Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
~loserman~ said:
Here is a funny tale....
We recently talked to IBM and they told us they will be shipping up to 2.5Ghz dual core 970's by January in their JS20 blade servers.
And what percentage of these new chips were being made available to Apple? IBM seems to focus on IBM and their own schedule of things. I love the PPC, but IBM has failed to deliver. I withhold judgement on Intel vs. PPC until I see a Mac running on Intel vs. a Mac running on PPC and get some side by side comparisons (without emulation).
 
ender78 said:
I believe that the margin that Apple enjoys today will diminish.
Apple's margins are significantly smaller than Dell's and Dell sells a much larger volume than Apple. How much smaller of a margin should Apple take to try to gain marketshare. Apple beats the competition and people are willing to pay a bit more for that. If prices drop with the switch ... great!! But if not, people will still buy Apple's better product (design, OSX, etc), even at a premium.

ender78 said:
Apple looses nothing as there is just no way that I would have bought a new machine at that point. Apple will reap the benefits of consistent sales [a Mac user will still buy a machine every X years] as well as some incremental updates. As it stands today, what can we really do with our machines?
Nothing? If people are able to upgrade their Apple computers with new motherboards / chips from Intel, Apple loses the ability to sell updated and upgraded computers and has to leave many of those option to other vendors like Intel, etc. Normal users buy a new Mac every X years ... but if upgrades and swappable motherboards are available, in the interim, users upgrade every X/2 to X years, and buy new every 2X or 3X years. Upgrading is cheaper than buying new. Intel benefits, Apple loses. JMO
 
Mechcozmo said:
Linkety to a post a few minutes ago by me. Explains this entire BIOS, Open Firmware mess.

So the fact that you can't play towers of Hanoi on the bios explains this "mess"?

I interact with neither bios nor OF on either of my computers on a daily, nigh monthly basis. The computational ability of my bios is of little to no importance to me.

How about something useful? ??? Does OF let me adjust my FSB? Does OF let me adjust my memory timing and latency?

I would FAR prefer something that allows me to USE my computer in a better / faster manner while in the OS than something that allows me to play tetris on bootup?!
 
Mr Maui said:
Apple's margins are significantly smaller than Dell's and Dell sells a much larger volume than Apple. How much smaller of a margin should Apple take to try to gain marketshare. Apple beats the competition and people are willing to pay a bit more for that. If prices drop with the switch ... great!! But if not, people will still buy Apple's better product (design, OSX, etc), even at a premium.


Nothing? If people are able to upgrade their Apple computers with new motherboards / chips from Intel, Apple loses the ability to sell updated and upgraded computers and has to leave many of those option to other vendors like Intel, etc. Normal users buy a new Mac every X years ... but if upgrades and swappable motherboards are available, in the interim, users upgrade every X/2 to X years, and buy new every 2X or 3X years. Upgrading is cheaper than buying new. Intel benefits, Apple loses. JMO

Dells margins are signifigantly less than Apples. Are you mistaking "Margin" with "Market share"?

The margin is the amount made per computer sold. Dell is in the cutthroat business. Apple is not. Apple makes a healthy chunk on every item sold. Dell slashes prices, and gets the volume.

I believe that in fact the mac mini was a "test" of sorts to see how apple could perform in the commodity hardware market. Push out a machine with lower than normal margin, and see if they could make it up with higher sales. I believe this test was a great success, and apple is beginning to comprehend that they could cater to both ends of the spectrum, with high margin servers / workstations, and high volume but low margin consumer systems. Increasing marketshare, profits, and program base all at the same time.
 
I do not thnink DRM ist the point. IBM will make chips for XBOX360 and PS3, so they can make chips with DRM (even for a PPC based platform). Apple could have DRM with IBM if they wanted to.

Apple has never taken that to serious, look at DRM in ITMS (you can get unprotected copies with iMove) or iPod (you can copy songs back to a computer quite easily) and they also have ITMS on Windows ( MS does not allow DRMed WMAs to be played back on MAc because they "have no controll over the OS"...).
When they introduce HD-DVD (or Blueray) to the Mac they will have to use some DRM, but I do not think they will do anything more than absolutely necessary.
This is because open source is important to the Mac. OSX would simply not exist without.

Christian

spinko said:
(...) they (IBM) surely didn't invest huge amounts of $'s in a factory if that wasn't the ultimate goal .. I must be missing something... (built-in DRM ?)
 
IBM isn't making the Xbox 360 chips

Mr Maui said:
IBM is interested in producing lots of chips for Microsoft, Sony and Nintendo for the gameboxes and not producing newer, better, or additional chips for Apple.

