Dual boot x86 is a slow death. Microsoft has already figured out how to screw this up. Refer to old Be articles and see what they were complaining and suing about. Microsoft changed the boot process so they could get dual boot to work correctly.
On PPC NT 3.5, it was not upgraded because no one bought it. You had the OS, but none of the Apps. Most of the ISV didn't want to support it, so it didn't sell and died. Again, the example of IBM/s O/S 2 comes into play here. Why support many when you can support 1 and get most of the same sales dollars.
What most of the people i've talked to want is to run OS X on generic x86 boxes. They aren't going to buy Apple hardware. No hardware sales = no $$$. The want to be able to run Windows as a safety valve.
I see a transition to IBM's POWER4/5 before an x86 transition. Altivec or no Altivec.
<strong>If Apple move to another chip, for the sake of argument say it's Intel's next big offering, then there are several things that are going to be a given:
1. Apple will be the only manufacturer to build the Mac.
2. OS X might well run on an Intel chip but there's no way it's going to run on anyone else's hardware except Apple's.
3. There's no way Apple's hardware will EVER allow you to install/run Windows of any form, outside of emulation or terminal clients.</strong><hr></blockquote>
I'm afraid anybody who thinks Apple can lock X to their hardware or block Windows hasn't been around computers that long!
I would give it 2 months max before you could hack any old PC to run X. The same with Windows. You would probably even get it going on 'Xbox2'
My first Mac was an Amiga 4000. All it had in common was the CPU but a 400k (or so) shareware program had a perfectly usable MacOS 7.1 running on it with no problems whatsoever.
I'm afraid anybody who thinks Apple can lock X to their hardware or block Windows hasn't been around computers that long!
</strong><hr></blockquote>
hey blackcat, i'm happy to admit ignorance of how it might be done, but why wasn't this done to OS9? why wouldn't WinNT for PPC run on the Mac?, or BeOS 5 on the G3?...it's all in the boot Rom and if Apple don't give that out I don't see how you can do it legally.
But i'm happy to be wrong, I just think Steve binned off the cloners for a reason, quality of hardware, and he'll fight to keep it that way, after all a Mac is more than the sum of it's parts, but only because Apple keep it tight.
<strong>That URL for the dual machine has results of 10,564,827 & 10,557,765.
I almost hate to bring this up, but the URL for all PPC machines lists the result for
a 1600(re: 1.6GHz?????) machine as 16,991,648.
Conclusions???
1- the 1600 machine is bogus, but has been there a long time(why hasn't it been removed)
2-the results for the 1000MHz machines are for single processors. If so where did said 1GHz single processor machines come from?
3-the result for the 1600MHz machine is for a dual 800MHz machine.
4-Some "wild and crazy guy" overclocked a single to 1.6GHz
Since I'm technically challenged could someone correct me </strong><hr></blockquote>
The most important thing to understand is that the RC5 challenge is strictly CPU-limited. The processor fetches a few bytes, loops for an eternity over a few dozen instructions, fetches a few new bytes, etc. The distributed.net client just lets each CPU compute their own block of keys, so there is almost no communication between processors. Ergo, a dual 800 Mhz will be just as fast as a single 1.6 Ghz. And a single 1 Ghz will be just as fast as half the result you get with a dual 1Ghz.
As such, you can be quite creative and report the dual 800 Mhz result as the result of a 1.6 Ghz, divide the result from your dual 1Ghz by two to get the result for one CPU or calculate the result for the fabled 16 CPU 1 Ghz Xworkstation.
<strong>On the PC side, they are becoming more and more Mac like in the regard that more and more things are on the motherboard and can't be replaced. Most of these parts are inferior to the Mac equivelent and also slow the machine down considerably from whatever sky high numbers the processor produces. So you can have a 2 ghz $599 wintel box, but it will have integrated video, audio, and networking all lugging down that processor.</strong><hr></blockquote>
And you shouldn't forget that some of these components use the CPU for certain tasks. See the Win-modems that drain your CPU for instance.
