Slashdot Mirror
Apple Hedges Its Bet on New Intel Chips
Corrado writes "The Mercury News is reporting that Apple is still planning to use PowerPC chips well into 2008 for its low end and portable systems. Does this increase the "warm fuzzes" for the Intel move?
More information from TheStreet and lots more links from Google News."
Wasn't the one of the points of the switch to make more efficient use of electricity? Then why are they going to continue to use PPCs in their portable devices?
Has anyone ever heard of support? Apple may need the occasional extra lot of processors for years to come to support their existing support contracts.
That what was all this school was for... to teach us how to solve our own problems. -- janeowit
Steve said low end items were the first to go (mini, iBook, etc). What the hell are they talking about.
The Doormat
If you're not outraged, then you're not paying attention.
Apple is just covering their bases. This is merely a safeguard, not an indication that the move is behind schedule.
This doesn't help the "warm fuzzes" at all. It just shows that apple is not as confident as they originally were. If Apple truely believed that the Intel procs they were switching to were so much better this point would be mute.
Why doesn't apple just continue to use both architectures?
This has nothing to do with the change over to Intel. Apple needs to support the warranty its existing base of G4 Macs for at least three years.
Entirely outside the ADC NDA...
If you take a look at Apple's developer tools - specifically, XCode 2.1 and above, you'll find that building binaries for both platforms is fairly easy. I think that Apple not only wanted to allow developers to build binaries for Intel and PPC, but to allow itself some time for the transition. Apple hasn't locked itself into a position where it must switch to Intel on a certain date. This is a good thing.
Really, if we can consider Mac OS X as simply OpenStep 4 (or whatever), then the CPU - to a very large extant - becomes just another part of the machine. With the exception of low level hardware driver experts, do you really care what bridge / Firewire / USB chip is used? Think the same way about the CPU, and you have Apple's apparent perspective on using Intel chips - the OS is fairly independant from the CPU, the developer tools can target multiple platforms, and consumers really won't have too much to worry about.
/* Dang, I can't type that well. */
So presumably we will again have an extended transitional period where Mac binaries have x86 and PPC code rolled together. I wonder how big an OS X Office install will be now.
I am becoming gerund, destroyer of verbs.
Will this be a problem for support? They'll now have to support this much more hardware, and will have to have a fork of their OS X code; or will all code be done for Intel procs now, and 'just work (tm)' via the rosetta on powerpc procs? I think they'd have to do this, but still, I think it's going to taint the marketing a bit. Still, this hasn't been done before, and it's in sits like this that Apple usually does well. As long as production can keep up...
bad_outlook
--
Is this vague enough for you?
Apple have stated that the low end will switch to Intel first, so I don't really know what the basis for this `story' is. It seems much more likely that, if they are extending their purchasing options for G4s to 2008, they will stop selling G4s at the end of this year. This would then give them a supply of G4s to use in replacements until the end of the 3-year AppleCare period for the last G4 units sold.
I am TheRaven on Soylent News
Good, because I just plain don't want an Intel chip. I don't care if it's the future of Apple's support, I want to keep buying PPC as long as I possibly can. I don't care if you think I'm crazy or stupid. Personally I just have a whole bunch of personal Altivec code and I don't want to have to rewrite it.
Posted as AC because every time I express that I do not want to be forced to use Intel chips in order to continue using OS X, I get screamed at for being a "zealot". I find it a bit funny that disagreeing with Apple gets you branded as an Apple Zealot now, but there it is.
Apple did say the transition would take a couple of years.
It's obvious that after the Intel bomb shell they dropped that they now need to secure supplies of current processors until the entire line has moved onto Intel.
This is common sense. (But you don't have to expect this from news sites that report, even, that Apple might be back-tracking on the Intel switch.)
Google Link and just click on the url that shows up on that page. Worked for me :)
I wish my lawn was emo, so it would cut itself.
Of course, but you're missing the point.
This is a perfect opportunity to blather on with uninformed speculation, rumors, pipe dreams of dual core Antaries laptops, etc. . .
Intel has made their whole company strategy around low power high performance chips. That was the stated reason for moving to Intel. Therefore, I would think that Portables would be the first thing to move. For example, there is no G5 portable and we've been waiting a long time for a portable with a better chip than the G4. My powerBook is getting long in the tooth, but I won't replace it with another G4 powerBook - what would be the point? - unless it quit working.
The G5 desktops are still very fast and I could see staying with PPC there for a while, but not on portables.
Avoid Missing Ball for High Score
When an auto manufacturer ships a new engine, they don't immediately halt production of the old ones that it is destined to supplant. A phased transition is simply a reality of the manufacturing business.
Apple doesn't have to rush out an entire new line of units in one big bang. Good engineering and facility planning take time.
