abandon the old. Not gonna' happen.

 

[imgod2u]

There are, of course, many advantages to having a closed-form platform. You control everything and you don't need approval by the rest of the industry in order to switch platforms. This, I think, is the main reason Apple so stubbornly sticks with the PowerPC. Athough I do think this is kind of pointless. Just because they would switch to an x86 processor (from Intel or AMD) doesn't mean they'd have to support everything out there. They could still claim stuff doesn't "work on a Mac" and people would still buy it. Compatibility will only improve when moving to an x86 platform, not decrease, so the current Mac fanbase will stick with it (because it's so perty....) and they may even gain some enthusiaste market share. Of course, they'd loose control of where their OS will be installed on. But then again, they could choose not to sell their OSX commercially.

Oh, and as far as the processor speed. There is a definite way of comparison. Take software that runs on both platforms, and compare how well they perform. Easy enough isn't it? This "they're different platforms and you can't compare" BS is nothing more than some excuse Mac fanatics spit out because they realized their beloved Macs would perform poorly compared to x86 platforms in software that is available on both (with a few exceptions).

Performance of software can be attributed to many things, not just processor type. Besides, photoshop runs about the same on my 1Ghz G4 as it does on my 1.8 athlon (work machines), so wouldn't that indicate that the Mac processor does it better because it offers the same performance at a lower clock speed?

This same arguement again and again. This is almost as rediculous (what am I kidding, it is as rediculous) as the "MHz is more power" arguement. How much it can do per clock means absolutely nothing if the chip can only scale to 1/2 of the clockspeed.

Say what you will about RISC being superior and all that, but the bottom line is, both AMD and Intel, the two major MPU designers in the world, are sticking with x86 for the time being for the desktop platform. While Motorola may have a processor supporting a better ISA, it does not have the dedicated resources or army of engineers that Intel has nor the devotion to the CPU market like AMD has. Apple can market all they want and throw around words like "we're 128-bit" or "we're RISC" but the bottom line is, there is no way Motorola can keep pumping out chips like Intel or AMD. When those companies decide to switch away from the x86 platform, that's when it will happen. And there's not a thing Motorola or Apple can do about it.

 

[RemyCanad]

I would just like to state that this is one pointless topic. Why hasn't apple switched to x86? Why would they? I don't see one reason. If they did they would loose money for sure. Think of converting all that software, then still supporting the legacy hardware. And the x86 is gonna be dead soon enough. (At least in my mind. thats a hole nother topic) The G4 is a 128 bit processor. AMD and Intel are just now getting to there 64 bit processors.

Just a week ago I contacted apple and asked if I could buy a seperate parts and they said no. Hmm I don't know what to think of this. Do you think they would be making any more money or would people just start competing with them...

Finally I would like to state my dual 1GHz G4 is much faster than my AMD XP 2000+.... So switching for speed is not there.

 

[imgod2u]

Originally posted by: RemyCanad
I would just like to state that this is one pointless topic. Why hasn't apple switched to x86? Why would they? I don't see one reason. If they did they would loose money for sure. Think of converting all that software, then still supporting the legacy hardware.

Not quite. Just because your software supports x86 doesn't mean it has to support everything else. Look at Linux. It runs perfectly fine on the x86 platform yet it doesn't have to support every piece of legacy hardware out there. Apple designs their own computers and has the power to make their computers legacy-free and yet still based on an x86 processor. This "it has to support everything" crap is a myth. The processor does not make the whole computer. Just because Apple has their software running on THEIR x86-based computer doesn't mean the software HAS to work on all OTHER x86-based computers.

And the x86 is gonna be dead soon enough. (At least in my mind. thats a hole nother topic) The G4 is a 128 bit processor. AMD and Intel are just now getting to there 64 bit processors.

