What news, revelations, rumors, or analyses have you seen to support or contest Adrien's thoughts? Let's talk about it!
You must be logged in to the O'Reilly Network to post a talkback.

Showing messages 1 through 41 of 41.

• Missing the point about Core Duo
  2006-04-03 11:04:36  nolaMatt [Reply | View]
  
  
  One thing about the new MacIntels that seems to be ignored here is the staggering number of Photoshop users who are still running old G4 Macs. It is an absolute fact that most Mac users upgrade their machines far less frequently than PC users (this also helps contribute to the Mac's "low market share"). "Set it, and forget it!" indeed.
  
  I recently upgraded from a 1.25GHZ G4 eMac to a 2.0 GHZ MacBook Pro. Quite honestly, while I haven't bothered to waste my time sitting around with a stopwatch benchmarking my Photoshop performance, it "feels" like Photoshop runs approximately equal to or somewhat better than my old machine.
  
  Meanwhile, almost all of the other apps I use on a regular basis have already been updated to Universal Binaries. So for me, other than Photoshop and MS Word, I am experiencing a STUNNING increase in performance with the MacBook. When the Universal version of Photoshop is released (which apparently will be part of Adobe CS 3), I'll upgrade. MS Office runs fine. A little laggy on the scrolling, but no big deal. For now, it's OK with me.
  
  If I had been using a dual or quad G5, then I suspect I'd be disappointed with the MacBook's Photoshop performance. Then again, I would have been disappointed with the PowerBook G4's performance as well!
  
  Basically, if you're currently using a top-of-the-line (or close to it) PowerPC Mac, and you are a heavy user of Adobe apps, stick with your current machine until Adobe gets around to updating the software.
  
  If you're using an older Mac, and/or you're more of a "casual" or occasional Photoshop user, go ahead and upgrade to an Intel Mac now. You won't notice much if any of a performance hit on the Adobe apps (if your current Mac is old enough, you may well see a performance *increase*), and you'll LOVE the performance increase you'll get on everything else.
  
  The best part of the Intel switch is that it ends the "megahertz wars" once and for all. No longer will Mac users have to suffer all of the ill-informed "dOOd, Mac sux. Mac has only 1.25 GHZ? My P4 has 3 GHZ!!! LOLOLOLOL!!!!" garbage. Now Apple has access to the same processors as the PC guys. Intel upgrades the processor? Great! Order some and send them over to the factory in China!
  
  The worst part of the Intel switch is that never again will Jobs be able to get on stage and say "Check out our new processor...this absolutely CRUSHES the Intel PC". FWIW, *sometimes* this was true when the PPC chip made a particular breakthrough. Unfortunately, Intel would usually catch up and surpass the PPC chip pretty quickly, but the "party line" out of Apple would keep up the "Pentium-Crushing Performance" blather, even if it was no longer true.
  
  As far as the Cell chip goes...there is one and only one question that has any bearing on whether or not Apple will need to worry about and/or begin using the chip in its computers:
  
  "Does it run Windows?"
  
  As long as the answer to that question is "NO", then Apple has nothing to worry about from the competition. Apple will continue to have access to the fastest processors used in the general PC market, and that will be Good Enough™.
  
  If the Cell ends up being "all that," and Apple decides to use it in order to crush their enemies, see them driven before them, and hear the lamentations of their women, it appears to me that they will have a much easier time of switching to the Cell processor than Microsoft, since they already seem to be moving towards a processor-independent development environment.
• Missing the point about Core Duo
  2006-04-04 14:51:03  AdrienLamothe [Reply | View]
  
  
  "... it appears to me that they will have a much easier time of switching to the Cell processor than Microsoft, since they already seem to be moving towards a processor-independent development environment."
  
  -- Absolutely. Any Unix-based operating system, Mac OS X included, will be easier to port to Cell than MS Windows.
  
  "As far as the Cell chip goes...there is one and only one question that has any bearing on whether or not Apple will need to worry about and/or begin using the chip in its computers:
  
  "Does it run Windows?" "
  
  -- Not necessarily. The Cell appears to offer video performance vastly superior to anything else, and the new Nvidia GPU isn't even available for it yet. Photoshop power-users can definitely benefit from such video performance. For the casual Photoshop user, it shouldn't make much of a difference.
  
  "The best part of the Intel switch is that it ends the "megahertz wars" once and for all. No longer will Mac users have to suffer all of the ill-informed "dOOd, Mac sux. Mac has only 1.25 GHZ? My P4 has 3 GHZ!!!"
  
  -- Definitely. The PowerMac-Intel comparisons were not very useful anyway, because Power is a RISC architecture and Intel is CISC. AMD was the first company to work at educating the public about the fallacy of MHz comparisons.
  
  Cheers.
  
  

• Drivel
  2006-03-31 12:57:14  dborod [Reply | View]
  
  
  This article is a mish-mash of corporate press releases and other drivel. What a bunch of hogwash.
• Drivel
  2006-03-31 20:12:45  AdrienLamothe [Reply | View]
  
  
  Thank you for your carefully crafted arguments. We now have a much better idea of where you're coming from.
  
  

• Orders of magnitude
  2006-03-29 11:28:31  jimothy [Reply | View]
  
  
  "Marketing talk says that PlayStation 3 will be 50 times faster than PlayStation 2. It is undoubtedly significantly faster (more likely by orders of magnitude) than Apple's new Core Duo systems."
  
