Sidebar Sidebar
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

Y Apple will never release a x86 version of Mac OS X!

i don' think AMD will bother will apple until apple's market share goes up...if it does

intel is a big and slow company, so AMD being smaller and more innovative can always give intel competition but will not overtake intel anytime soon

home comuting processors is only a part of what intel does so they have to worry about competition from other fronts

like i said, for now, apple is with ibm and motorola...for the forseeable future
 
Re: Y Apple will never release a x86 version of Mac OS X!

Originally posted by peter2002
If Apple ported their Mac OS X to x86, it would be no time before hackers would reverse engineer and eliminate any hardware or software protection and make the free version available on KaZaA, iMesh, Gnutella, etc. Why buy a Mac when you download one for free? Only die hard fans would buy a new one.

With the free warez version of Mac OS X, anybody could download and run any Mac OS X program using a dual boot partition program on a WIntel/AMD PC without having to buy a new or old Mac, and run OS X faster to boot with the new 3.06GHZ P4HT.
Click to expand...

I personally hope that Apple is able to survive without moving to x86. However, I wanted to add my opinion on why it would not be so simple to achieve this miracle hack you speak of.

Even if someone could figure out how to make a version of OS X on x86 boot on a non-Apple machine, they could only use the same processor/motherboard/chipset combination that Apple used unless they wrote additional drivers to support the other hardware options. Right now the x86 version of Darwin only runs on Intel processors with a 440BX chipset. Unless additional support code is written, it won't run on any other hardware. (AMD,VIA,SiS,NVidia,etc.) Same thing with the full OS X version -- it would only run if you matched the hardware in machines supplied by Apple.

So, while it might be possible to hack it to run on a non-Apple machine, it would take a lot of time and work to make it run on all x86 hardware.
 
Re: Re: Re: Ok

Originally posted by alex_ant
You speak as if "bringin new chips out" is like baking cookies or something - just finding the right ingredients in the right proportions and mixing them together and baking them and voila. It's a HUGE undertaking that requires MANY MONTHS of time and hundreds of millions of dollars in expenditures. If you think IBM is taking a long time with the 970, imagine how long AMD would take bringing out the same chip (in its plant which doesn't even have the technology to build them). Can you say 2008?
Click to expand...
(Hammer) Opteron and Athlon 64, 1H03.
PowerPC 970, 2H03.

Both are very similar in terms what they can do and how they are manufactured.

Hammer can do 32bit and 64bit, so does *gasp* IBM's PowerPC 970 and AMD is beating IBM to it.
 
Re: Re: Re: Re: Ok

Originally posted by MacCoaster

(Hammer) Opteron and Athlon 64, 1H03.
PowerPC 970, 2H03.

Both are very similar in terms what they can do and how they are manufactured.

Hammer can do 32bit and 64bit, so does *gasp* IBM's PowerPC 970 and AMD is beating IBM to it.
Click to expand...
Yes, Hammer can "do" 32 and 64-bit. 32 and 64-bit x86. It will run PowerPC code how? And it will feature what kind of replacement for AltiVec? And it's been delayed how many times? I don't understand. You're basically saying you want Apple to switch to AMD because the Opteron will be out <6 months sooner?

And in terms of moving to 64-bit, AMD is one of the last processor companies on earth to do this. IBM was years ahead of both AMD and Intel with 64-bit PPC chips - it's just that they were never used in Macs.
 
Re: Re: Re: Re: Re: Ok

Originally posted by alex_ant

Yes, Hammer can "do" 32 and 64-bit. 32 and 64-bit x86. It will run PowerPC code how? And it will feature what kind of replacement for AltiVec? And it's been delayed how many times? I don't understand. You're basically saying you want Apple to switch to AMD because the Opteron will be out <6 months sooner?

And in terms of moving to 64-bit, AMD is one of the last processor companies on earth to do this. IBM was years ahead of both AMD and Intel with 64-bit PPC chips - it's just that they were never used in Macs.
Click to expand...
You never said it had to exclusively run PowerPC code.

IIRC, Hammer has 3DNow!, SSE, SSE2, and so on. Similar enough to Altivec.

I want Apple to switch to Hammer because it will be out earlier, run faster than the 970s, etc.

Yes, IBM has had 64bit for a long time. But you were talking about the PowerPC 970. That's both 32bit and 64bit AND it is a desktop chip. POWER4 is not a desktop chip.

Both the Hammer series and PowerPC 970 are desktop/workstation processors and AMD has announced theirs for a while, has worked a very long time on it, and is releasing it 1H03.
 
Re: Re: Re: Re: Re: Re: Ok

Originally posted by MacCoaster

You never said it had to exclusively run PowerPC code.

IIRC, Hammer has 3DNow!, SSE, SSE2, and so on. Similar enough to Altivec.

I want Apple to switch to Hammer because it will be out earlier, run faster than the 970s, etc.
Click to expand...

Not only is AltiVec far beyond these in performance and implementation, it's what Mac software already uses. No porting necessary.

The 970 will be slightly slower than the Hammer at integer and fp speed, but it will have greater memory bandwidth and AltiVec-on-steroids which no PC processor will be able to compete with. It will rock your socks in the multimedia & graphics & audio apps which are Apple's bread & butter.
Yes, IBM has had 64bit for a long time. But you were talking about the PowerPC 970. That's both 32bit and 64bit AND it is a desktop chip. POWER4 is not a desktop chip.
Click to expand...

