Slashdot Mirror
2005 Good Year for Power Architecture
An anonymous reader wrote to mention an IBM article looking back on what the piece calls the best year ever for the Power Architecture. From the article: "While IBM is considered by many to be an 800lb. gorilla, in the microelectronics space, it is actually very small -- last year IBM was way down at number 21 on the iSuppli list of the top 25 semiconductor suppliers worldwide. Now, that isn't necessarily a bad thing: for instance, it means that IBM Semiconductor solutions is small and nimble and competitive -- and this agility (coupled with the fact that we do get to share Research and some other resources with the parts of IBM that are 800lb. gorillas) has led some to predict that the IBM chips division will be named the fastest growing semiconductor supplier of 2005. In fact, there is a very good chance that IBM may regain the coveted #18 spot on iSuppli's list this year!"
"there is a very good chance that IBM may regain the coveted #18 spot on iSuppli's list this year!"
Whats so special about spot 18?
Viable Slashdot alternatives: https://pipedot.org/ and http://soylentnews.org/
Apple was/is a relatively minor customer. The business from next-gen consoles (all of which use POWER chips) is/will be at least an order of magnitude larger than Apple.
It's hard to be religious when certain people are never incinerated by bolts of lightning.
Well, in terms of volume the XBox 360, PS3, and Revolution are going to get the Power architecture in a heck of a lot more homes than Apple ever could have.
I don't have statistics to back it up, but I can prettymuch guarantee that with the release of the Xbox360 and PS3 both using POWER based chips, Apple aren't IBM's biggest customer.
My 3D Texturing Skinning work (under construction)
This article is FROM IBM about IBM. The article is what is quoted and of course they're talking about themselves, so "WE" is a perfectly acceptable term to use in their own article.
Now, it may not seem newsworthy to some to put a navel gazing press release up on the front page, but some will find it interesting to get a glimpse of part of IBM's internal workings. Take it or leave it, the article is not some kind of Google ranking ploy.
A lot of printers use Power architecture (there's one near me, with a 600MHz processor, right now, and the speed with which it renders a full color A4 PDF is quite impressive.) Power is very good wherever there isn't a load of dead weight to keep supporting, which is why it seems to do so well in the embedded or non-"PC" market. As for Apple's decision - well, I fortunately don't have any shares in Apple. I'm not convinced that they will be able to make the world's best X86 portables, and that is the task they seem to have set themselves. I have an attic full of old Macs, and I now have no reason at all to acquire another one.
Pining for the fjords
"Well, consider that only 400k or so XBoxes have shipped and 0 PS3s have shipped, and Apple shipped 600k iMacs and PowerMacs in Q4 2005 (both of which use Power based G5s), Apple is STILL IBM's biggest customer.
.05%? Heck some vendors just NOW started to fully support OSX! RedHat Linux Desktop and SuSE Linux on the destkop will have more clients that OSX on X86. Perhaps Ubuntu will even have more desktops! I wish Apple well, but it will be far far easier on them to reign in their sales guys and send out a message to ANYONE to help support OSX on X86, be that open source apps and or Java applications.
Now Sony, Microsoft, and Nintendo may have potentially larger markets, but right now Apple is still shipping more G5s than Microsoft, Sony, and Nintendo combined, that and each G5 is higher margin than a Cell based CPU."
Ok, I don't know where you get 400k. I hope that you are correct, but from what I here there will be around 3 million 360's sold by July of this year if they hold their current pace. Now as far as PS3's go, it would be a fair estimate to say that they will be close to their 90 million PS2's sold in 5 years. So lets just say they average 10 million a year for the next 5 years of PS3's. Now you also didn't mention all the hub manufacturers out there like Cisco. They also use power chips. When you start to add it up, the loss of Apple for the gain of Microsoft and Sony was huge win for IBM.
