What if Apple Had Stayed with IBM?

Posted:
in Current Mac Hardware edited January 2014
Where would Apple be today? I don't hear much about IBM's chip development anymore now that I'm fairly uninterested and now that most Mac sites share my feelings. Is IBM still making competitive desktop chips or is IBM only going for the console market and the high-end servers and supercomputers market?



My understanding is that IBM never really had chips that are cool enough for laptops (and still doesn't?) So if I had to take a guess, Apple would be in bad shape today since their current laptop lineups are Apple's most popular products next to iPods. There might have been problems getting some components for iPods and iPhones too had Apple sticked with IBM.



What do you guys think?



One thing's for sure though, IBM is happy to be in the console business...I mean, who wouldn't want to be in the business of manufacturing a chip that will last 4-5 years (consoles last about that long right?...and Apple is more than happy to have cool dual-core chips in its laptops.



I know I'm happy as hell with my Mac Pro. Probably the best purchase I've ever made...one year and zero kernel panics later, I only have good things to say about Apple's switch to Intel. I wish I could have said the same about my Quicksilver G4...no it wasn't the RAM...I pretty much swapped everything out but the motherboard and the G4 chip in that computer. And when I hear stories of water-cooling systems failing in the G5 computers, I can only feel sad for the people that bought the G5s.



I don't hate IBM...IBM saved Apple when Motorola was screwing the pooch. I'm grateful for that. It's just too bad IBM did the same thing Motorola did back in the days. Apple is a tough customer, I'm sure, but I would have thought IBM would have gone along with the ride. Had IBM done that it would be the chip supplier of a resurgent computer and gadget maker...'cuz let's face it, Apple's market share is about to grow... a lot...
«1

Comments

  • Reply 1 of 26
    sandausandau Posts: 1,230member
    last i heard, the latest powerpc chip hit 5ghz or some crazy number and was competitive in wattage and heat to the intel chips....so its much faster.



    i'm glad for the intel switch, makes everything easier with windows.
  • Reply 2 of 26
    kim kap solkim kap sol Posts: 2,987member
    Quote:
    Originally Posted by sandau View Post


    last i heard, the latest powerpc chip hit 5ghz or some crazy number and was competitive in wattage and heat to the intel chips....so its much faster.



    i'm glad for the intel switch, makes everything easier with windows.



    You're probably mistaking the 'Power' serie with the 'PowerPC' serie. Power6 is at 5GHz...but I highly doubt IBM jumped from 2.5GHz to 5GHz with their PowerPC architecture. Someone feel free to correct me though.
  • Reply 3 of 26
    backtomacbacktomac Posts: 4,579member
    I think IBM will rue the day they let Apple get away. Sure Apple is a tough customer. But as you point out, Apple is clearly on the ascent. Having Apple around would help to support chip development costs. It would spread out those costs over a much higher chip volume. Now they're going to have to pass those costs on to Sony and their other customers.



    Time will tell. Clearly it was a GREAT move for Apple.
  • Reply 4 of 26
    regreg Posts: 832member
    This was the latest that I have read. http://www-03.ibm.com/press/us/en/pr...ease/21580.wss . This chip is for high and medium range servers. IBM still has not come out with a lower temperature chip other than the G4 that would work in a laptop. Then again they don't make desktop or laptop machines any more. They sold that to Lenovo . If IBM were to go back into desktop chips it would be most likely be with the Cell Chip. Its simplier design and archetecture are suppose to make it a low power and heat chip. The only problems are that they had to remove some features and that also impacts performance. Then programs, they would all have to be rewritten to optimize performance. No I don't see IBM getting back into the desktop arena anytime soon.
  • Reply 5 of 26
    I agree with the pro-change crowd, although at the time I thought it might mean the end of Apple. I was wrong, full steam ahead, Cap'n!
  • Reply 6 of 26
    MarvinMarvin Posts: 15,171moderator
    I think Apple should have moved earlier (at least to avoid the double software upgrade we've had to go through OS9->OS X, PPC->x86). When X86 chips went past 1GHz, that was the first warning. When they were 2GHz+ and Apple were still below 1GHz, they should have been ready to jump instead of trying to tell us 800MHz G4s were faster than 2GHz+ Pentiums - I know clock speed isn't a great comparison but there's a limit to how far you can take that. PCs just kept creeping up and up until eventually Apple broke 1GHz and didn't do much better. The G5 was a big jump but the chips ran hot and they only really matched x86 performance not blow it out of the water - although the throughput was great.