  Now that, I really doubt. You're saying that the PS3 will be at least 100 times faster than Apple's new systems (one order of magnitude is 10 times faster, two orders of magnitude is 100, etc.)? If the PS3 is 50 times faster than the PS2 (likely an exaggeration in itself), then the PS2 must be at least twice as fast as Apple's new systems. I really don't think so.
  
  Perhaps you meant to say the PS3 will be several times faster than Apple's new Core Duo systems? Even that might be an exaggeration, but we will have to see.
• Orders of magnitude
  2006-03-30 17:10:02  pquam [Reply | View]
  
  
  On certain applications, its performance is astounding. Take a look at Mercury computing's promotional video about using it for radiology:
  
  http://www.mc.com/cell/demo.cfm
  
  Try doing that with a core duo mac!
• Orders of magnitude
  2006-03-29 12:06:18  AdrienLamothe [Reply | View]
  
  
  That is what the marketing talk says. The greatest speed increase will be for video performance, which is what the Cell is optimized for.
  
  At this point, what seems to be hampering the Cell is unavailability of the Nvidia graphic processor. Early demos have been conducted without it (and have still been impressive). Sony funded Nvidia to develop a new graphic processor specifically for the Cell. Personally, I think IBM and Sony should have contacted two or three video card companies and used all of them as sources, to avoid putting all of their eggs in one basket. It may make sense for IBM to aquire a video company; Matrox stands out as a good candidate (Matrox's 3D stuff would need some work, but that's feasible).
  
  
  We'll just have to wait and see what happens. I'm looking forward to it. If the Cell turns out to be as fast as expected, word is that other microprocessor companies will immediately copy the architecture.
  
  Cheers.
  
• Orders of magnitude
  2006-03-30 18:11:04  pquam [Reply | View]
  
  
  Sun has its multiprocessing ultrasparc T1 out which was just open-sourced. It was supposed to do for servers what Cell does for multimedia, but failed to outperform Athlons in most situations. This may change though. There will probably be mainstream quad core processors before too long.
  ...
  If chip-makers were really smart, they would implement specific functions in FPGAs, and have them be reprogrammed as needed. Linked together, FPGAs can give magnitudes more processing power than etched silicon and their abilities have been doubling every year for the past five years. They are no longer limited to the embedded world.

• Cellular phones
  2006-03-28 07:37:24  oddhs [Reply | View]
  
  
  "The Cell may bring new possibilities to cellular phones, a potentially large market for the chip."
  
  Do you have any idea of the power requirements of the Cell chip? To set things in perspective: There's a good reason they use ARM chips and not Pentium 4 chips in cell phones.
• Cellular phones
  2006-03-29 11:50:10  AdrienLamothe [Reply | View]
  
  
  There has been talk of IBM offering different models of the Cell, with lower clock speeds for the PPE and reduced numbers of SPEs, that are tailored for particular environments.
  

• OSDL desktop environment?
  2006-03-28 05:10:42  ahmedsaad [Reply | View]
  
  
  "Torvalds is in charge of OSDL's Desktop Linux initiative, which is scheduled to release a desktop environment this summer."
  
  an alternative to KDE/GNOME/.. ? couldn't find anyting on osdl.org, any links?
  
  
• OSDL desktop environment?
  2006-03-29 11:54:55  AdrienLamothe [Reply | View]
  
  
  If you go to http://www.osdl.org and look at the menu area on the left side of the page, you will find a menu item titled "Desktop Linux". That will take you to, among other things, the Desktop Linux Working Group.
  

• Cell's PPE is not at all related to the POWER5
  2006-03-27 21:24:23  riskin [Reply | View]
  
  
  I'm not sure where you came up with the statement that the Cell’s PPE is derived from the POWER 5, but in your own link to Anandtech's article on the cell it states "The PPE is a new core unlike any other PowerPC core made by IBM."
  
  It is a very simple, in-order execution processor with a high clock speed to help offset this in-order design. If it is derivative of any IBM design, it is certainly not the POWER 5.
  
  For a detailed analysis of the Cell, I recommend Jon "Hannibal" Stokes' articles here:
  
  Part 1: http://arstechnica.com/articles/paedia/cpu/cell-1.ars
  
  Part 2: http://arstechnica.com/articles/paedia/cpu/cell-2.ars
  
  
  As mentioned by a poster above me, you seem to be downplaying the difficulty of extracting performance from the Cell for general computing tasks. Yes, the Cell is a media processing beast and has the potential to be a boon in certain computing tasks. However, the majority of code run on personal computers (you know, the ones that Toshiba, Sony, and Apple all make) does not benefit at all from the additional SIMD processing elements in the Cell. So the only thing you're left with for day to day computing (web browsing, word processing, chat, email, book keeping, and so on) is the relatively weak processing power provided by the Cell's PPE. Granted, if you wanted to encode one of your DVDs into a format suitable for your Video iPod, the Cell would whip through it like a bulldozer through cheese.
  
  Quite simply, the Cell is not in any way well suited to general purpose computing at all. Yes, it may suffice in that role and under certain conditions, but every application of it deals with processing massive amounts of data: games, scientific and high performance computing (supercomputing), media processing, and other workstation (aka task-specific machines) environments. I'd love to see a Cell co-processor paired with a general purpose CPU. We see such pairings of powerful, highly parallel media processors with general purpose CPUs all the time. They're called video cards. They are great at what they do, but if you tried to run an OS on a graphics chip you’d kill yourself.
  
  I think you're pretty far off the mark to suggest that Apple has to fear the Cell being adopted by computer makers as a central CPU. In fact, I think your argument that Cell-based devices such as a Linux powered PS3 will do anything to impact Apple or Microsoft on the desktop. The Cell is just not designed to be a general purpose CPU, and attempting to argue otherwise misses the point.
  