I hope that's sarcasm. I have yet to see one benchmark in which the the winning Photoshop computer won but such a large margin.
x86 hardware IMO is simply not superior enough for me to want a switch personally. I would like to see another two years of Apple/IBM/Moto work before making a judgement.</strong><hr></blockquote>
One of the things Chris Cox said in one of the Forums about PS on the Mac was that it was partially slowed down because both PS and OSX are writing the screen to buffers at the same time and a fix was being made for this. No doubt Jaguar has influenced the work schedule
All the prices need lowering. It's better to sell 2 Macs at $2500 than 1 at $3500. Apple needs resellers like CompUSA and PCWorld on side too. I get fed up of seeing crashed Macs in stores.
</strong><hr></blockquote>
well, it all depends on what you mean by better...
if apple is looking to increase market share at all cost, then yes, it is better.
however, given the current margins (30%), the cost of a $3.5k machine is $2450. therefore, they'd be making $50 per machine, which would make it $100
for the two machines they sold given the price cut.
congratulations.
admitting the market share doubles, you've just found a way to divide apple's profits tenfold.
now, the one way to do this would be to cut the margin in half...
the same 3.5k machine would then cost $2975, but would it sell twice as much at that price point ? sorry, but i doubt it.
"LBB, I think your LAST RESORT description sums up Mac x86 well."
Hmm. For me. It still depends. Apple may consider it to get into certain markets. People who wouldn't otherwise consider a Mac.
They're only going to reach so many people with 'Switch' campaign. If they want to grow and grow. They may have to consider the big picture.
(Anyway, I'm sure Apple will not repeat IBM's mistake. Apple aren't IBM. Apple has OS 'X'. For the next year or so...Apple have got to concentrate everything on piling on pressure on XP. iApps crazy, push the QE envelope and Rendezvous/isync ideas and the switch campaign. They've got to show for multimedia they are king.)
This 'Dual boot' strat' may mean nothing more than bundling Virtual PC on a Power X style processor (ie a processor that(!) powerful...it feels you're funning a 1 gig Pentium in emulation...) I can't think if that many compelling software reasons to be 'PC' anyhow. The Mac kinda has the best of all apps now anyhow. Best OS...best 3D apps, cream of the games...the best Office app Best version of Photoshop...geez...
"There really isn't much wrong with the Mac in areas which count on the balance sheet. We're not too hot in 3D but that could well be about to change if we see the 7470 next week."
Begin to change. That hardware there...yer talkin' about...is catch up. It'll be contempary until the Hammer hits early next year and then they had BETTER have some Rio/7500 juicin' stuff up their sleeve.
It'll be interesting to see the 1.4 G4 on the 'let the animal loose' mobo DDR etc. If it repairs those damning Lightwave and Photoshop benches...then I'll gladly eat my words My bet is we'll pull just behind to near even. But then, there's the price...
"Perception and Price are what needs changing, not Processor."
Don't agree. Both need changing. The G4 is long overdue for consumer range only badge.
"A 1.4Ghz G4 should help - in real terms that's about the same speed as a 2Ghz P4, add DDR and AltiVec and it looks pretty good. In truth email and Word are fine on a 500Mhz chip on any platform."
We'll see if benches back you up.
"All the prices need lowering. It's better to sell 2 Macs at $2500 than 1 at $3500."
Agreed. The last POWERMac I bought some years ago...cost me an arm and a leg. If Macs were cheaper...I'd have upgraded twice by now. As it is...expensive and old tech' equals long update cycle. And APPLE don't get the sales/market penetration they want. It's a downward cycle. Hopefully the news of Apple increasing their purchasing of Taiwan components to make their products price competitive is true. ie goes on cheaper Macs not higher margins.
"Apple needs resellers like CompUSA and PCWorld on side too. I get fed up of seeing crashed Macs in stores."
Agreed. But Apple doesn't help themselves here in the Uk. Remember when Tescos sold old Crt imacs in stores. Apple didn't approve of these cheap grey market Macs. But hey...they were reaching markets and people they wouldn't normally reach. Their marketing is still too aloof. I'm surprised they're still in business.
"Finally Apple needs to buy a PC games house so that the best games are on Mac first. They need to ensure affordable after-market graphics cards are out there too."