So in short Apple switching to Intel is a small loss for power, BUT there are some serious advantages now for Power as well. Microsoft and Sony will not require a new fab to be built for ~5 years! Cisco cares about the performance of the Power chip, but they care more about the power consumption. Intel on the other hand needs to focus on mobile chips, desktop chips and server chips. They will need to build their 65nm fabs as fast as possible then spend billions on the next version. Granted Intel spends around 10Billion a year in R&D so they can handle it as long as Wintel desktops keep selling like they do.
Now the real quesiton is why did Apple switch? This is off topic, but it doesn't make sense that they switched to a 32bit chip FROM a 64bit chip. Now if they would have switched to x86-64, that would have made better sense. I do feel for all those poor saps who are going to buy PPC or X86 Mac in the next year or so. I feel for them because I went through the 68k to PPC migraiton and bought all the load of crap that Apple fed back then. It is the same load of crap they are trying to spin today.... fat binaries.... blah blah blah. The truth is that if you go with a new X86 machine a bunch of your old stuff will not work and you will be praying that someone will code a new version that works as well as your old one did. Now after this painfull migration is over in a couple of years and if Apple moves to X86-64, then they will probably be better off. They just better pray (not that many in Apple belive in God), that sales of their Ipod don't faulter.
Apple has some other issues that they need to address. One is their view of open source. They use it a TON in their OS yet their sales guys go around and bash it whenever they get a chance. The next is their view on technologies such as Java. Again their sales guys go around and bash it, yet all their Java apps will not need to be ported to this new architecture. One sales guy just recently said to me that he would NEVER load any JVM on his system because he hates slow Java applications. I then informed him that it was loaded by default. He smiled and told me he knew that and he was "just kidding", but I have to wonder how many poor saps he talks to that don't know that and then repeat the same crap he just said. Lastly, and perhaps the biggest issue is why on earth would our development shop write software for X86 Macintosh? Lets look at the marketshare as it is and as it will be for the next few years. What percentage of all new desktops do you believe will be running X86 OSX?
I
The more I learn about science, the more my faith in God increases.
Umm, no, Apple was never anywhere near their biggest customer. Apple was only notable for building PCs with power chips, but the vast majority of power chips never went into PCs. Apple was a rather small customer, and one that was constantly demanding special treatment.
=-=-=-=-=-=-=-=-=-=-=-=-=-=-
Friends don't let friends enable ecmascript.
in the microelectronics space, it is actually very small
I thought everything in the microelectronics space was very small...
Apple was never IBMs biggest customer, PPC is everywhere, cars, routers, smart devices of all types, granted that Freescale has the majority of these markets, but IBM has more than enough share to outweight Apple. I'll grant that Apple was IBMs highest-profile customer, but I've seen numbers that place revenue generated by Apple at fewer than 1% of IBMs business. Not only that, but they're going to be manufacturing CPUs for all of the next generation game consoles. The Xbox 360 has already sold more units than Apple would have over the ~4 years or more (keep in mind that the only IBM processor in the Macs is the G5, The G4s are still provided by freescale.) How many PS2s, Xboxes, and Gamecubes have been sold worldwide? 150+ million combined, and that number is growing with each new generation (plus we're not even at the end of this generation's sales!) So IBM is looking at basically guaranteed sales of ~200 million PPC-based parts over the next 5-6 years. Apple doesn't even drive PPC development, its driven on both ends by the embedded and big-iron/server markets.
IBM won't be hurting by loosing Apple, neither will PPC. I am a bit sad to loose the only "mainstream" consumer-level PPC hardware, but there are some smaller PPC manufacturers such as Genesi to provide for the hardcore PPC fan market.
Personally I thought it was a particularly good year for Sun's Sparc processors - see this Forrester research article for example. Here are some recent Sun SPECjbb performance benchmarks against IBM's Power P5.
But since Sun isn't a leading Linux advocate, I don't expect them to get Slashdot front page coverage like IBM seems to...
Zen tips: Pay attention. Don't take it personally. Believe nothing.