Note that TSMC will be fabbing the triple-core chips for the Xbox 360 - not IBM and therefore not at Fishkill.
 
don't forget about QA and support

Fukui said:
I dont see why not. All most small-time devs need to do is click a check box...

If they don't care about quality and support, that's all they need to do.

If they do test the product before shipping, then they'll have twice as much testing to do. If they support the product after shipping, then they'll have additional costs to train support personnel and track the two versions of the app.

Note that this will be hardest for the Mac-only apps - those are the most likely to have subtle, hard to debug issues due to code that assumes big-endian CPUs.

It's very misleading to say that "it took two hours to port Mathematica" - this is an app that already runs on Intel, and runs on a large number of little and big-endian platforms. Multi-platform apps with years of development have already hit and solved these bugs. It also ignores Wolfram's QA time...
 
AidenShaw said:
Note that TSMC will be fabbing the triple-core chips for the Xbox 360 - not IBM and therefore not at Fishkill.

This is a VERY important point everyone here is missing out on. IBM will NOT be manufacturing the chips for most of the next gen consoles. They helped design the CPU's, however the console manufacturers OWN the chip designs, not IBM, and can go elsewhere for less expensive chip fab. THIS was the key reason Microsoft switched from intel... Now they can shop the chip anytime they like for a cheaper fab. Before, Intel owned the chip and design, and Microsoft was tied to them.
 
Mr Maui said:
Nothing? If people are able to upgrade their Apple computers with new motherboards / chips from Intel, Apple loses the ability to sell updated and upgraded computers and has to leave many of those option to other vendors like Intel, etc. Normal users buy a new Mac every X years ... but if upgrades and swappable motherboards are available, in the interim, users upgrade every X/2 to X years, and buy new every 2X or 3X years. Upgrading is cheaper than buying new. Intel benefits, Apple loses. JMO

Mac systems have much longer longevity than Wintel machines. While most of the people on this board are using a machine at least 2-3 revisions old, such a concept would be unheard of in the Wintel world. If I spend $500 on upgrades every year, it will in no way diminish my need to buy a new machine in 3 years, I wasnt going to spend ANYTHING in the interim, I'm confident that others are the same.
 
A year to go and all this talk. Love it.

I guess we finally know what the "i' has really stood for, eh?

jk of course.
 
~loserman~ said:
My Dual and Quad Opteron boxes scale memory bandwidth almost perfectly.(with NUMA)

By the way all my programs are single threaded CFD codes and we run all our Opteron Clusters with LINUX using a NUMA kernel.
Since you have an embarrassingly parallel application with a NUMA-aware kernel, then the Opteron architecture is a good match for you.

On the other hand, if you had a multi-threaded application with more threads than CPUs per NUMA node - then you would definitely have a memory scaling issue.


~loserman~ said:
The AMD [dual-core] cpus unlike the INTELS at least don't have to exit the die and go to the FSB to see ea others memory. They have an internal crossbar between the cores besides hypertransport.
This is true for the first Intel dual-core chips (the Pentium D and Pentium EE). Logically, they are almost equivalent to a dual single-core system (except cheaper, since the IPC doesn't have to go through the Northbridge).

Intel has at least another dozen multi-core chips under development, however, so don't believe that the next generation won't trump AMD.

And, BTW, you realize that the dual-core Yonah has a large shared cache (similar to the IBM POWER4 dual core), right? So, the chips that Apple is most likely to use will have on-die IPC.


~loserman~ said:
A dual dual core XEON based system will have to share the memory with all four cores. Talk about memory starvation.

At least with AMD's offering they will get close to twice the bandwidth.
On the other hand, the Xeon-based system won't suffer from NUMA issues and the high-latencies of the cHT links.

Just because on paper you can take the sum of the memory controller bandwidths doesn't mean that real programs will see the sum. AMD's architecture has real latency issues.

But, it's a great match for your embarrassingly parallel application, so it's what you should buy!

I wonder if the Steve's seen Intel's next generation memory controller plans, and if he left a little puddle of drool on the floor in Santa Clara.... ;)
 
Rhapsody was actually supported on x86...