  However, you are correct that the PS3 is an excellent vehicle to expose a large number of people to Linux.
  
• Cell's PPE is not at all related to the POWER5
  2006-03-30 17:39:12  pquam [Reply | View]
  
  
  If by general purpose computing, you mean email, internet, the office suite, etc., then all you need is a PIII and some fast drives. In other words, the excitement about general purpose computing is dead. The only way to make general purpose computing fun anymore is to pack it into something that weighs 6 oz., or give it unbelievable performance per watt.
  
  At up to 4ghz with absolutely phenomenal I/O, Cell is more than fast enough for this stuff; all of the overhead generated by techniques to try to feed the processor (like out-of-order processing) continuously actually slowed down the P4 and its sucessors, in my opinion. The consensus in the processor design community is that chips had become unnecessarily complicated with these techniques, and that they didn't pay enough attention to I/O. Cell was specifically designed to remedy that.
  
  Cell is a specialized processor, and its specialty is in what consumers care most about: multimedia. This was also Apple's specialty until recently, which is what I assume led to this article's focus on Apple. Cell has over 200 fully programmable gigaflops, which is like 30 times more than your average processor these days (or about 3 times more than a quad G5). Cell is a consumer electronics dream.
  
  
  
• Cell's PPE is not at all related to the POWER5
  2006-03-29 13:08:11  AdrienLamothe [Reply | View]
  
  
  Digital media is where the commercial computing market is heading, it is where most of the action will be. Apple has a core group of customers, who have been with Apple a long time, who use Macs primarily for digital photos, sound and film. Cell is optimized for digital media and computational work; this makes it very attractive to people who work with digital media. You can connect the dots from here.
  
  Regarding the Cell processor's roots in the Power5 architecture, perhaps it is best to quote an article in the IBM Journal of Research and Development, authored by the Cell architects. You can find this article at:
  
  http://www.research.ibm.com/journal/rd/494/kahle.html
  
  Three sentences stand out:
  
  1. "Indications were that a completely new architecture can easily require ten years to develop, especially if one includes the time required for software development. Hence, the Power Architecture* was used as the basis for Cell."
  
  2. "By the end of 2000 an architectural concept had been agreed on that combined the 64-bit Power Architecture* [4] with memory flow control and “synergistic” processors in order to provide the required computational density and power efficiency."
  
  3. "The Broadband Processor Architecture extends the 64-bit Power Architecture with cooperative offload processors (“synergistic processors”), with the direct memory access (DMA) and synchronization mechanisms to communicate with them (“memory flow control”), and with enhancements for real-time management."
  
  Regarding the difficulty of extracting Cell processor performance from a programming perspective, a large effort has been underway to provide developers with the necessary programming tools. Intel has been doing the same thing for their Core Duo architecture. A muti-core, multi-processing environment is about to become the new norm; those who refuse to adapt will suffer the same fate as others in the past who didn't change with the technology. If you want to develop relatively simple applications, then you will likely be able to continue using your current programming model. If you program on top of engines such as Apache, then you also won't likely need to worry about it; the Apache developers will port to the new environment and you won't have to deal with it.
  
  Cheers,
  
  Adrien
  
• Cell's PPE is not at all related to the POWER5
  2006-03-30 21:37:14  ohfoo [Reply | View]
  
  
  This comment seems accurate to me. The cell's controlling core is described as a 970 derivative, unlike the other cell "cores." I find it ODD that people think Apple has left behind the Power line of processors. They have picked up the Intel, but they haven't dropped the Power line, and have no economic reason to. They are well postioned over both lines now, with OS X and "universal" binaries. Jobs never said "we're abandoning all IBM options." The main article we're commenting on misreads the limitations on Apple, it seems to me. Apple can "wait and see" on IBM developments.

• Did you get lost on the way to PC Magazine?
  2006-03-27 18:54:03  lurchvt [Reply | View]
  
  
  Rambling text that almost seems focused...check
  Arguments that resemble "The sky is blue, so blue chips are sure to rise"...check
  Harping on the eventual demise of Apple...check
  Arguing for opposite actions (i.e., Apple supporting two architectures is bad for business...Apple should switch architectures again)...check
  
  Dvorak, is that you?
  
  P.S. I subscribed to PC Magazine for a couple years, and it was reasonably good, but Dvorak's column always seemed like it belonged at the supermarket checkout aisle next to "500lb. woman gives birth to alien twin of Stalin".
• Did you get lost on the way to PC Magazine?
  2006-03-29 12:09:46  AdrienLamothe [Reply | View]
  
  
  No, this isn't Dvorak.
  
  Steve, is that you?
  
  
  
• Did you get lost on the way to PC Magazine?
  2006-03-29 13:46:07  AdrienLamothe [Reply | View]
  
  
  You know, I had hoped some serious discussion on the issues would result. Such as, "Hey, how can Apple manage processor migration while minimizing partner and customer discomfort?" Instead we get drivel.
  
  Where did I predict the demise of Apple? I don't think you read the article.
  

• Universal Binary is not a certification program
  2006-03-24 08:15:31  jdodds [Reply | View]
  
  
  Universal binary is a file format for executable files, not a certification program.
  
  A universal binary supports multiple architectures, e.g. PPC and Intel, in one file. When a program is shipped as a universal binary there are not "two copies of [the] program in each box." There is one copy that can run on either architecture.
  
  There is a universal logo program for promoting software delivered as a universal binary but that shouldn't be confused with what a universal binary is.
  