Umm, for like no money, they've secured access to a chip. If they hadn't, they'd be liable to lawsuits for not securing chips in case there is a problem with the transition. CYA material, plain and simple.
Apple has to support current computers with their Applecare program. Applesinsider has discussed that these go into 2008. So really, this is probably nothing more than the winding down period.
http://www.appleinsider.com/article.php?id=1248
But with PPCs in portables til 2008, that means software will probably support PPC into 2010, if not longer.
The Aztecs predicted the end of the world to happen on December 12, 2012. Seven years out of a mac portable? I'm satisfied.
eric http://www.ericdfields.com/
No, Apple does not plan to keep using G4s in the lower-end stuff. It'd make no sense. Apple is likely securing this contract so they have a supply of G4 chips for product repairs for the next three years, as AppleCare is a 3-year agreement.
Its amazing how most news articles will not give you the full story, or worse, you get their slanted version of events.
Reading multiple articles (not something /.'ers are likely to do since we can't even get them to RTFA) lets you get all the facts so you can draw your own conclusions.
Just my 2 pennies
[Fuck Beta]
o0t!
Heat? What the hell
The number one reason you havent seen a G5 laptop is heat issues. I don't see any problems running newer and newer x86 CPUs in laptops.
Hell the G5 towers need to be *water* cooled.
Furthermore, while the CISC/RISC business is correct every single report Ive read about the dev OSX86 machines (which are just regular P4s) are that they boot faster, perform faster, and are overall considerably faster than a G5.
Drop the argument, even Apple realizes it's dead.
Clarus says Moof!
The Pentium M is lower voltage and has a lower power dissipation than the current line of mobile G4s. I too wish the PowerPC would continue, because it seems like a really elegant architecture, but Intel currently has the PowerPC beat in terms of mobile efficiency.
PPC is one of the best platforms ever for both sound, gfx and of course heat production.
Is this based on more than just personal feelings about the architecture? Honestly, I like the PPC. It's a great implementation of the classic RISC principles: lots of registers, simple and fast instructions, no hardware stack, etc. But, really, this is just geeky fawning over a pretty design. The x86 is certainly ugly in some ways, mostly in terms of the huge legacy instruction set, but it's not so bad overall. Having hardware stack support is very nice. The limited number of registers makes function call overhead very low. If you disassemble code for typical PPC applications, you may see dozens of instructions for entering and leaving a function. And with each of those instructions being 4 bytes, that's a big deal in terms of instruction cache usage. So it's not entirely clear that in the modern world a classic RISC architecture is better.
Apple is not only extending the line for compatibility. If the Cell proves to be a sucesefull processor i think Apple will probably look into IBM for a nice chip contract. This way, what Apple is trying to secretly doing is to decide wich chip will target the high segment and wich one will remains for lower budget machines. Just my 2cts
People may laugh when you say `Intel' and `Roadmap' together, but they don't laugh nearly as hard as when you say `Freescale' and `Roadmap' together. I remember these guys promising 3.6GHz+ G6 chips being ready by 2002 back in the late '90s. Next to them, Intel roadmaps look positively reliable...
I am TheRaven on Soylent News
This move makes sense from the standpoint of smoothing the transition to Intel chips. It may mean additional challenges for Apple's support personnel, but the bottom line is that Apple's typical customer doesn't care what chips are in the box as long as Macs act like Macs and iPods act like iPods, etc... Apple managed the transition to PPC pretty well, so there is a good chance that they'll be okay going to Intel, afterall, there isn't any big rush to do this that I've heard about.
To the making of books there is no end, so let's get started
That's not what the deal is. Apple isn't hedging its bets. From the various articles: " Apple Computer Inc. said it has an option to keep buying microprocessors from Freescale Semiconductor Inc., three months after saying Macs will switch to Intel Corp. chips next year." Apple has negotiated for an option to buy the processors, and there is no set number of processors they will buy. Freescale just agreed to supply whatever Apple orders for the next three years. Apple could order 0. Apple could order a bunch. But probably not. They're most likely buying these processors to support the existing user base, to produce enough existing models to meet existing contractual obligations, and to be able to meet any unexpected surges in demand during the transition period. Where do people get that Apple is "hedging its bets?" Apple has bet the farm on Mac on Intel. Believe it. And don't think that for one minute the underlying hardware makes no difference. There will always be certain types of applications that will very much be hardware dependent, and yes, it DOES make a difference.
My blog
lipo -remove i386
-- i am jack's amusing sig file
Was that not the entire point of switching? That Jobs was not pleased with the PPC options for portables?
And what does he mean by 'low end' ? Does that mean as better PPC chips come out ( like the G6 ) they will just be ignored?
