Behind the Apple-Intel Deal, IBM Surprised

[fluidinclusion]

YEAH Great Question

As it stands, peripheral support from some companies is lackluster/shakey at best when it comes to the Mac. I can't help but wonder how many manufacturers will decide this move to Intel is the final straw and decide to drop Mac support for their peripherals and/or devices? Anyone have an idea how much work will be required for drivers to work with OS-X on Intel?

Agreed, that is the largest problem I had (mostly in the past with scanner support, etc.) with Apple. However, I wouldn't think it would matter. Couldn't Rosetta do the same for drivers too?

 

[iMeowbot]

Yes killing the clones was almost 10years (7.5 or so) ago but I think it fairly closely coincided (perhaps coincidentally) with Motorola's lack of enthusiasm for investing lots of money in PPC development for PCs... which triggered a whole sequence of events leading to the present situation. Of course Motorola's inability to improve 68K performance was the thing that led to the switch to PPC in the first place.

There's nothing coincidental about it, Moto announced as much at the time. This is why G3 came from IBM, and G4 was an embedded CPU with some anemic interface glue tacked on. Altivec was already more than half done at the time of the clone disaster, and Moto wanted it for embedded anyway. IBM would be doing a lightweight POWER4 anyway because they were entering the blade market at the time; they had no interest in adopting SIMD until Nintendo came by with half a billion to invest on Gekko. That's really how it's been a while, Apple were getting the leftovers from embedded and game development.

 

[BenRoethig]

The relationship between Apple and IBM may have soured. IBM will still have to produce the chips that have been contracted. Wouldn't be surprised that if any further development would be halted.

I have my doubts those chips are contracted for a further two years.

 

[fluidinclusion]

5 years total

I have my doubts those chips are contracted for a further two years.

I can't remember for sure, but wasn't the original contract for 5 years with IBM, or am I just remembering wrong?

 

[Photorun]

"...whine whine 'I think' 'I think' 'I know' 'I think' whine whine..." is about all this and the other forums are devolving into. Pathetic.

And another reason escaping, oh, like 99% of everyone here is the laptops, the G5 wasn't going into the Powerbook, or rather COULDN'T, not now, not in the near future, and not in the past as I used to flamebait people on this hear forums all the time after hearing all those pathetic "oooh, G5 Powerbooks must be abouuutt... now, no wait... NOW... no NOW!" Wasn't happening.

Now those here who'd like to stick with IBM raise their hands if this means doing so means no new faster iBooks, Powerbooks, and Mac Minis for two, maybe three years, possibly longer. Okay show of hands people. Show of hands. Anyone? Anyone? Okay. Now that that's settled. Show of hands who doesn't want Apple in any other business but desktop tower machines? Okay, anyone? Anyone? The latter group that raised their hands would be the only market Apple has left if they stayed with IBM.

If you don't like where Apple's heading... get out! Seriously, go buy a Windoze machine, enjoy your XPee and viruses and spyware and ugly hardware. Don't let the door hit you on the arse on the way out fanboys. Those of you who will stick this out in the hopes for the brighter future, stick around. It may get a bit bumpy, it may not, and quite possibly fascinating ride with faster, cooler running, lower power machines and potentially new worlds of possibilities. Enjoy the ride!

 

[AidenShaw]

sorry, missed that factor

The fact that he bought it this afternoon entitles him to an upgrade.

You're right - sorry, I forgot about the up-to-date program applying to old models purchased between 12 April and 9 July.

My reading of the up-to-date conditions (http://www.apple.com/macosx/uptodate/) seems to be that you have to pay the $9.95 at least once per invoice, however. (If you bought two systems on the same invoice, you can pay $9.95. If you bought two systems separately, pay $9.95 per system.)

Until Apple sends the kit, however, the iMac is technically running a pirated copy.

(BTW, I think that $9.95 per system for the upgrade is very reasonable....)

 

[zigmoon]

Intel is strong

Intel is a good hardware ! But the real probleme it's juste "****dows".

 

[VanNess]

I think it's overstating it a bit to say that IBM royally screwed up... First we (or at least I) don't know exactly what the agreement they had with Apple was, other than apparently (based on what Jobs said) they promised 3GHz by last year. Second a 2.7GHz G5 is quite competitive with a 3.6GHz Xeon for applications it was designed for (see the recent arstechnica article). Although agreed that it lags a bit behind AMD.