Agreed. Yeesh. Put a bid in for id games. I'd love to see PC uproar about that in light of M$ buying Bungie! Buy the makers of Unreal Tournement. Don't have to buy many games companies. Just one or two, the ones that matter. And cut PC support! Give 'gamers' a reason to buy mac.
"These things would offer far more ROI than going up against MS as a PC builder."
Over $4 billion. They could use some of that to secure their future as a software company. A merger with Adobe. Buy out Quark. Buy Maya.
Then Apple are in control. Doesn't matter a shat which processor they run anything on then!
Lemon Bon Bon
PS. Nothing wrong with irrational opinions. We're on a rumours board.
I think people aren't getting Moki's point -- he is saying that Apple may introduce an additional line of Macintoshes which are based on an x86 processor. They will still be Macs, and as such be capable of running (a future version of) MacOSX. This does not imply that this version of MacOSX will automatically run on any x86-based PC, just like Solaris won't. They also may not be able to run Windows (more interesting would be a "Classic-like" environment for Windows apps, but we don't know what MS' or Apple's position on that would be).
Sure some people will figure out how to hack it to run on some PCs, but these people will be running the OS illegally and it won't have solid and/or fast drivers to support it. This will not amount to a significant number of people, and no corporation will do it and risk the wrath of Apple Legal... not to mention the complete lack of software support. If a few home users do it whom wouldn't have bought a Mac anyhow, then all they are doing is potentially expanding the Mac software market a little (but probably not since they are likely to be pirates anyhow). These hacks are not enough of a market to worry about, and they aren't the sort of market from which money can be extracted.
The PowerPC line of Macs will continue, and I wouldn't be surprised to see it thrive over time due to the backing and R&D work of IBM. This puts Apple in the rather nice position of being able to leverage the best of both worlds -- Intel & AMD seem to be gambling on pushing faster and faster clock rates, whereas IBM is pushing on more and more cores. MacOSX is pretty well position to leverage both. Take your pick... 1 50GHz Pentium9, or a stack of 8 quad core 2 GHz PowerPC G7s. :eek:
I could also see one or two MacOSX licensees whom Apple brings on board with specific and limited contracts to simply to be secondary suppliers for x86 and PowerPC hardware on which to run Apple's OS. This is important in a bunch of business and government markets that Apple no doubt wants to be in. These tend to be high margin environments which emphasize solutions and support, so by choosing appropriate partners (e.g. IBM) Apple doesn't have to worry about them suddenly jumping into Apple's traditional markets and doing what PowerComputing et al did.
This does explain Jobs' comment fairly well, but if you believe it then you have to give Moki more credibility and we know how much he hates that!
<strong>Exactly the same way IBM's BIOS was reverse engineered by Compaq. Took Compaq a couple months and a couple dollars, but it is all quite legal to do. Being able to undercut Apple HW prices by imitating motherboard designs and setting in a reingineered BIOS ROM without having to do the basic R&D would be ridiculously easy for an HP or DELL (even though DELL doesn't really do R&D). There are a few quite well defined legal wickets a company would need to comply with along the way, but it is an age old and legally acceptable process.</strong><hr></blockquote>
I think this lesson has been learned in the industry and Apple wouldn't go there unless they could truly prevent it. No PC manufacturer is going to bother with Apple waving its legal stick around -- its not worth the relatively small market share they are likely to gain. Remember, Compaq did it to get into IBM's market and IBM was the giant.
How difficult would it be to put AltiVec on a DSP or coprocessor?
If that's doable, combined with RapidIO or HT, then I would think that it wouldn't matter what brand of CPU was used.
I really don't think that Apple would go the x86 route, but this would open possibilities for pretty much anything (as long as it was RIO/HT enabled). For that matter, if we're gonna talk about using POWER*, then why not infiniband (I think that's right, no?)?
As an aside, I've been wondering lately about the possiblity of a secondary on-board graphics chip for doing the OS-level stuff (window redraws, etc), freeing up the main GPU for 3D calculations and such. How possible would this be? Would this increase speed more with Jaguar?