This whole thing is rather confusing now.
Personally, im glad to see the PPC will still be around in Apple products for a while longer, but gheesh, lets make a decision and stick with it...
---- Booth was a patriot ----
"If you take a look at Apple's developer tools - specifically, XCode 2.1 and above, you'll find that building binaries for both platforms is fairly easy."
If you don't feel the need to do testing on both binaries...
I rarely criticize things I don't care about.
No thanks, if I want to see fat strippers I'll go here: http://www.bigburlesque.com/home.html
Your example is not a good one. I disassemble x86 code and see instructions pushing parameters onto the stack. I disassemble PPC code and see instructions moving parameters into registers. How good is x86 with functions that have a large amount of local storage and a large number of parameters?
My initial reaction to this was... Hmmm, maybe I can go ahead and make the plunge to buy that new G4 iBook
But has anyone thought that this may have only been done to generate that very reaction. I have been holding out on purchasing an Apple because of the upcoming Intel switch. I will, again, have to wait to see what Apple is going to do...
Note that the agreement does not require Apple to buy any addtional G4 chips
Apple is the best computer I almost bought!
Caveat Emptor, I say.
Well, this is not so much about Apple hedging its bets, as it is about timing. A lot of folks anticipate an announcment of new hardware at MWSF in January. It ain't gonna happen!
Having now seen last weeks Intel announcement, it makes me believe it is unlikely they will launch Intel based Macs with 32-bit processors. Both iMac, Pro Mac and XServe are already 64-bit and they will stay that way. Anything else would be seen as a complete failure by the market.
For the mini and portables, the picture is a bit more tricky. From what I can gather from the latest announcements from IBM and Freescale, what I think will happen is that Apple will introduce a mini with a dual core processor from IBM perhaps even in September at MacWorld Paris, and follow up with similar announcements for the rest of the product line. Exception is of course the portables where they for thermal reasons have to stay at G4 until the switch to Intel, hence the agreement with Freescale.
I have a little more detail about this in an article I wrote a couple of days ago.The future is in beta
As per Job's Keynote they will be doing a full move to Intel by the end of 2007. The chips are for WARRANTY computers. They have to have a supply to fix computers, NOT sell them.
> Hell the G5 towers need to be *water* cooled.
No they don't. Only the first gen 2.7GHz were water cooled, and when they got updated the water cooling was removed.
/* FUCK - The F-word is here so that you can grep for it */
last week I finally experienced what is speculated as one of the major reasons Apple is moving to Intel from PowerPC. Friend let my wife borrow his mac laptop so she could run something that was mac-only. That thing is HOT. Heated up the whole damn room. Was suprized at just how hot it actually got.
Your example is not a good one. I disassemble x86 code and see instructions pushing parameters onto the stack. I disassemble PPC code and see instructions moving parameters into registers. How good is x86 with functions that have a large amount of local storage and a large number of parameters?
You're talking about the caller. I'm talking about what happens in the function that you're calling. On the x86, the return address is pushed on the stack as part of the "call" instruction (and popped as part of "ret'). On the PPC, both of these translate into multiple instructions (for non-leaf functions). On the x86 you typically only save and restore a handful of registers: ebx, esi, edi. On the PPC you potentially have to save and restore a dozen or more registers. I routinely see functions with 48 or more bytes of set-up and tear-down, but it can be much larger.
XCode can't magically correct any endian assumptions your code makes.
If you use the right libraries endian issues are automatically corrected.
So basically all you need for testing is one PPC mac and one Intel Mac. But when you get to the point where both work, it's pretty easy to maintain the endian compatibilty going forward (as long as you remeber to keep using thr right libraries) which is what makes years of dual proc support not such a big deal.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
Apple will support PPC macs in OS X probably about seven years after the last one is sold.
That's based on older models that OS X supports today, every release it seems like back support from OS releases is about five to seven years.
We know that not all mac will switch at once, say high-end Intel macs start coming out the year after next. That's 2007, which means there will be solid PPC support from Apple until at least 2012.
WHat about software vendors? Well you can imagine they would have strong motivation to keep software working well on PPC macs until the percentage of Intel macs is a lot higher than PPC macs. But that will take some time, so I think in the end you'll see universal binaries from just about all Mac software makers until at least 2012 - and it costs them nothing to keep making the universal binaries if they decided to drop testing support for PPC versions, which could extend it out longer.
In short, buy the powerbook.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
> deltree /y C:\Windows
> _
It appears likely that having a chip supply would allow Apple to make a G5 plugin card for new Intel Macs if some high powered hold-outs demanded G5's remain available.