Did IBM agree to target laptops with the G5? Again, I personally don't know, maybe they did. However, it could be the deal was that IBM would provide bragging rights on high-end pro-desktops and Moto/Freescale would take care of laptops/consumer machines.

True IBM had problems with their process, but apparently so did the rest of the industry.

I don't think IBM could be surprised that Jobs was unhappy with their current relationship, but I imagine the way the split went down might have been unexpected.

I'm not privy to the agreement either, but the results of it up (taking into account ALL of the developments) until now speak volumes. And it's not just the present absence of a 3ghz G5 or G5 based laptop, I think Jobs was honest in that regard. I think there's more...

It may be true that the entire industry "hit the wall" as Jobs said when everyone began moving to a smaller process, but with IBM's G5, it seemed more like a hit and run. The updates were slow in coming, they were mediocre by just about any measure, and from the 2.5 G5 to today's 2.7, it needed a special, Apple designed liquid cooling system just so it could be put into use. Sure, it's a little progress, but where's the roadmap now? The 3ghz is more than a year late, and you just can't go on selling last year's excuses forever. Especially when the competitors seemed to have already found ways to improve performance in their products. Apple certainly hasn't done anything with it's designs to prevent the implementation of faster G5's, in fact it looks like they've gone out of their way to accommodate whatever IBM tosses their way.

I don't recall Apple or IBM ever stating that the G5 was implemented only for desktop configurations, as one would expect to hear. In fact, when Apple was asked about the status of a G5 laptop (and it was asked often enough) Apple has repeatedly stated that their were "significant engineering issues" involved, but never denied it was not in consideration. Apple simply stated that we shouldn't expect one anytime soon. And as Job's lamented in the keynote, "we haven't be able to deliver one to you yet." So there's nothing to indicate that the G5 had been absolutely removed from consideration for inclusion in a laptop previously, by either Apple or IBM

So along with the developmental gridlock of the G5, there was the fab problems, the countless shipment delays, and the unprecedented announcement by Apple that the introduction of the new G5 iMac would have to be delayed - for months - due to the unavailability of IBM's G5 chip. IBM's "surprise" seems to have clouded it's memory of how that story rippled in the financial press at the time.

Sure, maybe there were acceptable (not desirable, but acceptable) technical reasons for all this. Maybe there were unanticipated issues involved with the development of the PPC roadmap that hindered the G5 beyond what could have been reasonably foreseen at the time the 3ghz promise was made. IBM was doing everything it could. Or was it...

Enter the final straw that broke the camel's back, and it's name is Xbox 2. When IBM announced it was supplying not just PPC chips for the Xbox, but multi-core PPC chips clocking in excess of 3ghz, that was more than enough for Apple to see exactly what IBM's R&D had been up too all this time. I think that announcement was astonishing to Apple, not because the Xbox chips could also be used in Apple gear, but because it showed a virtual collapse of IBM's resolve to bring the Apple G5 roadmap forward into at least the end of the decade (not to mention just make good on the original and still unfulfilled 3ghz promise). And the fact that this was for Microsoft, Apple's chief rival and competitor, certainly made the decision to bring in Intel that much easier. In any event, it should be of no coincidence in retrospect that the original Wall Street Journal article (not the Cnet article) that described a possible Apple/Intel alliance (and largely dismissed by everyone) occurred relatively soon after the Xbox announcement.

Game over, IBM.

 

[dabirdwell]

An interesting aside to this, IBM has announced a new low-power PowerPC chip.

It's the next in the processor family that brought Apple the G3, a high quality processor in my opinion. Can someone more technically inclined than I interpret these power usage numbers? Thanks.

I realize this model tops out at 1Ghz, but it looks like IBM has been able to implement some of the power management technology that they have been talking about for awhile now. Seems like in the past I heard it called PowerTune technology or something. It allows dynamic FSB multiplier changes, I think.

Well, let me know if it's anything interesting.

750GL

Edit: Also look at this. IBM has a new feature on their site promoting the future of PowerPC and POWER processors as great. Check it out.

 

[steeldrivingjon]

Yes but how will 32bit dual core Yonah stack up against the 32bit dual core G4 that Freescale announced 9 months ago? Remember Jobs said "Exciting PowerPC products are still in the pipeline" I don't think by "Exciting" he meant another 100Mhz speed bump. However it might mean a dual core G4 Powerbook! 🙄

Yeah, Freescale and IBM might manage to pinch out some new PowerPC chips that are a little faster, or maybe even dual-core, in the next two years.

But then what? How long will it take for them to squeeze out the next slight upgrade? It'd be a mistake to think that Apple would be all set if only the next CPU updates could ship.