    The problem with the IBM chips is they were fast using Altivec but because Macs take up such a small portion of the market, developers didn't invest the effort learning how to take advantage of it so programs still ran slower on the Mac side but were getting SSE optimization on the PC side.



    IBM are still having these issues with the consoles. The chips run too hot and in the case of the PS3, the chips are too difficult to optimize code for. Plus the price is high too. I wonder how much the PS3 would have cost if it used an Intel chip with 8 generic cores. They could probably have gotten away with a quad given that the XBox360 only has 3 generic cores.
  • Reply 7 of 26
    lundylundy Posts: 4,466member
    I think Jobs wanted to go Intel from the day he took over. It was clear that being dependent on Scrotorola and/or IBM wasn't exactly a good way to stay competitive with the x86 PCs.



    And as they began to bring together OS X, the Developer Previews and the Public Beta, they kept an Intel build in parallel, not too challenging since NextStep ran on Intel anyway.



    But there was one major reason that Jobs could not move to x86 hardware until 2005.



    OS 9.



    Lots of people were sticking with OS 9 because of inertia, Quark, and Adobe among others.



    The confluence of IBM not having a laptop chip, IBM not meeting its promise of 3 gHz desktop chip, Motorola staying frozen on a 167 mHz frontside bus, and OS X gaining enough users was the trigger. Jobs unleashed the Intel switch, and everything having already been on the launching pad for a long time, the transition went very smoothly.



    Remember, he told the Motorola execs "I can't wait until we don't need you guys any more."
  • Reply 8 of 26
    backtomacbacktomac Posts: 4,579member
    Quote:
    Originally Posted by lundy View Post


    I think Jobs wanted to go Intel from the day he took over. It was clear that being dependent on Scrotorola and/or IBM wasn't exactly a good way to stay competitive with the x86 PCs.



    And as they began to bring together OS X, the Developer Previews and the Public Beta, they kept an Intel build in parallel, not too challenging since NextStep ran on Intel anyway.



    But there was one major reason that Jobs could not move to x86 hardware until 2005.



    OS 9.



    Lots of people were sticking with OS 9 because of inertia, Quark, and Adobe among others.



    The confluence of IBM not having a laptop chip, IBM not meeting its promise of 3 gHz desktop chip, Motorola staying frozen on a 167 mHz frontside bus, and OS X gaining enough users was the trigger. Jobs unleashed the Intel switch, and everything having already been on the launching pad for a long time, the transition went very smoothly.



    Remember, he told the Motorola execs "I can't wait until we don't need you guys any more."





    Good points. Especially about OS 9. Once he got the Apple herd far enough away from OS 9 he knew he could ditch ppc.



    As it turned out, that turned out to be a good time to walk away from ppc chips as x86 really evolved into a superior chip platform at about that time. Especially in mobile chips.
  • Reply 9 of 26
    Quote:
    Originally Posted by lundy View Post


    I think Jobs wanted to go Intel from the day he took over. It was clear that being dependent on Scrotorola and/or IBM wasn't exactly a good way to stay competitive with the x86 PCs.



    And as they began to bring together OS X, the Developer Previews and the Public Beta, they kept an Intel build in parallel, not too challenging since NextStep ran on Intel anyway.



    But there was one major reason that Jobs could not move to x86 hardware until 2005.



    OS 9.



    Lots of people were sticking with OS 9 because of inertia, Quark, and Adobe among others.



    The confluence of IBM not having a laptop chip, IBM not meeting its promise of 3 gHz desktop chip, Motorola staying frozen on a 167 mHz frontside bus, and OS X gaining enough users was the trigger. Jobs unleashed the Intel switch, and everything having already been on the launching pad for a long time, the transition went very smoothly.



    Remember, he told the Motorola execs "I can't wait until we don't need you guys any more."



    the G5 had a much better BUS then 167 mHz
  • Reply 10 of 26
    lundylundy Posts: 4,466member
    Quote:

    the G5 had a much better BUS then 167 mHz



    Yes, Joe12345678, I know that. I also said that IBM had no laptop chip, which meant that Apple had no chip for their laptops except the G4. And Jobs knew the Intel roadmap, and that it was going to blow the G4 away very soon. They had to flush the G4, and there was no G5 to replace it. The G5 wasn't any faster than the G4 anyway, except for the frontside bus.



    It was fortunate that Apple was able to drag out the Motorola junk as long as they did - dual processors when running OS 9 didn't speed anything up - they were solely a marketing tool, but it's all they had. They could not switch to Intel until enough customers were on OS X, and Adobe and Quark were making that tough.