I see no concern about purchasing chips for a tech Apple claims it's dumping. It actually makes it more likely they'll be prepared to deal with stragglers and hold-outs so that they can officially abandon PowerPC retail computers sooner (if not the chips and architecture itself).
Apple considered the Cell and it was just not suitable.
The Cell is a general purpose PPC processor with 8 other vector processors. If your app can't be vectorized then the 8 other processors are about worthless.
However the rest of the world is going dual core general purpose CPU's. To go dual general purpose with the Cell would require 18 processors and a ton of heat.
Plus programming for Cell would be hard because nobody has done it, so there is a lack of talent out there. It would be Apple fighting the tide all over again.
Apple is now about going with the tide and controling where it will go. Going after market share and developers, which are x86 based and trained.
IBM has the big server and cluser market sown up anyway, so Apple can't sell any hardware there.
You just have to wonder about the chips Intel is going to produce for Apple. Most likely they'll end up in iPods and Macs. When Apple came out with the iMac it was revolutionary in that it had only USB for serial communications. Wouldn't it be nice if they could convince Intel to do the same for their processors? I mean Apple doesn't need x86 backward compatibility (just low power and high performance).
One thing that many people don't consider - does an intel chip HAVE to the of the x86 variety?
- Chris - PC, Apple and guitar Geek
Dual processor motherboards, has two sockets. One for an Intel chip, and one for a PowerPC chip. You choose which one you wish to use at the factory when you order it. If you chose one, and decided you need to run programs designed for the other, you can order that chip and have an Apple Tech put it in for you. If you cannot decide, order both chips and use them in a parallel processing.
// or PC on a card for the original PowerMacs.
Special dual-use Macintosh motherboards can be designed to work with an Intel chip, a PowerPC chip, or both.
For those who bought an Intel or PowerPC Mac without the dual-use capability, Apple will be happy to provide an expansion card with the other chip on it to run programs designed for that processor. Just like they used to have an Apple
How about we put a fire under the MAME/MESS community and they design a PCI card standard that uses Intel X86, PowerPC, 680X0, Z80, 6502, 6509, Z800, MIPS, SH4, and other processors on it, or have the sockets to add processors to it for emulators and other systems to use to run non-Native code on the processor card. Then your F/OSS software can use the PowerPC on the PCI card to run PearPC on your Intel X86 Linux PC at almost normal speeds? Stick a G4 or G5 on a PCI card, and sell it for use with F/OSS software that uses the card.
How about the Asian community makes a PowerPC socket for their Intel X86 motherboards to use it as a co-processor for emulators and other things? Just buy a PowerPC chip, clip it in the socket, run PearPC, Bochs, or whatever, and be happy with your native run PowerPC OS of choice.
Remember, Slashdot does not have a -1 disagree moderation, and no, troll, flamebait, and overrated are not substitutes.
Donate free food here
"The G4 is a pretty energy efficient processor,"
So is an 8086, but you don't use it in a modern computer.
I think Jobs merely wants to exercise his new weapon of choice (sorry... just love that video) as much as possible. The longer he can keep PPC on life-support, the longer he has an extra negotiating chip.
Thanks to G3's holding down the low-end of Apple's line for so long, nearly every app on the market has a G3 code path (otherwise you'd be dumping an awful lot of relatively recent iBooks, iMacs, eMacs, etc). Rosetta simulates a G3; the application will simply take the G3 instruction path and run fine, just a tad slower. Emulating a vector instruction set like that across platforms would have been hell, and likely slower than the G3 codepath in the first place.
Only applications that are G4-only will have any trouble (damn few outside of Apple, and theirs are already Intel optimized).
I don't know what kind of crack I was on, but I suspect it was decaf.
If you went to the pain of adding Altivec enhancements by hand, then you're probably going to have no problem recoding them for Intel's SSE and such. Most major products are cross-platform and have this code written already, and most minor products just used compiler optimizations. Either the hard work is already done, or Intel will do it for them, gratis.
And as I said elsewhere, nearly every program has a G3 codepath, so Rosetta will run them just fine. Those that really got a significant boost from the G4 will be hurt the hardest, so those will be the ones you upgrade first to Intel native versions. Since such apps clearly have made an investment in the platform and have a need for speed, they're also the most likely to be ready first.
I don't know what kind of crack I was on, but I suspect it was decaf.
1. Produce an Intel version of OS-X
2. Make a pretend announcement to change over to Intel
3. Leak the Intel version of OS-X
4. Let Windoze users salivate over OS-X.
5. Change back to PPC.
6. Windoze users buy PPC Macs.
7. PROFIT!!!
I'm not sure how Apple making a decision to keep their options open constitutes reason for anyone to twist and turn. To me it sounds like smart business, which is something the old Apple didn't even know existed.