(And, really, if their upcoming offerings, for *two years* are *that* meagre and questionable, why would Apple want to stick with them?)

Maybe everyone should send Freescale and IBM some ex-lax.

The PowerPC is a better architecture, but plenty of vaporware has nice architectures. If you can't get the implementations that you need, because it's too expensive, the beauty of the architecture means nothing.

 

[steeldrivingjon]

Agreed, that is the largest problem I had (mostly in the past with scanner support, etc.) with Apple. However, I wouldn't think it would matter. Couldn't Rosetta do the same for drivers too?

You wouldn't want it to. Driver code needs to be fast. Rosetta probably wouldn't be a good idea to have in the kernel space, and probably isn't allowed.

Rosetta might be okay for more complicated drivers which are split into a kernel-space driver and a user-space driver that communicate. The user-space driver might be able to run in Rosetta.

(These split drivers are done because there are limits to what you can do in kernel space, limits to what APIs you can access, and you don't want to do too much in kernel space, because bugs can bring down the whole OS.)

 

[MarkCollette]

How could the PPC be better than Intel yesterday, but now Intel is so much better than the PPC used in the Mac? When were we lied to. My money would go to both times.

I see a lot of people asking this question in the forumns, so here's the answer as far as I know:

For years the G4 has been noticeably slower than a P4 or Athlon. The only thing it had going for it was Altivec, so it could win on some things like media encoding. But then, its slow bus became such a limitation, that the G4 just could not compete, especially as things like SSE and SSE2 came out.

When the G5 came out, it had a much faster bus than anything else, and had two floating point units, plus VMX (Altivec), so it really was as good or better than anything else. But, AMD was just releasing chips with integrated memory controllers, and Intel starting bumping up their bus speeds to almost as high as the G5's, and they put way more cache on their P4s. So, the G5 can still kick ass in some floating point stuff and VMX stuff, and it's pretty comparable elsewhere, but it's so hot that the fast ones are only in expensive PowerMacs, or lower speed ones in the iMac G5.

With the Pentium M ramping up, and looking to trounce the G4 in notebooks, and compete nicely with the G5 in desktops, it's looking like the place to be.

A little while ago, most computer manufacturers started selling more portables than desktops, so portable friendly CPUs are the future. As it looks right now, PPC has no portable future.

So, that's how, before, people could say that PPC was better, but now they can't, or at least they won't be able to in the near future.

 

[neonart]

How could the PPC be better than Intel yesterday, but now Intel is so much better than the PPC used in the Mac? When were we lied to. My money would go to both times.

I don't remember in the keynote anything about how the Mactels are so much better. I do remember a keynote that said we would have 3Ghz G5's a year ago. You don't see that as a problem?


Do you believe that IBM would deliver a dual core 3 GHz PPC to Apple at this time.

That was supposed to happen a year ago. The PPC architecture is lagging- sorry.

All others should put a boycott on all thing Apple. This could start with the Intel Macs for those that need their iPods. In recent years Apple has changed to a music company away from a computer company. This Intel move will divide the already small Mac market into pieces. Maybe too many to keep it alive.

Yup. A boycott is exactly what the "already small Mac market" needs. Good advice!

To many the processor does not matter. To many like me it is all about the processor. I chose the Mac 21 years ago & I still am just as anti-Intel as I was then. I was a Heath-Kit computer user before that time. They chose to make IBM Clones with Intel 8086 processors in them. Being anti Intel I chose not to follow. Now Apple is making the same decision. This time I will make the same decision, not to follow my computer manufacture. This is limiting my choices severly. That is why I must keep what I haverunning for as long as I can.

Did an Intel processor steal a girlfriend or something? I'm sorry, but I just don't see the basis for this hatred. I use Macs for the OS and the user experience. If I can have those two things in my next Powerbook with a Pentium M, I'm good.

Oh well.

 

[steeldrivingjon]

The fact that he bought it this afternoon entitles him to an upgrade. I would agreee that if he doesn't take advantage of the Mac up-to-date program, then he is pirating OS-X. On the other hand, if he installed an existing copy in anticipation of sending for and receiving his u-t-d copy, I see nothing wrong with that. As he provided little information regarding his intention, I'll reserve judgement.

Isn't the up-to-date cost just for media and handling, and not an "upgrade fee" for the license?

In that case, using the DVD he has just saves Apple the cost of sending him something he doesn't need. It'd be sketchy if he bought the mac used, but that's not the case.