<strong>All the prices need lowering. It's better to sell 2 Macs at $2500 than 1 at $3500.</strong>
and Originally posted by mattbr:
<strong>given the current margins (30%), the cost of a $3.5k machine is $2450. therefore, they'd be making $50 per machine</strong>
and Originally posted by Lemon Bon Bon:
<strong>Agreed. The last POWERMac I bought some years ago...cost me an arm and a leg. If Macs were cheaper...I'd have upgraded twice by now. As it is...expensive and old tech' equals long update cycle. And APPLE don't get the sales/market penetration they want. It's a downward cycle. Hopefully the news of Apple increasing their purchasing of Taiwan components to make their products price competitive is true. ie goes on cheaper Macs not higher margins.</strong>
<hr></blockquote>
a high gross margin (1 - (cost of goods / revenues)) does not necessarily indicate that a company could cut prices and remain profitable. in apple's case, their 28% margin drops to 12% when you amortize r&d expenses, and drops to under 2% when you include marketing and administrative expenses. using these numbers (assuming a flat distribution of costs and profits across all product lines), by reducing the price of a $3500 mac to $2500, apple would be losing over $940.
as a point of reference, dell has a gross margin of 17% which drops to 16% after r&d and to just under 7.5% after other operating expenses.
[quote]1- the 1600 machine is bogus, but has been there a long time(why hasn't it been removed)<hr></blockquote>
the 1600 machine is simply a dual 800.
the point of Apple going x86 while still offering kick ass ppcs is as moki mentionned: some corporations *require* computers to have intel chips or to be able to run windows.
How difficult would it be to put AltiVec on a DSP or coprocessor?
If that's doable, combined with RapidIO or HT, then I would think that it wouldn't matter what brand of CPU was used.
I really don't think that Apple would go the x86 route, but this would open possibilities for pretty much anything (as long as it was RIO/HT enabled). For that matter, if we're gonna talk about using POWER*, then why not infiniband (I think that's right, no?)?
As an aside, I've been wondering lately about the possiblity of a secondary on-board graphics chip for doing the OS-level stuff (window redraws, etc), freeing up the main GPU for 3D calculations and such. How possible would this be? Would this increase speed more with Jaguar?</strong><hr></blockquote>
altivec is a dsp already so why do you want to put it on another dsp? putting multiple cores into one die has been no secret and motorola's embedded mpc 8260 is having two core inside one die: one is central 603e core and another one is 32-bit risc core to handle communication interface. the interconnect is through local bus. i heard in their new serial mpc85xx, they are putting an internal crossbar switch between its core cpu and its peripheral cpu.
for graphical intensive application, faster cpu can not solve everything. take a look at an sgi machine. it has darn slow cpu, but i can gauarantee that the low end machine from sgi could easily beat any top speed pc or mac in many aspects.
there is nothing new on how sgi did it. and i am sure there might be new way to do better. it matters whether apple wants to use it or not. there are tons of smart guys in apple, so i am pissed for the fact that they have not yet used any to improve a mac.
i guess sj is correct that apple has to put os x out first and then focus on hardware. this comment is an insulting as well as enlightment. it is kind bad that apple kept using old technology to lure money from their loyal customers, but they need money to survive. now i think apple will focus on the hardware or more grand idea on how to ride on os x. but the bad thing is that it means that i will have to wait another year for a power new macintosh.
<strong>As an aside, I've been wondering lately about the possiblity of a secondary on-board graphics chip for doing the OS-level stuff (window redraws, etc), freeing up the main GPU for 3D calculations and such. How possible would this be? Would this increase speed more with Jaguar?</strong><hr></blockquote>
That ties up really nicely with rumour that Apple might usee Nforce2 chipset.
Part of it can include an on motherboard Gforce4 MX, which would make every Nforce2 based Mac Quartz Extreme capable regardless of the graphics card you add (obviously adding a non OpenGL card would mess things up) so people could upgrade to better cards or dual monitor support more easily.
Quartz Extreme is another reason why PPC is less required too I suppose.
All the prices need lowering. It's better to sell 2 Macs at $2500 than 1 at $3500.</strong><hr></blockquote>
This is a misperception I see repeated over and over. Selling two Macs at $2500 does bring in $5000, which is more than $3500, but what if Apple's cost on each Mac is $2300? Selling two of those at $2500 gives them $400 profit while selling one of them at $3500 gives them $1200 profit. Business isn't about the greatest possible gross sales revenue, it's about profits.