For the last several years Apple has been running off a new playbook. I think it is fair to say that talking about its dumb moves from six years ago or more amounts to archaeology.
Read the EFF's Fair Use FAQ
What a load of FUD and bullshit. All I'm going to point to is the two very successful transitions Apple has pulled off with 68K->PPC and OS 9 -> OS X, not to mention the flood of early reports that every damn app people run in Rosetta runs just fine. Photoshop, Office, they all just run.
Please, come up with more than FUD for your doom and gloom scenario.
I don't know what kind of crack I was on, but I suspect it was decaf.
This story is just an artifact from the announcement that Apple contracted Freescale until 2008 to provide processors. This may be simply a CYA manuver in case something goes south with the intel deal (be it delays etc).
So I'm not sure that this means you won't get your portables until 2008, just that Apple has a backup plan in case, with intel portables in 2008 as a longest possible estimate.
-T.
It is for warranty and repair contracts. You can get a 3 year warranty with AppleCare. The are telling there corporate/scientific users that they will honor this going forward so to reduce the risk for those buying machines now. They need to have replacement parts and machines if you buy one on the last day they will use the PowerPC. This is the reason! People over think everything Apple does.
his has been going on since Apple was late with color displays ("The Mac doesn't need color. It has resolution"), through the day Jobs bowed down to Gates on the big screen...
Apple supported large high-resolution color displays before Windows 2.0 even shipped. Look it up. Yes, the original Mac was B&W. But starting with the Mac II the platform held a huge lead in color graphics support for years and years.
Why we won't see a G5 PowerBook/iBook...
Re:Why we won't see a G5 PowerBook/iBook...
You might also want to check out Apple's Universal Binary Programming Guidelines. Note the section titled Conventions (emphasis mine):
Apple has only give developers an IA-32 roadmap for Intel. Not a single Apple engineer has given any information about future 64-bit support on Intel processors. The reasons for this are outlined in my links above, but the bottom line is that 64-bit x86 chips cannot run the same way that 64-bit PPC chips can. A 64-bit PPC chip (the G5) can run a 32-bit operating system (Mac OS X 10.4) and both 32-bit and 64-bit applications at the same time. However, a 64-bit x86 chip can only run 32-bit and 64-bit applications at the same time under a 64-bit operating system.
This is why IA-32 will be first. Apple has not yet announced what they are going to do for 64-bit support on Intel nor have they given developers the roadmap that we need to prepare for it.
infested with jello like fishes no melotron wishes
>For example, many Mac users still run Microsoft Office
>on the Mac. Microsoft isn't obligated to offer that on
>x86 for the Mac, now that the five year deal has expired.
Why did Microsoft come on stage at the Keynote and publicly promise to offer Office/x86 then?
> This is more about keeping options open in
> case the Intel transition doesn't come off.
I don't think you can know that now.
The P4M is pretty much gone. Everything shifted to the Pentium-M a while ago with Apple looking to the the Merom for their Intel generation computers. The Merom is dual core @ 35W max with 1-2W nominal @ 65nm. I'm not a rah-rah Intel kind of guy (I own stock in AMD), but they really got something with this next line if they can deliver on their 65nm process.
[RIAA] says its concern is artists. That's true, in just the sense that a cattle rancher is concerned about its cattle.
Apple will be transitioning all their G4 product beginning in 2006. They contract Frescale till the end of 2008, which 3 years - the length of a typical applecare extended contract. So, I guess they only intend to sell these systems till the end of this year. Given the recent announcements of Intel processor lineup, I guess Apple will release new Powerbook, iBook, and mac mini's along side Intels release of dual-core and single Yonahs.
You don't have to be smart to use a Mac, you just have to be smart enough to buy one
For one thing, no matter how they try to play it down, their execs at the pentium d launch announcement went on at length about how the "D" in pentium d stood for DRM.
Add to that the fact that pentium chips add on kludge after kludge, and the fact that my last pentium system had a bug which resulted in it not being capable of properly encoding video, and i'm sticking with ppc until it's good and dead.
To that end i've sacrificed a lot of my budget to buy a high end g5 with as much as i can cram into it for the long wait.
Hopefully by then they'll have worked the damned bugs out of their stupid chips, and the populace or hackers will have successfully thrown off the yolk of intel's processor embedded DRM.
VLC FOR MAC IS DYING! IF YOU DEVELOP, PLEASE SAVE IT!!
You don't have to save anything you don't use, and you (the compiler) should only use extra registers if it is advantageous to do so (taking into account the cost to save/restore them). Having more registers available simply gives you more options.
That's the optimist's view. In reality it's less clear than that. When you see 20+ instructions to handle entry and exit code, then that strikes me as RISC gone wrong. And this is with a good compiler, not just some student project.