Does the family 5-pack include 5 DVDs? If not, that means Apple recognizes the concept of a "right to use".

 

[steeldrivingjon]

Yes. But the thing I still don't understand is why Jobs would unveil this information so early. People have been saying all along that this is really going to drive down the sales of the current line up. How could Jobs overlook that fact? Does he not want to sell computers for another what few years before the Intel line-up comes to existence?

Apple is making the switch to Intel and ditching the old stuff. Who would want to buy the old stuff now? I mean does anyone else see how this could really hurt Apple? I just don't see the logic.

For one thing, over the next year we'll probably be seeing a stream of application updates, where the updated app is a universal binary, even though nobody will be able to use it but other developers with test machines.

That'll help Mac users chill out about the transition next year.

I have to think that, if Apple tried to move developers over secretly, there'd be a risk of the big guys taking a wait-and-see attitude about it. That'd result in the Intel port having less software at the start, and being less viable because of it.

But with so many people moving to universal binaries out of the gate, there'll be more pressure on Adobe and Intuit, etc, to keep up.

Most developers won't need the full year to go universal, but the big outfits like Adobe and Microsoft probably will - if not more than a year.

If Apple just told the big guys, in secret, then the big guys might wait too long before getting serious, which hurts Apple. Plus, rumors would fly, which would be more damaging than just laying it all out on the line. Little developers would wonder what they weren't hearing about and why they weren't being told.

Basically, this is the only way.

All Apple needs to do to avoid the Osborne effect is to commit to supporting current PowerPC Macs with new OS upgrades until the end of the decade. That ought to suffice for anyone.

 

[MarkCollette]

2) Intel

Intel's history before the Pentium Classic is immaterial to this discussion. Intel's Pentium Classic (P5 architecture) was nothing special in it's time and still isn't. However the original replacement for it, the Pentium Pro (beginnings of P6), was a pretty nice upgrade to it.

Actually, the Pentium was a noticeably big step from the 486. It added support for EDO DRAM which was noticeably faster than Fast Page RAM. It had a much better FPU. It was the 68k killer. Up until then the 68k could compete with x86, but not any more. And of course, socket 7 introduced the era of all the x86 knock-offs coming into their own.

 

[MarkCollette]

If this is true... does this mean cheaper Macs? 🙂 Or does Apple just make more money?

Most likely a bit of both.

I don't think that most Mac people really complain about the cost of Macs, we've only really complained about the performance per dollar, which this will address.

 

[dongmin]

Most likely a bit of both.

I don't think that most Mac people really complain about the cost of Macs, we've only really complained about the performance per dollar, which this will address.

Yeah, I don't really see Macs getting all that much cheaper. Apple will probably get a bit of a discount using Intel chipsets and mobos, but I imagine the Mac mobos now are already pretty cheap. Apple doesn't exactly use cutting-edge technology in their Macs. It's all pretty standard PC stuff; in fact, Macs lag PCs in several areas (RAM, graphics, I/O).

The only way Macs will get significantly cheaper is if they start moving a lot more Macs and save money from economies of scale. This won't happen over night.

 

[neonart]

Apple will probably get a bit of a discount using Intel chipsets and mobos

This is a good point! I really hadn't thought about this, but what if Apple is getting a double whammy on savings by having intel make their processors AND motherboards? (Hopefully not the graphics.)

 

[m-dogg]

fortunately? i would think the shareholders would like a lot more customers like that!

hehhh - yeah, I guess that's true...

I took his original post as saying that all mac users (becaue he and all his friends do) buy new machines every 9 months. now, becasue of this, they are going to boycott macs, which is going to cause apple to crumble b/c no one is buying new macs every 9 months. which, in reality, not a whole lot of people are doing anyway...and I think apple will be just fine through this transition...

 

[rotorblade]

Isn't the up-to-date cost just for media and handling, and not an "upgrade fee" for the license?

In that case, using the DVD he has just saves Apple the cost of sending him something he doesn't need. It'd be sketchy if he bought the mac used, but that's not the case.

Does the family 5-pack include 5 DVDs? If not, that means Apple recognizes the concept of a "right to use".

That's correct. The UTD program is for customers who purchase a qualifying new Macintosh computer on or after a given date that did not have the most recent version of OS-X installed on it when purchased. By most recent version, I'm refering to dot versions, i.e.; the computer came with 10.3.9 and 10.4 is on the store shelf.

The family pack comes with 1 DVD that is used for all computers.