Does anyone remember that Apple Rhapsody was shipped to developers with Intel x86 support in the original Developer Release and Developer Release 2? That just continued the Intel x86 support that was in NEXTSTEP.

pre-Apple acquisition:

NeXTstep 0.8 - NEXTSTEP 2.0: Motorola 68K support
NEXTSTEP 3.1 - 3.2: Motorola 68K/Intel x86 support
NEXTSTEP 3.3: Motorola 68K/Intel x86/Sun SPARC/HP PA-RISC support
OPENSTEP for Mach 4.0 - 4.2: Motorola 68K/Intel x86/Sun SPARC/HP PA-RISC support

post-Apple acquisition:

Rhapsody 5.0 (DR) - 5.1 (DR2): Intel x86/PowerPC support
Rhapsody 5.2 (1.0) - Rhapsody 5.6 (OS X Server 1.2v3): PowerPC support (secret Intel x86 support?)
OS X 10.0 - 10.3: PowerPC support (secret Intel x86 support per Steve Jobs)
OS X 10.4: PowerPC/Intel x86 support

Sounds to me like the "secret double life" was just that Intel x86 support was simply never released to the public since Rhapsody 5.2, but it was not really dropped. I wonder if Motorola 68K/Sun SPARC/HP PA-RISC support lives on as other "secret lives"?

Screenshot of Apple Rhapsody 5.1 (DR2) running on Intel Pentium from Nathan's Toasty Technology GUI Gallery's Apple Rhapsody page.

More interestingly, maybe there are other unknown "secret lives"? Without having to invent anything, what if they've also simply kept up Apple's Rhapsody Yellow Box for Windows? That used Microsoft Windows instead of Mach/Darwin for its kernel and other services. What if there's another "secret life" of OS X that's OS X (Cocoa and maybe more) running on a Microsoft Windows XP kernel "just in case"?

Screenshots of Apple Yellow Box for Windows/Rhapsody 5.1 (DR2): About Apple Software and Preview running on Windows XP from this German NEXT info site.
 
Mr. Zarniwoop said:
Does anyone remember that Apple Rhapsody was shipped to developers with Intel x86 support in the original Developer Release and Developer Release 2? That just continued the Intel x86 support that was in NEXTSTEP.

pre-Apple acquisition:

NeXTstep 0.8 - NEXTSTEP 2.0: Motorola 68K support
NEXTSTEP 3.1 - 3.2: Motorola 68K/Intel x86 support
NEXTSTEP 3.3: Motorola 68K/Intel x86/Sun SPARC/HP PA-RISC support
OPENSTEP for Mach 4.0 - 4.2: Motorola 68K/Intel x86/Sun SPARC/HP PA-RISC support

post-Apple acquisition:

Rhapsody 5.0 (DR) - 5.1 (DR2): Intel x86/PowerPC support
Rhapsody 5.2 (1.0) - Rhapsody 5.6 (OS X Server 1.2v3): PowerPC support (secret Intel x86 support?)
OS X 10.0 - 10.3: PowerPC support (secret Intel x86 support per Steve Jobs)
OS X 10.4: PowerPC/Intel x86 support

Sounds to me like the "secret double life" was just that Intel x86 support was simply never released to the public since Rhapsody 5.2, but it was not really dropped. I wonder if Motorola 68K/Sun SPARC/HP PA-RISC support lives on as other "secret lives"?

Screenshot of Apple Rhapsody 5.1 (DR2) running on Intel Pentium from Nathan's Toasty Technology GUI Gallery's Apple Rhapsody page.

More interestingly, maybe there are other unknown "secret lives"? Without having to invent anything, what if they've also simply kept up Apple's Rhapsody Yellow Box for Windows? That used Microsoft Windows instead of Mach/Darwin for its kernel and other services. What if there's another "secret life" of OS X that's OS X (Cocoa and maybe more) running on a Microsoft Windows XP kernel "just in case"?

Screenshots of Apple Yellow Box for Windows/Rhapsody 5.1 (DR2): About Apple Software and Preview running on Windows XP from this German NEXT info site.
Thank God! Someone with some sense about the whole thing! I wish your post went to every other thread on this board to quell some of the naysayers. Kudos to you!

This should be the easiest transition yet! Most people can't figure that out! :)
 
Apple's Strategic Competitive Advantage

skellener said:
Thank God! Someone with some sense about the whole thing! I wish your post went to every other thread on this board to quell some of the naysayers. Kudos to you!

This should be the easiest transition yet! Most people can't figure that out! :)
Thanks. I guess I don't see this transition as big a deal as others do, since OS X (formerly Rhapsody/OPENSTEP/NEXTSTEP) ran on Intel x86 long before it ran on PowerPC. I just wanted to remind everyone of that.

In terms of Apple's business, I think some people see the transition as a big deal if they think Apple's strategic competitive advantage lies deep in their low-level hardware designs. Basically, Intel CPUs don't turn their geek knobs as much as PowerPC CPUs do.