Comments
On PPC NT 3.5, it was not upgraded because no one bought it. You had the OS, but none of the Apps. Most of the ISV didn't want to support it, so it didn't sell and died. Again, the example of IBM/s O/S 2 comes into play here. Why support many when you can support 1 and get most of the same sales dollars.
What most of the people i've talked to want is to run OS X on generic x86 boxes. They aren't going to buy Apple hardware. No hardware sales = no $$$. The want to be able to run Windows as a safety valve.
I see a transition to IBM's POWER4/5 before an x86 transition. Altivec or no Altivec.
<strong>If Apple move to another chip, for the sake of argument say it's Intel's next big offering, then there are several things that are going to be a given:
1. Apple will be the only manufacturer to build the Mac.
2. OS X might well run on an Intel chip but there's no way it's going to run on anyone else's hardware except Apple's.
3. There's no way Apple's hardware will EVER allow you to install/run Windows of any form, outside of emulation or terminal clients.</strong><hr></blockquote>
I'm afraid anybody who thinks Apple can lock X to their hardware or block Windows hasn't been around computers that long!
I would give it 2 months max before you could hack any old PC to run X. The same with Windows. You would probably even get it going on 'Xbox2'
My first Mac was an Amiga 4000. All it had in common was the CPU but a 400k (or so) shareware program had a perfectly usable MacOS 7.1 running on it with no problems whatsoever.
as for those distributed.net scores, I interpreted them as 2x500MHz =1Ghz and 2x800MHz = 1.6Ghz.
So the 2x1000Mhz would probably score around 20'000'000 keys/sec.
G-news
<strong>
I'm afraid anybody who thinks Apple can lock X to their hardware or block Windows hasn't been around computers that long!
</strong><hr></blockquote>
hey blackcat, i'm happy to admit ignorance of how it might be done, but why wasn't this done to OS9? why wouldn't WinNT for PPC run on the Mac?, or BeOS 5 on the G3?...it's all in the boot Rom and if Apple don't give that out I don't see how you can do it legally.
But i'm happy to be wrong, I just think Steve binned off the cloners for a reason, quality of hardware, and he'll fight to keep it that way, after all a Mac is more than the sum of it's parts, but only because Apple keep it tight.
[quote] <hr></blockquote>
<strong>That URL for the dual machine has results of 10,564,827 & 10,557,765.
I almost hate to bring this up, but the URL for all PPC machines lists the result for
a 1600(re: 1.6GHz?????) machine as 16,991,648.
Conclusions???
1- the 1600 machine is bogus, but has been there a long time(why hasn't it been removed)
2-the results for the 1000MHz machines are for single processors. If so where did said 1GHz single processor machines come from?
3-the result for the 1600MHz machine is for a dual 800MHz machine.
4-Some "wild and crazy guy" overclocked a single to 1.6GHz
Since I'm technically challenged could someone correct me
The most important thing to understand is that the RC5 challenge is strictly CPU-limited. The processor fetches a few bytes, loops for an eternity over a few dozen instructions, fetches a few new bytes, etc. The distributed.net client just lets each CPU compute their own block of keys, so there is almost no communication between processors. Ergo, a dual 800 Mhz will be just as fast as a single 1.6 Ghz. And a single 1 Ghz will be just as fast as half the result you get with a dual 1Ghz.
As such, you can be quite creative and report the dual 800 Mhz result as the result of a 1.6 Ghz, divide the result from your dual 1Ghz by two to get the result for one CPU or calculate the result for the fabled 16 CPU 1 Ghz Xworkstation.
<strong>On the PC side, they are becoming more and more Mac like in the regard that more and more things are on the motherboard and can't be replaced. Most of these parts are inferior to the Mac equivelent and also slow the machine down considerably from whatever sky high numbers the processor produces. So you can have a 2 ghz $599 wintel box, but it will have integrated video, audio, and networking all lugging down that processor.</strong><hr></blockquote>
And you shouldn't forget that some of these components use the CPU for certain tasks. See the Win-modems that drain your CPU for instance.