As an aside, when you pass parameters in registers you often have to turn around immediately and copy them to different registers or write them out to memory. It's not hard to come up with cases where it's faster to pass parameters on the stack.
See SufficientlySmartCompiler.
They need a piece of news to keep people interested in PPC Apple products.
Been to Fry's since the announcement? No electricity in the air.. Just a bunch of high-priced, under-equipped machinery.
Those G4 iBooks won't sell themselves, especially not when they are presumed to be the last of the breed. Who wants that?
I have a second gen 2.7 dual.. it's heat syncs have 'liquid cooling" warnings. but as for the grandparent.. BS! no stupid p4 is faster than a g5. they're obviously stripping something down on it for it to boot that fast.
VLC FOR MAC IS DYING! IF YOU DEVELOP, PLEASE SAVE IT!!
So I'm somewhat curious as to why you feel so strongly about using a PPC chip.
Did you read what he wrote? The original poster said:
Personally I just have a whole bunch of personal Altivec code and I don't want to have to rewrite it.
That's why. His code won't run without rewriting, so getting his job done in the quickest manner means using the optimised code he already has. Rewriting it for SSE3 or whatever not only may not be as fast, it's a lot of extra effort.
If your comment title says 'Re: Foo', I'm not likely to read it.
See IDF news for the answer to that... Apple will incorporate Intel's newest processors from 06/07, meaning processors like Yonah, Merom, Conroe, and Woodcrest. They are all based on an entirely different microarchitecture than the Pentium 4.
FWIW, the PPC also has an instruction with which you can save an arbitrary number of integer registers (even all of them) in one instruction. It's not used because it's slower than the combination of the individual replacements. Just like on 80x86 things like lodsb, loop and jecxz are slower than the alternatives consisting of more instructions (on modern processors).
To pass them on the stack you often first have to load them from elsewhere in memory first. By passing them via the registers you can distribute the load on the memory interface by scheduling the loads with code at the caller side and the stores by code at the caller side. If you don't have to load them from memory, then you don't have to load them from memory either when passing them via registers.And in other cases it is not necessary at all to copy them back to memory, so you also win.
Donate free food here
One possible conversion tool is macstl (which comes with some scripts that claim to help in converting Altivec to SSE) - I haven't used it myself, only read the blurb.
Huh?? How much programming experience do you have?
The 68K-PPC transition was only "successful" to those who live in the Jobs Reality Distortion Field. In reality, Apple market share dropped substantially through the transition. Before the transition, Apple was a serious #2. After the transition, Apple was a niche player.
Ars has a much more thorough and thoughtful take from someone who actually follows Apple and has some common sense.
Except the PPC has instructions to save multiple registers at once to memory. So you can save registers 3-8 to the stack in one 32-bit instruction. I assume the compilers you are using do not do this for parameters because it's faster not to. Much like x86 xor'ing a register with itself instead of setting it to zero.
So I would question what kind of functions these are that you say are takin 20+ instructions. But also, ultimately performance has little to do with the overhead of functions since the vast majority of time is spent in simple loops.
What you say is correct though, that functions are more wasteful on PPC, but another way to look at it is that languages like Java will run insanely fast on PPC since they hot spots are often inlined 5 or more functions deep -- something that is basically impossible with a language like C or C++. That means no function calls, and using up all the registers for something useful. RISC may be a poor choice for C... true dat.
Its clearly missing the "?????????" point....
HI O WISE PRINCE. WHT TOOK U SO DAM LONG?
That has a whole lot more to do with the fact that Michael Spindler was running the company. That's when OS development went down the tubes (Copland), hardware went down the tubes (ten bajillion Macs that were configured the same, yet had different model numbers - Performa, anyone?), no marketing at all, no strategy, etc. Everyone was being told to develop for Copland (which was going to break all kinds of software), Apple developed several large software packages that went nowhere (PowerTalk, QuickDraw GX, PlainTalk) and started losing money hand over fist to clone makers. Yeah, they managed to do all of that in the 1994-1996 era, which is after PowerPC and prior to Amelio/Jobs.
None of that has a damn thing to do with the 68K to PPC transition, which was one of the smoothest the industry has seen. Your software "just worked", albeit slightly slower. By the second generation of PowerPC, emulated software ran faster than on the original hardware (jumping from 40Mhz to 120Mhz in a year and a half didn't hurt things). If you used Codewarrior (the savior of Apple at that time) then a checkbox made your code PowerPC native. The System 7 transition (32-bit clean, virtual memory, 32-bit QuickDraw, aliases, etc) was FAR worse than the 68K->PPC transition, yet that didn't scare off developers.
So, as someone who not only lived it but was paying attention, please take your FUD and misinformation and politely shove it.