I happen to think they're wrong about why Apple's been successful. I'm pretty sure Apple's strategic competitive advantage is the end-user experience, which is hard for other people to sustainably copy. Their customers go to Apple stores because of the nicer experience, it's really different than the pages of low-level hardware specs you get on Dell's website or in a Best Buy computer department. They buy iPods in all shapes and sizes because of the experience of how they work, they're different than the "technically superior" and less expensive MP3 players that focus on low-level specifications and features. They buy Macs across form factors because of the experience of how OS X "sings" coupled with the Apple design/look and feel, they're different from faster/cheaper/"better" beige boxes that suffer from overcomplexity in the OS requiring end-users to be really familiar with rebooting and reinstallation. They buy music from iTunes Music Store because of the buying and playing experience, it's different from the "cheaper" or even free (but illegal) alternatives that just aren't as simple, easy, and elegant to use.

Apple's success does not hinge on down-deep technological superiority. They never really had "better" technology, except in a few niche areas that were trumpeted loudly in Apple's marketing. As a matter of fact, they were usually behind when competitively benchmarked or compared strictly by feature/function on most of their technology products and probably still are behind today. You can buy faster/"better" alternatives to any Mac, any iPod, or songs from iTMS. But people still buy the Apple products, and love them, in real emotional ways because of the experience, the design, the brand. And based on what I saw in the WWDC keynote, that's not going to change.

I think the Developer Transition Systems are also an indicator of this. They're not about the low-level hardware details. Without really digging around, like taking the cover off or holding down strange key combinations while booting, you can't tell what CPU they run. You just know they're Macs, because of the Mac experience: they look like a Mac, and feel like like a Mac. I personally can't wait until Apple product engineers and designers get their turn at making the new Macs, I bet it'll be the same Mac experience again where people flock to stores to see them, buy them, and take them home.

And most Apple customers will have no idea what kind of CPU is inside.
 
AidenShaw said:
Note that this will be hardest for the Mac-only apps - those are the most likely to have subtle, hard to debug issues due to code that assumes big-endian CPUs.
I think maybe we mean different things
by small-time developers.
I'm talking about those developers that
are building new apps using cocoa and carbon,
not those still using the toolbox API.

If they are saving thier files using keyed archiving
(possible from both) then there's not gonna be
much endian issues. Most small time devs anyways
aren't gonna be dealing with assembly and endian
switching I would think. Images saved in different
endian formats should be handled my NSImage
automatically. Theres already a couple apps being
delivered in Universal formats! And they arent even
shipping Intel systems yet!
 
Mr. Zarniwoop said:
More interestingly, maybe there are other unknown "secret lives"? Without having to invent anything, what if they've also simply kept up Apple's Rhapsody Yellow Box for Windows? That used Microsoft Windows instead of Mach/Darwin for its kernel and other services.

Actually, YellowBox on Windows included processes that duplicated some of the expected Mach services which Windows lacks. I'm pretty sure it also had a WindowServer process to handle the interpreting of display postscript.

It wasn't just an unreleased part of Rhapsody, either. Before it was "Yellowbox" it was just OpenStep for NT, a commercial product. I worked at a big bank for 3 years in the late 90s, where we moved a trading system over from OpenStep on Mach to OpenStep on NT, which let the trading system be used via Citrix.

I believe it was also used for the WebObjects development tools on NT, though I think those changed to Java at some point.
 
anyone who has paid the slightest bit of attention to apple's innovation in design would realize that creating their own custom motherboards is an ESSENTIAL practice for them and i don't see them changing that one bit.

to those who theorize the power macs will shrink because the dev kit motherboard is so much smaller.... have you been paying attention to anything?! the current power mac g5 is not huge because they couldn't design a small motherboard. it's huge because they designed it to be that way, with independent thermal zones and all that. the motherboard adapts such that everything is exactly where apple needs it to be in conjunction with its case design.

you guys are just like music fans who talk trash on the "new" album of your favorite artist, until the next one comes out and then that last one becomes amazing. get a grip. you are all paranoid and instigating nonsense. you honestly all sound like windows users. we're not even supposed to have to care about all this crap. apple will come through with great products. so... please, shut up.
 
Mr. Zarniwoop said:
I wonder if Motorola 68K/Sun SPARC/HP PA-RISC support lives on as other "secret lives"?

The 68K is a dead processor for the desktop market, so I'd guess no.

Sun uses x86 chips now...

HP wanted to use Itaniums, the Itaniums never appeared, and so I'm pretty sure they don't use PA-RISC or, for that matter, anything anymore.
 
BillHarrison said:
Dells margins are signifigantly less than Apples. Are you mistaking "Margin" with "Market share"?