  Apple's Xcode development tools support building universal binaries. It's a build configuration option in the Xcode UI.
  
  Generalizations about universal binary uptake among software vendors can not be made based solely on Adobe. Many developers were shipping universal binaries for months before the January MacWorld Expo.
  
  Adobe's Mac OS development is tied to the Metrowerks tools. Metrowerks is now part of Freescale and doesn't appear to even offer Mac OS X development tools anymore. To ship universal binaries Adobe first needs to migrate from Metrowerks to Xcode.
  
  Metrowerks was a hero in the 68K to PPC transition but in terms of Mac OS support they have been in decline for a long time. I'm sure there are other vendors in the same situation as Adobe but I'm also sure there are many more that recognized Metrowerks' decline and heeded Apple's advice to move to Xcode.
• Universal Binary is not a certification program
  2006-03-24 15:56:33  AdrienLamothe [Reply | View]
  
  
  Universal Binary is indeed an Apple branding and certification program. The technical details are but one aspect of it (albeit the most important.) Check out http://developer.apple.com/softwarelicensing/agreements/maclogo.html for an overview. Software developers have to sign a licensing agreement before they can use the Universal Binary logo on their products.
  
  Regarding the fact that both architecture's binaries are contained in the same file: It is correct to express that as "two versions of the program come in the same box" (or more accurately, on the same CD.)
  
  Apple has done a good job leveraging open-source technology for Xcode. It still isn't an accepted international standard, which means you are dependent on Apple (or whoever) for future enhancements. There is nothing wrong with that, if you don't mind putting your coding effort in the hands of other people. That is the trade-off you accept when you opt for a multi-platform development tool.
  
• Universal Binary is not a certification program
  2006-03-25 09:49:32  jdodds [Reply | View]
  
  
  You're still confusing the logo program with a file format. Yes, developers need to sign a licensing agreement to use the logo. But I don't need to sign anything to create a universal binary. And licensing the logo doesn't involve any certification by Apple.
  
  I thought I understood your point about Xcode up until the last sentence. Then you lost me. I don't know what you mean by:
  
  That is the trade-off you accept when you opt for a multi-platform development tool.
  
  Did you mean single platform instead of multi-platform?
  
  I thought you were making a point about open source tools versus proprietary tools.
• Universal Binary is not a certification program
  2006-03-29 12:33:15  AdrienLamothe [Reply | View]
  
  
  When you opt to use a multi-platform development kit, you have to utilize a new set of system calls in your code, link to new libraries, and in some cases utilize a run-time engine. Unless you plan to maintain the kit yourself (and have the right to do so,) you are dependant on other people for enhancements, bug fixes and insuring the kit will work for future versions of the target operating systems.
  
  Trolltech, developer of the QT multi-platform SDK, actually open-sources their code under GPL. If you use the kit for commercial purposes, then you pay Trolltech a licensing fee, otherwise their kit is free to use. If Trolltech ever goes out of business, people can continue to enhance the kit, because they have the source code.
  
  I've been in a situation where a company had based a critical piece of their software infrastructure on proprietary middleware. The middleware vendor was then sold to another company, at which point the brilliant founder, who was also the lead developer, left the company. From that point on, no updates were released for the product. The source code wasn't available. Worse, not having updates prevented the user from upgrading to new hardware and operating system versions. And they couldn't just throw the system away, their business depended on it. Overall, a bad situation to be in.
  
  Cheers,
  
  Adrien
  
  
  
• Universal Binary is not a certification program
  2006-03-30 06:53:43  jdodds [Reply | View]
  
  
  Xcode is an IDE. It's not a 'multi-platform development kit.' It is not an alternative to something like Qt. But, as Trolltech's website notes, Xcode is a supported IDE for Qt/Mac development.
  
  Adobe put themselves in a tight spot by not recognizing (or at least not reacting to) the handwriting on the wall regarding Metrowerks. Adobe's dependence on Metrowerks' IDE and libraries is similar to your story about the proprietary middleware.
  
  I don't think Xcode represents the same risk. First, if you're targeting Mac OS X then you're already depending on Apple. If Apple loses interest in computers and becomes a consumer electronics company your investment in Mac OS X development may be a lost regardless of your choice to use Xcode or not. Second, Xcode has little or no direct competition right now. Paradoxically that's good. Apple needs developers so Apple needs to support good developer tools.
  
  You are always "putting your coding effort in the hands of other people." Even when you are using open source. The potential advantage of open source is that by providing more options the risk factors are changed.
• Universal Binary is not a certification program
  2006-03-30 10:27:12  AdrienLamothe [Reply | View]
  
  
  Quoting one of your earlier statements:
  
  "Apple's Xcode development tools support building universal binaries. It's a build configuration option in the Xcode UI."
  
  Your earlier assertion was that software developers can smoothly migrate to whatever new microprocessor Apple uses, simply by using Xcode. Do you still assert this?
  
  QT also generates Mac OS X programs. Adobe has used QT in the past, to build Photoshop Album.
  
  Using open-source tools does potentially mitigate risk. Even better is when you can code to open, established standards. Linux is a POSIX-compliant Unix operating system, that is both open-source and adheres to standards. This separates it from other operating systems, that appear to be open but are in fact only open to a certain degree. Lots of software developed for other Unix platforms will compile and run on Linux; this demonstrates the power of standards.
  
