Look out MS

[Tiger]

Mac OS X on x86

 

[igiveup]

Yeah, this has been floating around for a while, but I must admit that this is the first thing I have seen to suggest it is "complete" and just being maintained.

Honestly, Steve Jobs would be a fool not to keep his options open and I for one think its a great idea. The one thing that has kept me from trying out apple and their associated products and software is the fact that they are proprietary hardware based. I hope they smarten up and keep it open.

 

[n0cmonkey]

We all knew they were developing it. Darwin has already been ported. If Apple ever *switches* to x86 they have lost atleast one customer, and gained a whole lot of trouble.

 

[TheOmegaCode]

It would be sad to see Apple switch to x86, but I would like to get my hands on OSX for x86...

 

[boran]

make that three.... I'll try it at least ... then we'll see if it's better than win2k

 

[CTho9305]

Originally posted by: n0cmonkey
We all knew they were developing it. Darwin has already been ported. If Apple ever *switches* to x86 they have lost atleast one customer, and gained a whole lot of trouble.

What do you have against faster x86 chips? You'd rather be loyal to motorola and have relatively slow processors? OS X on x86 would be awesome, especially if it had something like wine (unlikely though 🙁)

edit: and I HIGHLY doubt it would run on commodity hardware. The reason OS X works so well is the intergration with the underlying hardware. Most likely apple would design VERY proprietary boards that just use x86 processors. Probably normal ram, and normal PCI, but I would think they would be VERY strict about approving a driver or not so that it won't run on normal computers.

 

[NogginBoink]

VERY interesting. This would certainly change the entire landscape of the PC industry. I don't see why it's taken Apple this long to do this, or why they're hesitating now.

 

[Tiger]

If they officially port the thing to x86 and then tie it to a proprietary mobo, they deserve their 4% market share.
It would just be another in a long line of Steve Jobs ego driven fubars. If I was a major stockholder I'd want his head on a platter.

 

[Nothinman]

The problem would be app support. As soon as OS X is put on an x86 box it can't run all the PPC or 68k apps, all the current OS X apps need ported and none of the Classic apps will run, unless they write a PPC/68k->x86 translator which would be terribly slow. And everything that uses Altivec needs a lot of asm rewritten.

OS X is better than Win2K, easily. But if none of it's most important apps (i.e. Photoshop, quark, Office, etc) don't run, it's useless to the existing Mac users and just becomes another BSD fork.

 

[n0cmonkey]

Originally posted by: CTho9305

Originally posted by: n0cmonkey
We all knew they were developing it. Darwin has already been ported. If Apple ever *switches* to x86 they have lost atleast one customer, and gained a whole lot of trouble.

What do you have against faster x86 chips?

Nothing, but I have other OSes for x86 platform.

You'd rather be loyal to motorola and have relatively slow processors?

Not at all. The g3 Sahara in the current iBooks I believe is a non-Motorola chip, but Im not sure. I am a fan of Motorola, just not in this area.

OS X on x86 would be awesome, especially if it had something like wine (unlikely though 🙁)

Yes, it would be awesome, I just wouldnt buy it.

edit: and I HIGHLY doubt it would run on commodity hardware. The reason OS X works so well is the intergration with the underlying hardware. Most likely apple would design VERY proprietary boards that just use x86 processors. Probably normal ram, and normal PCI, but I would think they would be VERY strict about approving a driver or not so that it won't run on normal computers.

That way they could possibly stay alive.

 

[n0cmonkey]

Originally posted by: NogginBoink
VERY interesting. This would certainly change the entire landscape of the PC industry. I don't see why it's taken Apple this long to do this, or why they're hesitating now.

They have had x86 ports for a long time. Back in the day, m68k and PPC chips were faster than comparable x86 chips. The mhz gap has changed things. Rhapsody and OpenStep apparently had working x86 ports released to beta testers/developers.

 

[n0cmonkey]

Originally posted by: Tiger
If they officially port the thing to x86 and then tie it to a proprietary mobo, they deserve their 4% market share.
It would just be another in a long line of Steve Jobs ego driven fubars. If I was a major stockholder I'd want his head on a platter.

They are not a software company. How would they make money if you could run their OS on any old POS x86 machine?

 

[n0cmonkey]

Originally posted by: Nothinman
The problem would be app support. As soon as OS X is put on an x86 box it can't run all the PPC or 68k apps, all the current OS X apps need ported and none of the Classic apps will run, unless they write a PPC/68k->x86 translator which would be terribly slow. And everything that uses Altivec needs a lot of asm rewritten.