[ 08-05-2002: Message edited by: wfzelle ]</p>
<strong>
I hope that's sarcasm. I have yet to see one benchmark in which the the winning Photoshop computer won but such a large margin.
x86 hardware IMO is simply not superior enough for me to want a switch personally. I would like to see another two years of Apple/IBM/Moto work before making a judgement.</strong><hr></blockquote>
One of the things Chris Cox said in one of the Forums about PS on the Mac was that it was partially slowed down because both PS and OSX are writing the screen to buffers at the same time and a fix was being made for this. No doubt Jaguar has influenced the work schedule
Thank you
<strong>G-News & wfzelle
Thank you</strong><hr></blockquote>
You're welcome. <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />
<strong>
All the prices need lowering. It's better to sell 2 Macs at $2500 than 1 at $3500. Apple needs resellers like CompUSA and PCWorld on side too. I get fed up of seeing crashed Macs in stores.
</strong><hr></blockquote>
well, it all depends on what you mean by better...
if apple is looking to increase market share at all cost, then yes, it is better.
however, given the current margins (30%), the cost of a $3.5k machine is $2450. therefore, they'd be making $50 per machine, which would make it $100
for the two machines they sold given the price cut.
congratulations.
admitting the market share doubles, you've just found a way to divide apple's profits tenfold.
now, the one way to do this would be to cut the margin in half...
the same 3.5k machine would then cost $2975, but would it sell twice as much at that price point ? sorry, but i doubt it.
Hmm. For me. It still depends. Apple may consider it to get into certain markets. People who wouldn't otherwise consider a Mac.
They're only going to reach so many people with 'Switch' campaign. If they want to grow and grow. They may have to consider the big picture.
(Anyway, I'm sure Apple will not repeat IBM's mistake. Apple aren't IBM. Apple has OS 'X'. For the next year or so...Apple have got to concentrate everything on piling on pressure on XP. iApps crazy, push the QE envelope and Rendezvous/isync ideas and the switch campaign. They've got to show for multimedia they are king.)
This 'Dual boot' strat' may mean nothing more than bundling Virtual PC on a Power X style processor (ie a processor that(!) powerful...it feels you're funning a 1 gig Pentium in emulation...) I can't think if that many compelling software reasons to be 'PC' anyhow. The Mac kinda has the best of all apps now anyhow. Best OS...best 3D apps, cream of the games...the best Office app
"There really isn't much wrong with the Mac in areas which count on the balance sheet. We're not too hot in 3D but that could well be about to change if we see the 7470 next week."
Begin to change. That hardware there...yer talkin' about...is catch up. It'll be contempary until the Hammer hits early next year and then they had BETTER have some Rio/7500 juicin' stuff up their sleeve.
It'll be interesting to see the 1.4 G4 on the 'let the animal loose' mobo DDR etc. If it repairs those damning Lightwave and Photoshop benches...then I'll gladly eat my words
"Perception and Price are what needs changing, not Processor."
Don't agree. Both need changing. The G4 is long overdue for consumer range only badge.
"A 1.4Ghz G4 should help - in real terms that's about the same speed as a 2Ghz P4, add DDR and AltiVec and it looks pretty good. In truth email and Word are fine on a 500Mhz chip on any platform."
We'll see if benches back you up.
"All the prices need lowering. It's better to sell 2 Macs at $2500 than 1 at $3500."
Agreed. The last POWERMac I bought some years ago...cost me an arm and a leg. If Macs were cheaper...I'd have upgraded twice by now. As it is...expensive and old tech' equals long update cycle. And APPLE don't get the sales/market penetration they want. It's a downward cycle. Hopefully the news of Apple increasing their purchasing of Taiwan components to make their products price competitive is true. ie goes on cheaper Macs not higher margins.
"Apple needs resellers like CompUSA and PCWorld on side too. I get fed up of seeing crashed Macs in stores."
Agreed. But Apple doesn't help themselves here in the Uk. Remember when Tescos sold old Crt imacs in stores. Apple didn't approve of these cheap grey market Macs. But hey...they were reaching markets and people they wouldn't normally reach. Their marketing is still too aloof. I'm surprised they're still in business.