  Of course, there are abstractions above the operating system, that are not official standards, so we end up developing to those de-facto standards. Today's de-facto standards can become tomorrow's official standards. So it is also good to avoid dogmatism and pragmatically utilize the best tools for the job. QT may be a candidate for future standardization, as may any new abstractions Apple can develop.
  
  Cheers.
  
  
• Universal Binary is not a certification program
  2006-03-31 06:06:19  jdodds [Reply | View]
  
  
  Xcode is no magic bullet. It won't save you if you have written non-portable code. But Universal Binary can accommodate as many processor architectures as Apple needs and it's reasonable to assume that Xcode will continue to support Universal Binary.
  
  I'm not familiar with Qt but libraries like Qt generally need to make compromises to abstract away some of the differences between the UIs they support. If you are developing to multiple UI platforms the trade-off may be acceptable.
  
  But if you are writing a Mac OS X app for Mac OS X only I don't see value in writing to Qt over writing directly to Carbon/Cocoa.
• Universal Binary is not a certification program
  2006-03-31 09:04:24  AdrienLamothe [Reply | View]
  
  
  Universal Binary will support a new processor architecture after Apple (or someone) develops the compiler for it. I believe Apple is using the GNU compiler for Core Duo. So, there is always work involved when supporting a new microprocessor. Speaking of GNU, it is amazing how we now live in a world where companies can deliver innovative products based on open-source technology. Ten years ago we couldn't have had this discussion.
  
  Apple made a smart move migrating to Core Duo, because they needed something new. I was surprised they didn't migrate to AMD, because Opteron is so good and is 64 bit. Core Duo is a better choice for notebook computers (a very important market segment.) The migration was also well timed from a marketing perspective. Apple has held on to public mind share because of constant product introduction (mostly iPods) the past couple of years. One of Steve Jobs' strengths is his initiative. Apple is proof that a smaller company with initiative can outperform larger competitors. Larger companies tend to be very risk averse and also slow to react due to the bureaucratic nature of their decision making processes. I believe Sony is in this category and is fighting hard to change.
  
  I believe QT does abstract away some differences. It actually doesn't have to, it is just a matter of how much effort goes into developing QT. Some software companies have the resources to develop their own custom QT widgets. I agree that developing to Carbon/Cocoa is the way to go if you are writing only for the Mac.
  
  
  
• Universal Binary is not a certification program
  2006-03-31 10:15:59  jdodds [Reply | View]
  
  
  I agree with you that the choice of Intel over AMD was probably driven by notebooks.
  
  There was a recent article (can't remember where) that made a case that Apple doesn't act its age in the sense that it acts more like a startup than a 30 year old company.

• Market share
  2006-03-24 04:45:59  simon_hibbs [Reply | View]
  
  
  Apple had a market share of 2.3% in 2003, so where does your figure of 10% 'a few years ago' come from? I didn't think Apple had ever had a share that big. Oh, and didn't they recently pull ahead of Dell?
  
  Simon Hibbs
  
• Market share
  2006-03-24 15:28:42  AdrienLamothe [Reply | View]
  
  
  Apple did at one point have a market share of 10%.
  
  Dell has been the #1 personal computer seller in the world for a number of years now. I think you are referring to a recent blip where Apple sold more Intel core duo systems than Dell.
  

• This is a tech analysis about a company that isn't tech-focused
  2006-03-23 23:41:29  jmproffitt [Reply | View]
  
  
  Sorry, but this whole analysis misses the Steve Jobsian point of product development. It's not about the technology itself -- it's about the end user experience.
  
  Yap all ya like about Linux and PS3 and the Cell and all that jazz. But so long as Steve Jobs or someone born of his mold is sitting at the top and maintaining a vise-like grip on product development -- and fanatically focusing that development on the end user experience -- then all the rest of the elements should fall into place and Apple will remain a leader either in design and thought (Mac OS X) or in actual market share (iPod, iTunes).
  
  The problem with Windows and Linux is that no one controls the end user experience with enough vigor and attention to detail. Windows is doing better and better with each revision, but there are still too many cooks in that kitchen. Linux has no central leadership on the end user experience because it cannot, by definition, be uniformly led or controlled by any one person or group. All that said, Apple still has some cleaning up to do in its Mac OS X camp -- the Finder and lots of their bundled applications do not yet behave uniformly or maintain a consistent look/feel. So even with Jobs cracking heads, Mac OS X still has issues, but it's way out in front of platforms with no one leading from a position of end-user focus and authority.
  
  As for the shifting-processor threat, Apple is now successfully moving its Mac OS X platform into a processor-agnostic model. Follow their mantra -- use Xcode, use Xcode, use Xcode. They know they might have to jump processors again, and probably expect to jump. Stick with their development approach (Xcode) and you'll be able to follow them wherever they go. If they have to go to Cell, so be it. If the go to the Sun 8-way package, so be it. They've gotten it right at this point -- the presentation layer is the best in the business (though it could be better still), and the development layer has been largely abstracted from the hardware layer.
  
  And puhleeze stop with the Linux predictions of growth or dominance or whatever. Linux is a great OS, but it's impossible for end users to use -- no applications, nothing is easy, and there are a billion variations. Way, way, WAY too hard for end users to handle. And that will not change because no one's in charge of it. Again, great OS, but only when it's handed to end users in a closed package (like the proposed PS3 setup).
  
  Apple license OS X? Not likely. The only way that works is if they have a massive hardware certification program or license only specific manufacturers that stick to their straight and narrow hardware path. It's easier to make your own to your own specs than to hammer out contracts with third parties (like a Dell) to make things that meet your exacting standards. Keep in mind that Apple doesn't really make any of their own hardware anymore. So why contract with Dell when you can dictate terms to a Taiwanese manufacturer and get your preferred solution, then sell it yourself and get the best margins? End user simplicity and stability is the point, and that's impossible to maintain when Dell switches hard drive, processor, memory and DVD-RW models and suppliers every week.
  