I don't know what kind of crack I was on, but I suspect it was decaf.
I think you missed the point; he's not talking about high level code (i.e. C, C++, or whatever Apple uses), he's talking about machine code (compiled high level code).
A simple one line statement in a high level can very well take several lines/instructions to execute. I used to write assembly code for a Motorola micro-controller (don't remember the model number) and an instruction that would be extremely simple in C took several intructions in Assembly.
C code:
int i = 5;
i = i + 12 * 2;
Assembly:
Reserve memory in RAM
Set memory to 5
Store memory value in CPU register A
Do multiplication and store result register B
Do register addition and store result in register A
Store value of register A in RAM
Note: Some steps may be more than one instruction.
So I can understand how it may take 20+ instructions to handle entry and exit code. The CPU needs to worry about the state of the registers, RAM, stack, etc; so it takes quite a bit of extra instrcutions to check the status of the registers (and etc) before and after a function is called or whatever else is done.
BS! no stupid p4 is faster than a g5. they're obviously stripping something down on it for it to boot that fast.
Incorrect. P4's have continually been released in faster revisions while the G5's have been stuck at nearly the same clock speed for quite a while.
The performance of newer P4's easily exceeds that of the G5's.
Perhaps this is Bizarro-slashdot?
“Common sense is not so common.” — Voltaire
The new mobile G5 uses almost the same as the current G4 processors.
Apple has mobile G5s? I just checked and they don't have any G5 laptops, do you know when they go on sale?
FalconShould there be a Law?
Anyone else feel bad for their developers? Having to develop two parallel code branches for years to come? ouch...
Developers shouldn't need to be concerned about having to work with different codes or source branchs. Apple has worked on universal binaries, using Xcode developers can output both PPC and Intel code in one package.
FalconShould there be a Law?
Not correct.
The 2.5 and 2.7 ghz duals are liquid cooled
(9 fans and liquid cooling).
The 2.0 and 2.3 duals are air cooled (9 fans).
And I don't believe it is water cooling but rather
some synthetic liquid that enhances the cooling
process. Creepy.
It's almost as if people now accept what Kernigan and Ritchie was saying all those years ago. I'll wager that a move to x86 was on Apple's mind a long long time before the announcement, and thus reasoning behind moving to a *nix platform.
That's what they said when they annouced the switch, that they had OSX running on Intel for 5 years.
FalconShould there be a Law?
HD space isn't that cheap when you're talking about laptops. I have a 100 GB HD in my PowerBook that is nearly filled up. It kinda pisses me off that I have to lug an external 250 GB HD around with me all the time. I don't think there are bigger 2.5" HDs in mass production at this time.
Yea, I'm planning on getting a Powerbook early next year and hope they get bigger hds than the 100GB one they have now. I don't want to have to rely on an external hd for storage, especially if I also get a new DSLR camera. On the other hand so I don't have to have a laptop with me all the tyme if I get a DSLR I may just get a portable hd so I can offload my photos onto it.
FalconShould there be a Law?
Your Alpha, do you have NT4.0 on it, and if so did you have any trouble installing software on it? I got an Alpha several years ago but I could only install a few programs so it's been pretty much a waste. DEC's FX!32 didn't work that good for me.
FalconShould there be a Law?
Two reasons kept me away from the Mac in the early '80s - lack of color and cost.
I used an Amiga at the time, loved the color and video capabilites, low cost and multitasking.
Apple was first to the mass market with 8bit color screens in the Mac II (I believe in 1987), but the cost was astronomical. So yeah they were there with color but for years after the Mac IIs release, the lower end black and white Mac SE and SE 30s were commonly used. I think the Mac IIsi was probably the first relatively low cost color Mac. Even then the IIsi had slow video, since the video card used system memory rather than dedicated vram.
Sometimes my arms bend back.
yeah, and my dogs have lovely wide white wings, and they fly and fly around and around our house! I'm not some provincial mac zealot. I've been continuously exposed to varying models of p4, g4, and g5 systems, and I'm telling you right now that in terms of response time, a p4 system will only match that of a dual g4, never that of a g5.
VLC FOR MAC IS DYING! IF YOU DEVELOP, PLEASE SAVE IT!!
I think they were talking about the iBooks other then the power books. If my Powerbook died right now I would probably just save some cash and switch to an iBook. But if they were a G5 or Intel based PowerBook I would go with that. The Power Book line is starting to really drag.
I'm hoping they have Mactel Powerbooks by January, February at the latest. I'm planning on getting a Powerbook around then and am hoping that Intel based powerbooks will drive down the price of G4 Powerbooks. I also want to compare the price/performance of G4s and Intel laptops. If the Intel ones are better in this regard I may get one instead of a G4.