The margin is the amount made per computer sold. Dell is in the cutthroat business. Apple is not. Apple makes a healthy chunk on every item sold. Dell slashes prices, and gets the volume.
Beyond "Margin" vs "Market share", there is overall profitability on revenue.

Dell is more profitable than Apple for the same revenue. I'm sure I read some results which said Apple made 30% on their computers (30% margin, vs dell's 10%), but Apple's profit was 0.5%. This was because Apple spends a LOT more than Dell on R&D and OSX.

If you consider Apple a hardware company then OSX is an unnecessary cost burden. If you consider Apple a software company then the software makes a small profit (both from upgrades and from a share of 30% margins), and the hardware is a 0% margin business (of course though, if they sell 50% more Macs the software development & R&D costs are unchanged so the profit increases hugely).

In the end though, it's all about the Mac experience. ANY changes will have some type of effect (good or bad)
 
steeldrivingjon said:
It wasn't just an unreleased part of Rhapsody, either. Before it was "Yellowbox" it was just OpenStep for NT, a commercial product. I worked at a big bank for 3 years in the late 90s, where we moved a trading system over from OpenStep on Mach to OpenStep on NT, which let the trading system be used via Citrix.

I believe it was also used for the WebObjects development tools on NT, though I think those changed to Java at some point.

Yellow Box for Windows was an application environment based on OpenStep Enterprise for Windows.

While in developer release form, there were fewer differences between OpenStep and Yellow Box for Windows than there were between the OPENSTEP for Mach and Yellow Box in Rhapsody application environment.

Consequently, OpenStep Enterprise for Windows applications could (more often than not) run unmodified in the Yellow Box for Windows developer release. OPENSTEP applications were not runnable at all by the time that Rhapsody 5.1 (Developer Release 2) was out.

Yellow Box for Windows was the foundation of the WebObjects 4.x.x releases for Windows.

Any applications made for Yellow Box for Windows would run just fine on a Windows system with WebObjects 4.x.x installed.

As for what Windows systems Yellow Box can run on... I've successfully installed the environment on both Windows NT 4.0 and Windows 2000 Pro.

Mr. Zarniwoop said:
Does anyone remember that Apple Rhapsody was shipped to developers with Intel x86 support in the original Developer Release and Developer Release 2? That just continued the Intel x86 support that was in NEXTSTEP.
Yeah, I'm pretty sure I've heard of Rhapsody somewhere. ;)
 
Quartz Extreme said:
Are there many viruses on Linux for x86?
How about for BeOS, OS/2, Rhapsody, Darwin, UNIX?
In the 5 years that I've been running Rhapsody on my IBM ThinkPad (using it as my primary mobile system) I've never run across any type of virus.

As I'm one of the only people left in the world still running Rhapsody as a daily use OS, I doubt that I'm that much of a target market for malicious software. :p
 
OS X Intel Leak

CrazySteve said:
If this x86 version of OSX leaks, you will have every wintel person
running OSX 2 years before the new MACs come out. :mad:


I would suspect part of it has been leaked already, like up to 4-5 years ago.. :p Minus the pretty interface..

-Don

http://www.opensource.apple.com/darwinsource/

Darwin 8.0.1 Installer CD
The Darwin 8.0.1 Installer CD is available. Darwin 8.0.1 corresponds to the open source core of Mac OS X 10.4 and is available at the following URLs:
For PowerPC:
http://www.opensource.apple.com/darwinsource/images/darwinppc-801.cdr.gz
http://www.opendarwin.org/downloads/8.0.1/darwinppc-801.cdr.gz
MD5 (darwinppc-801.cdr.gz) = fe85def148896f76b00a753687d99144

For x86:
http://www.opensource.apple.com/darwinsource/images/darwinx86-801.iso.gz
http://www.opendarwin.org/downloads/8.0.1/darwinx86-801.iso.gz
MD5 (darwinx86-801.iso.gz) = dbd260dda994093a11c31afbe624aa34

The source code for Darwin 8.0.1 is available via the web. For more information, please see the Release Notes.

Come on what's so super secret, you folks are running a powerpc version of Darwin under the pretty mac interface with a lot of mac services.. I don't think Jobs ever made it a secret there was in intel version of this core :) It's been on Apple's Open Source site for a long time, so if you had a "Darwin" App you could run it under OS X or Darwin x86 with a recompile.. Hey wait how many apps is that?

MacOS X is really Unix. http://www.apple.com/macosx/features/unix/

This really shouldn't be a suprise to ANYONE...

There already is an Intel Darwin cult out there check out..

http://www.opendarwin.org/
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.