I wasn't talking about the Power4. There have been various 64-bit PPC-compatible desktop-worthy chips available for years. One being the 620 from the late '90s. If you want to make a case for Apple switching to AMD, feel free, but AMD is not beating anyone to 64-bit.
Both the Hammer series and PowerPC 970 are desktop/workstation processors and AMD has announced theirs for a while, has worked a very long time on it, and is releasing it 1H03.
Click to expand...
You're saying Apple should make the switch despite mere months of time difference between the release of these two chips, and I believe that would be a monumentally, catastrophically dumb decision, for reasons everyone else has already outlined in this thread, which I've not seen seriously challenged.
 
Re: Re: Re: Re: Re: Re: Re: Ok

Originally posted by alex_ant
Not only is AltiVec far beyond these in performance and implementation, it's what Mac software already uses. No porting necessary.
Click to expand...
Then why is the Pentium 4 and Athlon XP slapping the G4 silly. In fact, the people who report OS X and FCP running on an "AMD" box says it's significantly faster.
The 970 will be slightly slower than the Hammer at integer and fp speed, but it will have greater memory bandwidth and AltiVec-on-steroids which no PC processor will be able to compete with. It will rock your socks in the multimedia & graphics & audio apps which are Apple's bread & butter.
Click to expand...
Altivec-on-steroids? Last time a couple people informed Altivec on the PowerPC 970 didn't improve that much over the G4. But I'll have to see.

From what I remember, Hammer has HT 6.4GB/sec just like the PowerPC 970. Infact, the Hammer can have three HT links providing up to 15 or possibly much higher GB/sec of throughoutput than the 970.
You're saying Apple should make the switch despite mere months of time difference between the release of these two chips, and I believe that would be a monumentally, catastrophically dumb decision, for reasons everyone else has already outlined in this thread, which I've not seen seriously challenged.
Click to expand...
Yes, six months ahead is a lot. Just when the PowerPC 970 is going to be used in Power Macs, AMD will have had improved and released faster Hammers.
 
Re: Re: Re: Re: Re: Re: Re: Re: Ok

Originally posted by MacCoaster

Then why is the Pentium 4 and Athlon XP slapping the G4 silly. In fact, the people who report OS X and FCP running on an "AMD" box says it's significantly faster.
Click to expand...

Everybody knows the G4 sucks now. But AltiVec - the faster version of which will be implemented in the 970 - will be faster than anything the P4 or Athlon or Hammer will be able to compete with.
Altivec-on-steroids? Last time a couple people informed Altivec on the PowerPC 970 didn't improve that much over the G4. But I'll have to see.
Click to expand...

Even if the specs of the AltiVec unit were unchanged compared to the version on the G4, performance would still be much better due to the bandwidth problems being eliminated.
From what I remember, Hammer has HT 6.4GB/sec just like the PowerPC 970. Infact, the Hammer can have three HT links providing up to 15 or possibly much higher GB/sec of throughoutput than the 970.
Click to expand...

If I'm not mistaken, the Hammer's bus is 533MHz and capable of 5.4GB/s vs. the 900MHz 7.2GB/s bus of the 970. (At least according to the chart on this page)

Also, although the Hammer will be released before the 970, the 970 will be on .09 micron before the Hammer. Clearly the 970 will be a very competitive chip; it would be foolish to throw this away and bet the company on x86.
 
Re: Re: Re: Re: Re: Re: Re: Re: Re: Re: Ok

Originally posted by MacCoaster

Interesting that you know that this is true now. What are your sources? etc. :rolleyes:
Click to expand...
Well, cripes, how could it not be true? Intel and AMD have nothing on the horizon that could possibly compete with any kind of next-generation AltiVec unit that could possibly appear in the 970. Even if the 970's AltiVec unit is extremely weak - as "weak" as the G4's - it will still have much more memory bandwidth available to it and will be able to perform much better than it does now (since bandwidth is currently the limiting factor in its performance). Meaning if a well-tuned AltiVec program running on a 1GHz G4 can dust a similar program running on a 2GHz Athlon now, it would be a virtual certainty that the same program running on a 1.8GHz 970, with vastly greater memory bandwidth than a G4, would dust the same program optimized for a 2GHz Hammer.
 
Re: Re: Re: Re: Re: Re: Re: Re: Re: Re: Re: Ok

Originally posted by alex_ant

Well, cripes, how could it not be true? Intel and AMD have nothing on the horizon that could possibly compete with any kind of next-generation AltiVec unit that could possibly appear in the 970. Even if the 970's AltiVec unit is extremely weak - as "weak" as the G4's - it will still have much more memory bandwidth available to it and will be able to perform much better than it does now (since bandwidth is currently the limiting factor in its performance). Meaning if a well-tuned AltiVec program running on a 1GHz G4 can dust a similar program running on a 2GHz Athlon now, it would be a virtual certainty that the same program running on a 1.8GHz 970, with vastly greater memory bandwidth than a G4, would dust the same program optimized for a 2GHz Hammer.
Click to expand...

intel is way ahead...but maybe not in doing the programs we may like...intel has a 3 ghz chip already...can't be that bad
 
You must log in or register to reply here.
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.
Top Bottom
Back