 

[MikeAtari]

It's the next in the processor family that brought Apple the G3, a high quality processor in my opinion. Can someone more technically inclined than I interpret these power usage numbers? Thanks.

I realize this model tops out at 1Ghz, but it looks like IBM has been able to implement some of the power management technology that they have been talking about for awhile now. Seems like in the past I heard it called PowerTune technology or something. It allows dynamic FSB multiplier changes, I think.

Well, let me know if it's anything interesting.

750GL

Edit: Also look at this. IBM has a new feature on their site promoting the future of PowerPC and POWER processors as great. Check it out.

- Of course it's interestin that the Cell may go to 4 ghz.
However, it's just a Single User, with 8 dsp branches, chip.
It's also a simplified design.
- The 970 line is where IBM should have been spending their time and money. This line( for Apple ) and IBM's lower cost servers, could have fed sales to the POWER line.

For my part, I take this as an incredible blunder by IBM.
Power was getting a lot if interest, because of Apple's use of the chip.
What are those FOUR PHASES?
I guess I'm past denial, and now in the ANGER stage.
The thing is I KNOW IBM has got some of the smartest people in the world, so, I can't understand them not being able to address Apples needs.

IBM has good power mgmt, but, if your running your laptop at 80 to 100% it doesn't help. A hot chip will still be hot when it gets used.

 

[SiliconAddict]

I don't think you understand. Apple didn't do this simply because Jobs was pissed (although he would have every right to be). Apple did this because they had absolutely no other choice available to them. No matter how clever or innovative your plans are for a transition like this, no company would ever take on something this enormous at this time unless they had no other conceivable option available to them. You can bet the farm that if there was a sliver of a chance for IBM to keep pace with the rest of the industry, developing advancing and delivering the PPC, then the big switch to Intel would of never happened. There's been a lot of tabloid-like references to Jobs personality, but he's also almost universally regarded as pretty charismatic guy. As Apple's unofficial lead marketer, he has a unique ability (actually rare for a company CEO) to connect with ordinary people as well as industry professionals. He's also regarded as being very sharp, and as a technology visionary, he's been right on the money almost every time. But even all of those qualities won't save him (or Apple) if the products he sells come up empty-handed next to the competition.

It doesn't seem to be sinking in how royally IBM f'd up with the G5. IBM says they were "surprised," but how on earth could that possibly be? They've been sending the message consistently to Apple since the introduction of the G5 - via broken promises, massive shipment delays, and anemic G5 updates - that they were incapable of being a viable CPU solutions provider. Furthermore, IBM is no stranger to the computer technology industry. Having failed itself in the PC business, IBM probably better than most understands that maintaining key strategic technologies on your platform are essential for survival in today's market. To say that they were "surprised" by Apple's move to Intel assumes that they are either clueless about the very basics of computer industry competition, or they were unable to meet their technology commitments for internal reasons unrelated to cost (if cost was really an issue, that would have been vetted by the parties before this. Therefore no claim of "surprise"). No matter how you spin it, in the end IBM looks hopelessly feckless. And for Apple, that spells trouble with a capitol T.

In fact, if you want another clue regarding how bad this was for Apple, it was when Jobs declared that this was going to be "the year of HD video." Well, H264 is here. Now look at the processor requirements for playing H264 HD content on Apple's Quicktime website. Looks like they are pretty much locked into the upper end of the G5 line, not much headroom left over. Even the 1.6 G5 can't do it, and the G4's, which includes EVERY Apple laptop, need not apply. H264 is a massively important technology for Apple going forward on both the consumer and professional level, and about 3/4 of it's own computer line-up is presently shut out from playing just a 2 minute movie trailer.

Of course, that's not a reason to throw a tantrum. It's a reason to directly address the problem with real solutions for customers, not with another round of delays, excuses, and broken promises. The obsolescence clock is always ticking, whether your name is Steve Jobs or not.

By the way, thanks to scu and sacear for your comments earlier, Appreciate it.

That’s probably the sanest, most thought out post, I’ve read all weekend.

 

[savar]

Anyone care to enlighten us that don't subscribe to the post?

It was nytimes.com, but either way you need to go here:

bugmenot.com

 

[savar]

isnt that how it always is? im not surprised. and i wont be surprised if it turns out to be the best chip in the future.

The cell chip isn't really a general purpose computing chip. Read up on it at Ars Technica. While its great for certain types of computation, it seems that it will really serve video games, digital video/audio, and scientific applications best. Its not very good at heavily-branched, single threaded code.