  Sure, Apple has challenges, as do the rest of the companies noted in the analysis. I just don't buy the overall concept that 2006 is Apple's high water mark and 2007 begins the decline. I can look back in history and see a lot tougher business environments in which Apple has done just fine, thank you. After all, the biggest story about the Intel transition is that it's not much of a transition at all. For the end users. And that's the point.
• This is a tech analysis about a company that isn't tech-focused
  2006-03-30 21:50:35  ohfoo [Reply | View]
  
  
  Couldn't agree more with this comment. First person that explicitly realizes the "agnostic" position implied in Apple's efforts, that the Xcode tools hover over the two main cpu-line choices. Offloading video processing from the 970 version x to the rest of Cell isn't such a big deal.
• This is a tech analysis about a company that isn't tech-focused
  2006-03-24 17:38:29  AdrienLamothe [Reply | View]
  
  
  It is best to respond to this point-by-point:
  
  "Sorry, but this whole analysis misses the Steve Jobsian point of product development. It's not about the technology itself -- it's about the end user experience."
  
  --- I agree with the end-user experience part. However, Steve Jobs has ALWAYS hyped the concept of technological excellence in his products. His public statement about switching to Intel was that (sic) "Intel has the best processor roadmap from here on out." He always talks about the technology in the products, using it as a selling point. There are other people better suited to discuss the actual end user experiences of Macintosh owners; perhaps some of them will post to this talkback. ---
  
  "Yap all ya like about Linux and PS3 and the Cell and all that jazz. But so long as Steve Jobs or someone born of his mold is sitting at the top and maintaining a vise-like grip on product development -- and fanatically focusing that development on the end user experience -- then all the rest of the elements should fall into place and Apple will remain a leader either in design and thought (Mac OS X) or in actual market share (iPod, iTunes)."
  
  --- Yes, maintaining tight control on product insures a good user experience. Several Linux vendors (HP, IBM, Penguin Computing and others) already do this in the server space. Companies will start to do so in the mass market, when the time is right. Some companies are already starting to do this in select markets, South Africa is an example (see "Related Links" at end of article) ---
  
  "The problem with Windows and Linux is that no one controls the end user experience with enough vigor and attention to detail. Windows is doing better and better with each revision, but there are still too many cooks in that kitchen. Linux has no central leadership on the end user experience because it cannot, by definition, be uniformly led or controlled by any one person or group. All that said, Apple still has some cleaning up to do in its Mac OS X camp -- the Finder and lots of their bundled applications do not yet behave uniformly or maintain a consistent look/feel. So even with Jobs cracking heads, Mac OS X still has issues, but it's way out in front of platforms with no one leading from a position of end-user focus and authority."
  
  --- True openness, like true democracy, is a messy affair, but much more desirable than the alternatives. Regarding a lack of central authority, this has recently changed. The Open Source Development Lab and Linus Torvalds have devoted themselves to insuring a good end-user experience. As far as Mac OS X being "way out in front", I'm not sure what you mean by that, and I disagree with the statement. SuSE Linux has been the leader in providing an easy to install and use Linux distro, with a very nice fit and finish. I do agree that Mac OS X is the leader for offering a polished desktop, but it also has occasional glitches that require expert attention. Computers are extremely complex machines, it is doubtful they will ever be maintenance free. The closest thing I've seen to maintenance-free computers are the Linux boxes I've built for people. I have a friend who is almost computer illiterate, who was experiencing massive problems with his Windows computers. He has been using a Linux box for over two years now, and it has never crashed or experienced any problems, it just works for him every time. ---
  
  "As for the shifting-processor threat, Apple is now successfully moving its Mac OS X platform into a processor-agnostic model. Follow their mantra -- use Xcode, use Xcode, use Xcode. They know they might have to jump processors again, and probably expect to jump. Stick with their development approach (Xcode) and you'll be able to follow them wherever they go. If they have to go to Cell, so be it. If the go to the Sun 8-way package, so be it. They've gotten it right at this point -- the presentation layer is the best in the business (though it could be better still), and the development layer has been largely abstracted from the hardware layer."
  
  --- If you feel comfortable entrusting your programming effort to Xcode, then by all means use it. I prefer established standards, such as Unix. Binding your code to a non-standard abstraction layer also binds you to the organizations that control the layer. ---
  
  "And puhleeze stop with the Linux predictions of growth or dominance or whatever. Linux is a great OS, but it's impossible for end users to use -- no applications, nothing is easy, and there are a billion variations. Way, way, WAY too hard for end users to handle. And that will not change because no one's in charge of it. Again, great OS, but only when it's handed to end users in a closed package (like the proposed PS3 setup)."
  
  --- Wow, have you been living in the jungle the past few years? Impossible to use? No applications? Nothing is easy? A billion variations? No one in charge of it? Get serious. Do you work for Microsoft? I do agree that most Linux distros can be hard for the average user to install and configure. This is where, at some point, an enterprising computer vendor will decide to build a solid consumer market Linux box and offer support (this is starting to happen.) ---
  
  "Apple license OS X? Not likely. The only way that works is if they have a massive hardware certification program or license only specific manufacturers that stick to their straight and narrow hardware path. It's easier to make your own to your own specs than to hammer out contracts with third parties (like a Dell) to make things that meet your exacting standards. Keep in mind that Apple doesn't really make any of their own hardware anymore. So why contract with Dell when you can dictate terms to a Taiwanese manufacturer and get your preferred solution, then sell it yourself and get the best margins? End user simplicity and stability is the point, and that's impossible to maintain when Dell switches hard drive, processor, memory and DVD-RW models and suppliers every week."
  