There are m68k emulators out there right now (bassilisk?). PPC emulators would be much harder to design and impliment, and even then they would be ungodly slow. Apple could help fix the AltiVec issue by helping out the gcc project more (from what I understand) 😉

Much of OS X has altivec optimizations, so much of OS X would have to be reworked, debugged, tested, etc. Even with Apple's strive for excellence, this would be a painful peroid for a lot of people. Hell even OS X 10.0 was incomplete and probably shouldnt have been released as a full product.

 

[Tiger]

They are not a software company. How would they make money if you could run their OS on any old POS x86 machine?

Maybe they should be. If OS X is as good as everbody seems to think they could take a serious bite out of MS's market monopoly.

 

[n0cmonkey]

Originally posted by: Tiger

They are not a software company. How would they make money if you could run their OS on any old POS x86 machine?

Maybe they should be. If OS X is as good as everbody seems to think they could take a serious bite out of MS's market monopoly.

I disagree. Unfortunately, a lot of their "usefulness" is linked to Microsoft. Many people (not myself) complained about a lack of Office for Mac OS X for quite a while. Without it, a lot of people were unwilling to switch.

Breaking into Microsoft's stranglehold of the industry isnt easy, and I dont know if Apple would be willing to try. Why take unncessary risks?

 

[lowtech1]

You are only as fast/good as your worst part.

Darwin is quite good, but the lack of apps is a hindrance for many newer users & for the unhappy Wintel crowd.
At the present the bottleneck on the PPC is not only on the bus architecture but also on the lacking PPC core, which is much worst than the 30~33mhz bus verses the Wintel 50~66mhz bus speed 10 years ago. It doesn?t matter how good Mac engineers are & how great Altivec is, because at the advance rate of x86 32/64 bits architecture will make a Mac look like a dinosaur in the next few years. At the moment Mac need to pump more cash into PPC development or port to x86, because it will be a blip in history like the great Amiga if they don?t hurry up the development.

[edit] Ps. My Amiga 500/1200 & Apple IICwas once great machines, but now collecting dust in my basement.

 

[n0cmonkey]

Originally posted by: lowtech
You are only as fast/good as your worst part.

Darwin is quite good, but the lack of apps is a hindrance for many newer users & for the unhappy Wintel crowd.
At the present the bottleneck on the PPC is not only on the bus architecture but also on the lacking PPC core, which is much worst than the 30~33mhz bus verses the Wintel 50~66mhz bus speed 10 years ago.

Can you explain the bolded part?

EDIT: fixed my bolding

 

[lowtech1]

The problem would be app support. As soon as OS X is put on an x86 box it can't run all the PPC or 68k apps, all the current OS X apps need ported and none of the Classic apps will run, unless they write a PPC/68k->x86 translator which would be terribly slow. And everything that uses Altivec needs a lot of asm rewritten.

Backward compatibility is not a problem at all with Apple, because it never has been in their business plan.
I bought my Amigas, because Apple didn?t follow through their promise on legacy software support from the Apple to Mac transition (less than 40% of Apple apps could run on Mac OS during the late 80s to early 90s).

They has already lost this customer & I'm not going to ever go back.

 

[lowtech1]

Pre PowerPC chip the Mac bus stuck at 30~33mhz, while the x86 bus was pushing 50~66mhz. At that time the bus was a slight hinder for the Mac, but the superior of the CPU architecture (even those the CPU mhz was slightly slower) and OS make the Mac a much better machine to be had compare to Wintel 386/486. Once PPC went into full production it shown how superior Mac OS was compare to Windows/OS2.

At the moment there isn't even a rummor that Apple is going to use the Power4 for the Mac, and price/performance ratio would be poor if it happen. Another thing is that Mac is very late in the mhz game at the moment....my speculation that it would take Apple at least another year to begin to move production on to the Power4.

 

[n0cmonkey]

Originally posted by: lowtech
Pre PowerPC chip the Mac bus stuck at 30~33mhz, while the x86 bus was pushing 50~66mhz. At that time the bus was a slight hinder for the Mac, but the superior of the CPU architecture (even those the CPU mhz was slightly slower) and OS make the Mac a much better machine to be had compare to Wintel 386/486. Once PPC went into full production it shown how superior Mac OS was compare to Windows/OS2.