    That is why Jobs opened the WWDC with the "funeral" of OS 9, complete with eulogy, casket, and the casket descending into the floor after the lid was closed. We didn't really get the importance of it then, but now it's clear why he was pushing so hard for people to get off OS 9.
  • Reply 11 of 26
    jeffdmjeffdm Posts: 12,951member
    Quote:
    Originally Posted by backtomac View Post


    I think IBM will rue the day they let Apple get away. Sure Apple is a tough customer. But as you point out, Apple is clearly on the ascent. Having Apple around would help to support chip development costs. It would spread out those costs over a much higher chip volume. Now they're going to have to pass those costs on to Sony and their other customers.



    I think Apple might be on the ascent because they switched. I don't think IBM could justify making new chip designs. They had made notebook G5s, but they ran at 1.4 to 1.6GHz and were single core only.



    They can't spread the cost of development around between different products very well because each company needed different chips. The Cell is a huge vector engine and the XB360 chip took out complicated instruction optimizations, so they don't do much for general purpose computers. The consoles might sell as many as 100 million units with little change other than maybe minor fixes and die shrinks. Apple's needs would require chip updates every 5-10 million units to keep up with Intel.
  • Reply 12 of 26
    jeffdmjeffdm Posts: 12,951member
    Quote:
    Originally Posted by Marvin View Post


    IBM are still having these issues with the consoles. The chips run too hot and in the case of the PS3, the chips are too difficult to optimize code for. Plus the price is high too. I wonder how much the PS3 would have cost if it used an Intel chip with 8 generic cores. They could probably have gotten away with a quad given that the XBox360 only has 3 generic cores.



    Cell and what the has XB360 are very different. Still, I think both machines run their chips near 4GHz, so you should see why they are running a little hot. I think the chips are monsters in terms of die size and probably not going to be cool enough to run in a notebook even if they were down clocked below the speeds of a notebook.
  • Reply 13 of 26
    sc_marktsc_markt Posts: 1,401member
    I'd like to know the story about the motorola G5. I remember reading that the chip had really high performance but was blowing up inside of test boxes. I have no idea if this is true or not but I'd like to know.
  • Reply 14 of 26
    If Apple had stuck with IBM it would not be surging in every direction.
  • Reply 15 of 26
    jeffdmjeffdm Posts: 12,951member
    Quote:
    Originally Posted by sc_markt View Post


    I'd like to know the story about the motorola G5. I remember reading that the chip had really high performance but was blowing up inside of test boxes. I have no idea if this is true or not but I'd like to know.



    I've never heard of it, and it seems very unlikely. I don't know what a chip can do to explode, they might melt and smoke out.



    PA Semi was supposedly working on a notebook capable 2GHz dual core chip, but I don't know if that came out.
  • Reply 16 of 26
    regreg Posts: 832member
    I heard of the Motorola G5 had problems with production and was never mass produced. About that time Apple dumped Motorola for IBM, Motorola spun off its semiconductor unit as Freescale, which did well after awhile. Fires or melt-downs did not happen or were at least not wide knowledge.
  • Reply 17 of 26
    lundylundy Posts: 4,466member
    There was a G5, but as mentioned it ran so hot they abandoned the project. Motorola's engineers were being hired away by companies with management that actually knew what it was doing.



    There was a time when Moto couldn't even make G3s right and Apple had to get IBM to make those. Then there was the infamous 500-megahertz wall with the G4, where Apple had to downgrade an already-announced Mac to 450 mHz since there was not going to be a 500 mHz ready.
  • Reply 18 of 26
    backtomacbacktomac Posts: 4,579member
    Quote:
    Originally Posted by JeffDM View Post


    They can't spread the cost of development around between different products very well because each company needed different chips. The Cell is a huge vector engine and the XB360 chip took out complicated instruction optimizations, so they don't do much for general purpose computers. The consoles might sell as many as 100 million units with little change other than maybe minor fixes and die shrinks. Apple's needs would require chip updates every 5-10 million units to keep up with Intel.



    I'm not a chip engineer so what I suggest may not or probably isn't possible. But I don't know why IBM didn't develop a chip with a basic micro architecture and tweak it for uses in desktops, laptops and game consoles. Sort of what Intel has done with the core 2 chips. The only application they've not used it for is game consoles but I wonder if it would be possible.
  • Reply 19 of 26
    Intel = Boot Camp / Windows natively.



    HUGE switch/sales factor there.
  • Reply 20 of 26
    marcukmarcuk Posts: 4,442member
    I've always wondered, say we took the original 250nm G4 (7400 chip) and made it on todays 65nm process, what sort of speeds would it likely run at?
Sign In or Register to comment.