  --- Ah, we agree on something. Although the PC industry does a pretty amazing job of producing stable computers; that's because the various manufacturers agree to and for the most part adhere to industry standards. ---
  
  "...I can look back in history and see a lot tougher business environments in which Apple has done just fine, thank you."
  
  --- Apple will continue to do well, by maintaining a low head count, tight control, and high prices. We'll see if they can reach out to the mass market. ---
  
  "... After all, the biggest story about the Intel transition is that it's not much of a transition at all. For the end users. And that's the point."
  
  --- Tell that to the people I meet who would like to buy a Mac Core Duo system, but don't want to buy new copies of Photoshop. ---
  
  Cheers, it's been fun.
  
• This is a tech analysis about a company that isn't tech-focused
  2006-03-24 19:42:31  jmproffitt [Reply | View]
  
  
  1. Of course Jobs touts technology in his new product introductions. Consider the audience. The mass market computer users don't pay attention to those announcements (I do, but my wife doesn't, for example). The mass market sees iPod advertisements, not iPod introductions. Ditto for Mac OS X stuff, which is almost never advertised anymore. Jobs talks tech to techies, but he ensures the products don't. Look at any manual that comes with a new Mac. Or read the online help. It's a stark contrast from the product introductions and focus on technology you're proposing.
  
  2. Linux's only hope as a mass market operating system is through the help of a major manufacturer that develops a custom distribution and controls it tightly. That's entirely possible, but it's unlikely at best. Any custom distro of Linux will not be "true Linux" and will still not have sufficient applications for the mass market. I can go to Wal-Mart and buy software for a Windows box -- games, silly little address books, word processors, electronic encyclopedias and more. I can't buy Linux software (or at least not much) at Wal-Mart. I can't even buy Mac OS X software at Wal-Mart.
  
  Now, assuming even a big player like HP built a custom Linux distro, they'd also have to have a custom hardware platform and a custom set of core applications. They would have to replicate Apple's work with Mac OS X, essentially. I don't believe that there's anyone at HP (or a similar player) will have the guts to do this -- it would take too much faith and it would by definition cannibalize Windows sales and sour the relationship with Microsoft. It just won't happen at a big shop. And a little shop won't have the visibility or leadership position to pull it off. That's why you've built a nice distro of Linux for your computer illiterate friend, but despite your hard work, you're still not in charge of a multi-million or multi-billion dollar company making Linux distributions.
  
  3. I'm totally with you on the "openness is messy" concept. That's Linux's strength and weakness. It can be custom-built into a bulletproof appliance-like setup and it's slick. But that kind of setup lacks Windows' advantages of universal "compatibility" and availability. They're just at opposite ends of the spectrum there, and neither one is the "right" way to do computers.
  
  4. You mentioned the problem with Xcode being a nonstandard development environment. I agree -- it's not a universal standard. However, neither is Visual Studio. Nor is Eclipse. Indeed, nothing is a universal standard, as each methodology addresses a target platform and they all have their pros and cons. I think developers working for a living (rather than doing it for academic research or bragging rights) will develop in whatever way is profitable for them in the end. If that means adopting Apple's methods with Xcode, then that's that -- you'd better follow Apple's lead. If that means going with Visual Studio, then so be it. Sure, there are some development environments that can target multiple platforms, but nothing's universal and nothing's without its tradeoffs. Don't like drinking the Xcode Kool Aid? Great. Just don't develop for the Mac then, or do it all yourself the hard way and start all over again when they change processor platforms.
  
  5. Nope, not been living in a jungle. But I also don't sit around building my own operating systems in my spare time or for my job. I have a job and a business to operate at work, and I have a life at home that doesn't involve using compilers or using the word "distro" a lot. And I'm the geek at home. My wife is the mass market user and keeps me grounded in reality, as do the hopelessly computer-phobic people at the office. To sell Linux in a big way, it'll have to be sold to those people one by one. So far, Windows dominates by force of marketing and inertia and Mac OS X has a few converts that have been burned by Windows in the past and don't want to be geeks. Linux? That would scare the crap out of the non-techie people I know. To you, they live in the jungle, unaware of the salvation Linux offers. To me, they're just normal folks with better things to do with their time. This is the fundamental understanding Steve Jobs maintains at Apple. If he dies, that understanding will die with him.
  
  6. As for the Intel transition, the mass market users Apple targets (one part of their market, but not all of it), the lack of Adobe apps is not an issue. Pro users are pissed (and I'm a little miffed myself), but the mass market isn't going to run out and buy (or stay home and not buy) based on Adobe apps performance or Universal Binary availability.
  
  In the end, I think Apple's future is very much bound up with Steve Jobs. Without him, Apple would make the same decisions that all the other tech companies make when it comes to products and strategies. Then Sony, IBM and all the others could just crush Apple with sheer market pressure. I totally agree with you that if they don't control headcount and other expenditures, and if they can't really break into the consumer electronics market (aside from the iPod), then the future will be not-so-bright. Just as I cannot fathom a Linux-dominant world, I also cannot fathom a Mac OS X-dominant world.
  