"Finally Apple needs to buy a PC games house so that the best games are on Mac first. They need to ensure affordable after-market graphics cards are out there too."
Agreed. Yeesh. Put a bid in for id games. I'd love to see PC uproar about that in light of M$ buying Bungie! Buy the makers of Unreal Tournement. Don't have to buy many games companies. Just one or two, the ones that matter. And cut PC support!
"These things would offer far more ROI than going up against MS as a PC builder."
Over $4 billion. They could use some of that to secure their future as a software company. A merger with Adobe. Buy out Quark. Buy Maya.
Then Apple are in control. Doesn't matter a shat which processor they run anything on then!
Lemon Bon Bon
PS. Nothing wrong with irrational opinions. We're on a rumours board.
Sure some people will figure out how to hack it to run on some PCs, but these people will be running the OS illegally and it won't have solid and/or fast drivers to support it. This will not amount to a significant number of people, and no corporation will do it and risk the wrath of Apple Legal... not to mention the complete lack of software support. If a few home users do it whom wouldn't have bought a Mac anyhow, then all they are doing is potentially expanding the Mac software market a little (but probably not since they are likely to be pirates anyhow). These hacks are not enough of a market to worry about, and they aren't the sort of market from which money can be extracted.
The PowerPC line of Macs will continue, and I wouldn't be surprised to see it thrive over time due to the backing and R&D work of IBM. This puts Apple in the rather nice position of being able to leverage the best of both worlds -- Intel & AMD seem to be gambling on pushing faster and faster clock rates, whereas IBM is pushing on more and more cores. MacOSX is pretty well position to leverage both. Take your pick... 1 50GHz Pentium9, or a stack of 8 quad core 2 GHz PowerPC G7s. :eek:
I could also see one or two MacOSX licensees whom Apple brings on board with specific and limited contracts to simply to be secondary suppliers for x86 and PowerPC hardware on which to run Apple's OS. This is important in a bunch of business and government markets that Apple no doubt wants to be in. These tend to be high margin environments which emphasize solutions and support, so by choosing appropriate partners (e.g. IBM) Apple doesn't have to worry about them suddenly jumping into Apple's traditional markets and doing what PowerComputing et al did.
This does explain Jobs' comment fairly well, but if you believe it then you have to give Moki more credibility and we know how much he hates that!
<strong>Exactly the same way IBM's BIOS was reverse engineered by Compaq. Took Compaq a couple months and a couple dollars, but it is all quite legal to do. Being able to undercut Apple HW prices by imitating motherboard designs and setting in a reingineered BIOS ROM without having to do the basic R&D would be ridiculously easy for an HP or DELL (even though DELL doesn't really do R&D). There are a few quite well defined legal wickets a company would need to comply with along the way, but it is an age old and legally acceptable process.</strong><hr></blockquote>
I think this lesson has been learned in the industry and Apple wouldn't go there unless they could truly prevent it. No PC manufacturer is going to bother with Apple waving its legal stick around -- its not worth the relatively small market share they are likely to gain. Remember, Compaq did it to get into IBM's market and IBM was the giant.
How difficult would it be to put AltiVec on a DSP or coprocessor?
If that's doable, combined with RapidIO or HT, then I would think that it wouldn't matter what brand of CPU was used.
I really don't think that Apple would go the x86 route, but this would open possibilities for pretty much anything (as long as it was RIO/HT enabled). For that matter, if we're gonna talk about using POWER*, then why not infiniband (I think that's right, no?)?
As an aside, I've been wondering lately about the possiblity of a secondary on-board graphics chip for doing the OS-level stuff (window redraws, etc), freeing up the main GPU for 3D calculations and such. How possible would this be? Would this increase speed more with Jaguar?