Which bus? I can think of several off the top of my head, but my brain isnt working out well enough to figure out which one you mean (sorry, weekend fuzz is still there 🙁). PCI bus? Memory bus? 😕

At the moment there isn't even a rummor that Apple is going to use the Power4 for the Mac,

Uhhh, yeah there are quite a few of those rumors. Especially with the new "desktopish" POWER4 chip that happens to have an unnamed SIMD extension (something IBM is rumored to not be fond of) of 160 instructions that IBM is developing and will release/announce/whatever next month(?). AltiVec happens to have 160 instructions... Ill try and find linkage tomorrow.

[aq]nd price/performance ratio would be poor if it happen. Another thing is that Mac is very late in the mhz game at the moment....my speculation that it would take Apple at least another year to begin to move production on to the Power4.[/quote]

Code developed on Motorola's G4 or IBM/Motorola's g3 shouldn't have to have much more than a recompile to work on another PPC chip, if that! If that is the move they are planning on making, Im sure they have working code.

 

[n0cmonkey]

link

Session One: PC Processors
Kevin Krewell, Senior Editor, Microprocessor Report; General Manager, MDR

Breaking Through Compute Intensive Barriers - IBM's New 64-bit PowerPC Microprocessor

Peter Sandon, Senior Processor Architect, Power PC Organization, IBM Microelectronics

IBM is disclosing the technical details of a new 64-bit PowerPC microprocessor designed for desktops and entry-level servers. Based on the award winning Power4 design, this processor is an 8-way superscalar design that fully supports Symmetric MultiProcessing. The processor is further enhanced by a vector processing unit implementing over 160 specialized vector instructions and implements a system interface capable of up to 6.4GB/s.

Its possible its not intended for Apple, but the SIMD is not usually what you incude in a server chip, and do you think IBM is going to go after the PC market again?

 

[lowtech1]

PCI bus? Memory bus?

I believe that Mac uses Nubus architecture at that time instead of the superior VLB/MCA/PCI bus.

Code developed on Motorola's G4 or IBM/Motorola's g3 shouldn't have to have much more than a recompile to work on another PPC chip, if that! If that is the move they are planning on making, Im sure they have working code.

If they all ready have working example then Mac will have a bit of air left in their tank, and that is only if the Mac fan are willing to paid for expensive 64bits & SMP system that have to hobble along using recompile programs (it will take a few more years till 64bits apps to exist/mature). Even recompile the OS & apps will take some testing to ensure bug free status, and Apple wouldn?t want repeat the paper launch of OSX v10.0.


[edit] Too bad Motorola/IBM fuged up with the PPC/OS2, because Altivec is a much more elegant architecture than 3D-Now/SSE.

 

[n0cmonkey]

Originally posted by: lowtech

PCI bus? Memory bus?

I believe that Mac uses Nubus architecture at that time instead of the superior VLB/MCA/PCI bus.

Ok, you are talking old school, not today. NuBus failed just like the MCA(?) bus did for IBM.

Code developed on Motorola's G4 or IBM/Motorola's g3 shouldn't have to have much more than a recompile to work on another PPC chip, if that! If that is the move they are planning on making, Im sure they have working code.

If they all ready have working example then Mac will have a bit of air left in their tank, and that is only if the Mac fan are willing to paid for expensive 64bits & SMP system that have to hobble along using recompile programs (it will take a few more years till 64bits apps to exist/mature). Even recompile the OS & apps will take some testing to ensure bug free status, and Apple wouldn?t repeat the paper launch of OSX v10.0.

Im not positive about a recompile being necessary, it may have the backwards compatibility The Hammer is supposed to have, not sure.

Apple isnt a stupid company. Im sure they have various ways to go when different things happen. If Motorola does something stupid and Apple has to find something new they probably have a couple of different solutions worked out. One of them is probably IBM, who is very active in the PPC market and another may be an x86 chip. We will find out if the time comes.

Apple's PowerMac series is all dual processors at this point, so SMP is the norm. A 64 bit chip may not be much more expensive than the current 32bit chips. We will have to wait and see. The application support in a switch from current PPC chips to possible 64bit PPC chips of the future will be much less painful than the switch from the m68k processors to PPC or the switch from Mac OS to Mac OS X.

 

[CTho9305]

Apple is a company. They make a line of computers called "Macintosh".
Dell is a company. They make a line of computers called "Inspiron".

 

[lowtech1]

Apple is a company. They make a line of computers called "Macintosh".
Dell is a company. They make a line of computers called "Inspiron".

Apple is a company that make a line of Macintosh computer since the early 90s, but prior to that they make a line of computer that called Apple II.