  Thanks for the comments back. I bet you make fascinating lunchtime conversation! And I appreciate the analysis. I'm not actually that "down" on your analysis -- I just don't feel that Apple is so easily compared to other companies in the market, nor are they as impacted by the trends out there. They also don't shift the market around as much as people say they do.
• This is a tech analysis about a company that isn't tech-focused
  2006-03-24 21:24:21  pquam [Reply | View]
  
  
  >Linux? That would scare the crap out of the >non-techie people I know. To you, they live in >the jungle, unaware of the salvation Linux >offers. To me, they're just normal folks with >better things to do with their time.
  
  --I think there is a certain amount of truth to this. Especially, since some people are intimidated by Windows. However, Mac OS X is a prime example of putting a user-friendly interface on top of Linux/BSD.
  
  With most Linux apps, you could make them into Windows-like apps by jazzing up the GUI, pre-configuring a bunch of things, disabling some advanced features, and pre-loading it into memory at startup and running them in the background, and maybe adding a few macros/scripts to accomplish common tasks.
  
  I agree that Linux has come a long way recently.
  On the other hand, people are less and less
  tolerant of the learning curve to use computers.
  However, it would be a mistake to underestimate Linux's ripeness for popular adoption, esp. if it was pre-packaged correctly.

• Brilliant analysis
  2006-03-23 21:48:50  pquam [Reply | View]
  
  
  This analysis is right on the money.
  
  The XBox is not a PC; the PS3 is.
  If you can get a supercomputing Linux workstation with various convergence properties (home entertainment center, video conferencing with the eye toy, home computer....) for so cheap, sales of
  PS3 and other cell-based products will set back
  a number of very prominent computer manufacturers
  and distributors, and reestablish Sony as
  the king of computing. Apple is not the only
  company in trouble. KDE can absolutely be made
  user-friendly so that it competes with Wintel.
  
• Brilliant analysis
  2006-03-24 18:10:15  AdrienLamothe [Reply | View]
  
  
  I don't think Sony was ever the king of computing, but they are an excellent consumer electronics company. Sony has a lot of potential.
  
  Some important KDE development is happening this year. KDE's technical foundation is pretty solid.
  
  Cheers!
  

• Cell Processors
  2006-03-23 18:02:11  malibu [Reply | View]
  
  
  While the cell processor marketing is good the reality is still in question. Apple was offered the Cell and turned it down after evaluating it.
  
  The big issue, one very similar to the Itanium processor are the compilers and tools. In order to maximize the performance on the Cell archtecture you have to program in a different model then most programs are written. More concerency and specialty options must be used to maximize throughput. IBM has admitted as much as they are working on a project called "Octopile"that will take programs written in standard programming languages and standard models to get them to perform optimally on the Cell architecture. I am sure this will take some time to mature.
  
  These are the same problems that exist with the Itanium. It currently does not support an on board instruction re-ordering unit and thus all branch predictions and optimization MUST be done at compile time to avoid pipeline issues.
  
  The G5 suffers from this in some ways as well. The fact that the intel architecture has had over 10 years of compiler development in terms of gcc alone and that optimizations and intrinsic library support is significantly more mature provides for huge gains in performance. The G5 also lacks a good set of support chips surrounding it. Much of the support ASICs in Apple's boxes were developed by Apple. With the intel architecture much of the supporting architecture is done and Apple can concentrate on other areas.
• Cell Processors
  2006-03-24 17:58:21  AdrienLamothe [Reply | View]
  
  
  Initial reports of Cell demos have been very favorable, but only time will tell. Early demos do tend to be tightly controlled affairs.
  
  I don't expect the different programming model to be a show-stopper.
  
  Excellent point about the compiler optimizations and chipsets. A company with the scope, size and organizational capacity of IBM should be able to overcome those types of problems.
  
  
• Cell Processors
  2006-03-25 05:50:24  malibu [Reply | View]
  
  
  A show stopper no but a significant obsticle yes.
  
  A brief description of the coding issues can be found here
  
  http://arstechnica.com/news.ars/post/20060225-6265.html
  
  I would not be so sure about IBM overcoming this issue quickly and confidently. They had same said resources and look where the G5 ended.
• Cell Processors
  2006-03-25 09:39:49  pquam [Reply | View]
  
  
  I agree that this won't be solved overnight to
  anyone's long-term satisfaction. Concurrency
  is a hot topic in CS, and has been since the 80's
  or before. So, in a sense, it has been the
  subject of a generation's doctoral theses, and still, I would argue, it's not mature, and few
  people would disagree. There have also been
  some commerical failures, as in the 5th generation
  computer project--which was not a failure in
  terms of teaching us about what not to do.
  
  I don't think this negates the usefulness of the cell processor though. In the worst case, ignoring the SPEs, you have a simple risc-based dual-threaded processor clocked up to about 4ghz, that's compatible with all the powerpc code. That means that there is a huge code-base that will work already with it. This may or may not be enough to make this a competative product.
  
  There will, of-course, be sub-optimal usage of
  the SPE's for the forseeable future with most
  applications. This is not necessarily a bad
  thing. What dedicated graphics has taught us
  is that you can take certain processor intensive
  tasks, and off-load them relatively cheaply
  to dedicated chips. Even selective usage
  of the SPE's could make a huge performance
  difference on certain processor-intensive tasks.
  
  What I see as the key feature of the cell
  processor for 2006 is not its parallel
  execution, but its I/O bandwidth. It's an
  incremental, but very significant advance on the
  real bottle-neck in modern processors: bandwidth.
  
  Bandwidth should be reason enough to make Cell
  successful until the tools for painless and
  automatic compiling can be developed.