FalconShould there be a Law?
I can see that the warranty issue is a concern - heck I just bought an iMac G5 2 months ago w/ the Apple Care warranty.
... are the Intel chips Apple will be buying likely to be more expensive than deals they have with IBM on the G5? Afterall, if Intel were more affordable, the low end systems would have Intel in them only.
But I have to wonder - the article mentions the 'low end' systems will continue to use PPC chips (by 2008 likely still the G4??) then I have to wonder
Actually Microsoft released Windows to run more than just X86 compatibles. I know as I've got Windows NT4.0 running on a DEC Alpha.
FalconShould there be a Law?
just kidding...
They probably secured the deal just in case things get off to a rough start with Intel. Furthermore, bet they'll need a few extra G4 chips for future hardware repairs. The deal doesn't say that they WILL buy the chips - just they CAN buy the chips if they need them.
The bits on the bus go on and off... on and off... on and off...
World Thumpin' Friar, no doubt.
first they make the PPC-based used Macs prices fall down by telling the horror-story of moving to intel and now make hope again... clever boyz over there. too bad i didnt buy a G5 before this news
free 880 megs file hosting - www.FTPZ.US - best
Er, no. Reality distortion field alert.
68K floating point didn't work in the PPC's emulator. If you ported your engineering application to PPC, you had to go from 80-bit floats to 64-bit floats, making your data incompatible with the 68K program and the x86 version. Most engineering apps dropped the Mac at the PPC transition.
There was a third-party "SoftFPU", but it was slow and didn't work very well.
If you only test your code on one system, don't count on getting many PayPal dollars from people registering their copy.
resigned
I made this for work, but it's useful all over the place.
I rarely criticize things I don't care about.
If you look at the benchmarks (not the PR benchmarks released by Apple), you'll see that both the Athlon64 and the P4 outperform the G5 by quite a margin.
Even the developers who are testing OSX86 commented how the speed on the P4 developer systems is impressive. Hell, even Steve Jobs must have seen something on Intel's side.
To get the Mac to win common benchmarks, you need to compare a dual G5 system to a single P4 system. But single vs. single? Speed revisions for the G5 have been slow to come. The P4 surpassed it long ago with its constant speed increases.
benchmarks can be manipulated like any other statistics. bring forth the full spec on the benchmarks and i can tell you more. In the mean time.. as a power user i have my own standards of responsiveness and power. If you open more than 5 intensive programs on a p4 machine and swap between them you'll end up with lag which does not occur on a g5. As far as osX86, keep in mind it's a DEV edition. It is not the final product, and is likely stripped down or some other way altered from what is currently out there. If you want to talk speed.. my g5 can boot win2k in 25 seconds and run it in emulation as if it were on native. I'm particularly abusive and demanding upon my systems, and find it hard to overload a g5, but a pentium machine will start to complain at the drop of a hat.
VLC FOR MAC IS DYING! IF YOU DEVELOP, PLEASE SAVE IT!!
Think about it. Who else uses the chips that Apple may want? Nobody. They're ordering an unspecified number for a period of time so they will be able to put a new motherboard in your uncle's iBook in 2007. If they didn't get a deal like that, they'd be screwed if anything happened.
Are you using a SMP-G5? That would explain it. Lagging and stuttering does occur on uniprocessor-systems (all P4-systems are uniprocessor-systems), but SMP-systems the problem is practically eliminated (yes, that includes Xeons and Opterons, as well as G5's).
Lesbian Nazi Hookers Abducted by UFOs and Forced Into Weight Loss Programs - -all next week on Town Talk.
At this point you're just making up false "facts" to support your invalid argument.
The fastest P4s and Athlon64s are faster than the G5s. Despite your claim of not being a Mac fanboy, your actions speak louder than words. You'll never admit that your G5 is slower no matter how much proof you're confronted with. You *are* the typical Mac fanboy.
You coded yourself into a dead end. 'Altivec' is indeed rapidly becoming 'legacy.' You can file it where a lot of people file their 6809 assembly code.
Deal with it. Maybe you should port your code all over to the Dec Alpha processor. *smirk*
resigned
I was really looking forward to the intel move. I want to buy a high end macintosh so that I can dual boot windows and os x. I could use the windows partition for my gaming needs and the os x partition for my computing needs. For all you naysayers, imagine not being able to dual boot windows/linux. Imagine having to maintain two machines, one for playing games, and one for getting things done. Now I will have to split my resources two ways, buying a mediocre pc clone and a low end mac. How depressing. To those who say: Buy a console! Console games suck and are for mutants with multiple thumbs. PC gaming forever (hopefully sometime soon on a mac) that is all, blah