[quote]
Originally posted by Blackcat:
<strong>All the prices need lowering. It's better to sell 2 Macs at $2500 than 1 at $3500.</strong>
and Originally posted by mattbr:
<strong>given the current margins (30%), the cost of a $3.5k machine is $2450. therefore, they'd be making $50 per machine</strong>
and Originally posted by Lemon Bon Bon:
<strong>Agreed. The last POWERMac I bought some years ago...cost me an arm and a leg. If Macs were cheaper...I'd have upgraded twice by now. As it is...expensive and old tech' equals long update cycle. And APPLE don't get the sales/market penetration they want. It's a downward cycle. Hopefully the news of Apple increasing their purchasing of Taiwan components to make their products price competitive is true. ie goes on cheaper Macs not higher margins.</strong>
<hr></blockquote>
a high gross margin (1 - (cost of goods / revenues)) does not necessarily indicate that a company could cut prices and remain profitable. in apple's case, their 28% margin drops to 12% when you amortize r&d expenses, and drops to under 2% when you include marketing and administrative expenses. using these numbers (assuming a flat distribution of costs and profits across all product lines), by reducing the price of a $3500 mac to $2500, apple would be losing over $940.
as a point of reference, dell has a gross margin of 17% which drops to 16% after r&d and to just under 7.5% after other operating expenses.
the 1600 machine is simply a dual 800.
the point of Apple going x86 while still offering kick ass ppcs is as moki mentionned: some corporations *require* computers to have intel chips or to be able to run windows.
<strong>Ok, I'll ask again.
How difficult would it be to put AltiVec on a DSP or coprocessor?
If that's doable, combined with RapidIO or HT, then I would think that it wouldn't matter what brand of CPU was used.
I really don't think that Apple would go the x86 route, but this would open possibilities for pretty much anything (as long as it was RIO/HT enabled). For that matter, if we're gonna talk about using POWER*, then why not infiniband (I think that's right, no?)?
As an aside, I've been wondering lately about the possiblity of a secondary on-board graphics chip for doing the OS-level stuff (window redraws, etc), freeing up the main GPU for 3D calculations and such. How possible would this be? Would this increase speed more with Jaguar?</strong><hr></blockquote>
altivec is a dsp already so why do you want to put it on another dsp? putting multiple cores into one die has been no secret and motorola's embedded mpc 8260 is having two core inside one die: one is central 603e core and another one is 32-bit risc core to handle communication interface. the interconnect is through local bus. i heard in their new serial mpc85xx, they are putting an internal crossbar switch between its core cpu and its peripheral cpu.
for graphical intensive application, faster cpu can not solve everything. take a look at an sgi machine. it has darn slow cpu, but i can gauarantee that the low end machine from sgi could easily beat any top speed pc or mac in many aspects.
there is nothing new on how sgi did it. and i am sure there might be new way to do better. it matters whether apple wants to use it or not. there are tons of smart guys in apple, so i am pissed for the fact that they have not yet used any to improve a mac.
i guess sj is correct that apple has to put os x out first and then focus on hardware. this comment is an insulting as well as enlightment. it is kind bad that apple kept using old technology to lure money from their loyal customers, but they need money to survive. now i think apple will focus on the hardware or more grand idea on how to ride on os x. but the bad thing is that it means that i will have to wait another year for a power new macintosh.
i will wait...
<strong>As an aside, I've been wondering lately about the possiblity of a secondary on-board graphics chip for doing the OS-level stuff (window redraws, etc), freeing up the main GPU for 3D calculations and such. How possible would this be? Would this increase speed more with Jaguar?</strong><hr></blockquote>
That ties up really nicely with rumour that Apple might usee Nforce2 chipset.
Part of it can include an on motherboard Gforce4 MX, which would make every Nforce2 based Mac Quartz Extreme capable regardless of the graphics card you add (obviously adding a non OpenGL card would mess things up) so people could upgrade to better cards or dual monitor support more easily.
Quartz Extreme is another reason why PPC is less required too I suppose.
<strong>
All the prices need lowering. It's better to sell 2 Macs at $2500 than 1 at $3500.</strong><hr></blockquote>
This is a misperception I see repeated over and over. Selling two Macs at $2500 does bring in $5000, which is more than $3500, but what if Apple's cost on each Mac is $2300? Selling two of those at $2500 gives them $400 profit while selling one of them at $3500 gives them $1200 profit. Business isn't about the greatest possible gross sales revenue, it's about profits.