With Hammer and the P7 core scheduled to scale all the way to 10+ GHz. I doubt x86 is going to be dead anytime soon. And the G4 is not a 128-bit processor. That's some stupid statement some idiot mac fanatic prolly came up with after reading about the AltiVec extension in the G4's (which are 128-bit). The processor itself, is still 32-bit just like the P4/Athlon. Btw, just in case you feel like blowing AltiVec out of proportion. Intel's SSE2 extension is also 128-bit. AltiVec is Motorola's version of SIMD. It actually works pretty well for the few software that uses it. But it does not make it a 128-bit processor anymore than Intel's 128-bit SSE2 extensions make the P4 a 128-bit processor.

Just a week ago I contacted apple and asked if I could buy a seperate parts and they said no. Hmm I don't know what to think of this. Do you think they would be making any more money or would people just start competing with them...

Finally I would like to state my dual 1GHz G4 is much faster than my AMD XP 2000+.... So switching for speed is not there.

I'd like to see some benchmarks on that. Not to mention the cost of that dual 1 GHz compared to even a Dual MP 1.67 GHz. It's so entertaining how people like to just make unsupported statements like "mine is much faster" and support it by "it feels faster". When I'm bored it "feels" like time is getting slower, doesn't mean it is. In other words, the way you "feel" is by no means a valid indication of speed because it is far too relative to your mood, and most importantly in this case, biased views.

 

[Random Variable]

The G4 is a 128 bit processor

On what planet?

 

[n0cmonkey]

I would switch all of my systems over to ppc platform if it was cheaper. Its a wonderful platform and a breath of fresh air when you are stuck with the crap known as x86.

 

[n0cmonkey]

Stop comparing Apple to Intel. They do totally different things. Compare Motorola to Intel. Thats almost a fair comparison (Intel almost has enough products to compare to some of Motorola's). Compare Apple to Dell. Why dont Dell's machines perform as well as home built machines using comparable hardware?

 

[ViRGE]

Originally posted by: Vespasian

The G4 is a 128 bit processor

On what planet?

Not on this one I'm afraid. The G4 does however have 128bit registers, which is something even the P4 can't claim. These registers are why the G4 absolutely flys on the Distributed.net DC project "RC5-64". The larger registers open up new, faster doors to go about things. I have a box at work for example that can crunch RC5 at 17.4MegaKeys/second. This is a dual 1ghz machine, which totals 2ghz worth of power. A 2ghz Athlon(the fastest x86 chip in terms of MegaKeys/mhz), on the other hand, can do only 6.8MegaKeys/second(even if you went dual, you wouldn't beat the dual G4), and the fastest P4, at 2.53ghz, does 3.4MegaKeys/second. While the G4 is overbuilt when it comes to registers, those little registers can go a long ways in some cases.

 

[krazykilluh]

A little late, but I think there are basically two reasons why Apple doesn't go x86:

1) The x86 instruction set is crap. No wait, it's worse than that: It's crap built on top of crap that's built on top of crap. Take a look at the x86 assembly code that gcc generates...it can't be understood by mere mortals. Just be thankful you don't have the job of pipelining this junk...

2) Apple can't compete with Dell, Gateway, HP, etc. Macs sell for more than PCs today, but I bet Apple doesn't pay too much more for the components than Dell does. Apple can sell Macs at a higher profit margin because there are no competetors selling Macs. If you want to run Apple's software then you have to buy Apple's hardware, even if it is overpriced. Let's say Apple went x86 and you want to run the x86 version of OSX or whatever. Are you going to pay $1000 for the Dell box or $1500 for the Apple box that's the exact same except it comes in a fancy case??

Of course, if Apple went x86 I guarantee they would their software business would grow. OSX is awesome, I'd run it if I didn't have to pay $1300 for a Mac. But their hardware business would suffer.

 

[imgod2u]

Originally posted by: ViRGE

Originally posted by: Vespasian

The G4 is a 128 bit processor

On what planet?

Not on this one I'm afraid. The G4 does however have 128bit registers, which is something even the P4 can't claim. These registers are why the G4 absolutely flys on the Distributed.net DC project "RC5-64". The larger registers open up new, faster doors to go about things. I have a box at work for example that can crunch RC5 at 17.4MegaKeys/second. This is a dual 1ghz machine, which totals 2ghz worth of power. A 2ghz Athlon(the fastest x86 chip in terms of MegaKeys/mhz), on the other hand, can do only 6.8MegaKeys/second(even if you went dual, you wouldn't beat the dual G4), and the fastest P4, at 2.53ghz, does 3.4MegaKeys/second. While the G4 is overbuilt when it comes to registers, those little registers can go a long ways in some cases.

Not really. As I stated in a previous post, the SSE/SSE2 component of the P4 is 128-bit. It has a register buffer that is able to contain clusters of data (multiple FP or Integer data that combine to 128-bit) and schedule it for execution. The G4's AltiVec component is actually very similar to that of the P4. The reason the G4 is good in distributed computing, especially RC5 is due to its RISC-like nature. Without the limitations of x86 such as only 3 instruction decodes per clock and a massive integer execution engine, it can achieve much better results when performing pure integer math.

 

[TNTrulez]

All your points are valid. Email them to apple.

 

[kgraeme]

To answer the initial question, they have. They just haven't told anybody yet.

 

[BD231]

Isn't RISC processing more efficient than the x86 architecture for complex operations? I thought that was the sole reason scientists used them

 

[RemyCanad]

Originally posted by: imgod2u
With Hammer and the P7 core scheduled to scale all the way to 10+ GHz. I doubt x86 is going to be dead anytime soon. And the G4 is not a 128-bit processor. That's some stupid statement some idiot mac fanatic prolly came up with after reading about the AltiVec extension in the G4's (which are 128-bit). The processor itself, is still 32-bit just like the P4/Athlon. Btw, just in case you feel like blowing AltiVec out of proportion. Intel's SSE2 extension is also 128-bit. AltiVec is Motorola's version of SIMD. It actually works pretty well for the few software that uses it. But it does not make it a 128-bit processor anymore than Intel's 128-bit SSE2 extensions make the P4 a 128-bit processor.

(After pulling my foot out of my mouth)
Sorry, I must say I was completely wrong. Now I must find the site that I read this information on and mark it as being unreliable. If I remeber correctly it was actually a PC site.

Once again sorry about that.

 

[n0cmonkey]

Originally posted by: TNTrulez
All your points are valid. Email them to apple.

Which ones did you consider valid? Every arguement he brought up was just rehashed from past discussions and was barely valuable.

Apple would have to charge just as much for x86 machines because they would have to redesign a good portion of the hardware to make it so you cant just purchase a system off the shelf, a copy of X, and throw it on that machine.

 

[zephyrprime]

and tell me that if apple had started with cisc... they'd be where they are right now.

Apple DID start with CISC. Remember the 68000 series of processors? That's what macs used before the powerpc era.

 

[BFG10K]

Three main reasons:

(1) Steve Jobs is a retard.
(2) Apple can continue charge black market prices for PPC processors.
(3) The RDF coming from Cupertino has convinced Mac users that the dual 1 GHz G4e system is much faster than any x86 system, which of course is utter garbage. In fact the Palomino is almost clock for clock competitive with a G4e.

 

[imgod2u]

Originally posted by: bdog231
Isn't RISC processing more efficient than the x86 architecture for complex operations? I thought that was the sole reason scientists used them

I don't think there is such a thing as "complex" instructions in the RISC forms of ISA. That's the whole point. All instructions are of equal length (in the case of the PowerPC, this is 32-bit). That's what makes it so much simpler to make the processor design. As if all instructions are of equal length, it isn't very hard to decode. Modern x86 MPU's have to decode x86 instructions into risc-like micro-ops (which are of equal length), and after that, can process them very efficiently. RISC doesn't inherently make a processor more efficient, but it does allow simpler designs and whenever a design is simpler, you have less human error in the design that could hurt efficiency.
However, the fact that all RISC instructions have to be of equal length makes writing a compiler for a RISC-like architecture a huge pain.