“I have read many articles and debates on USB vs. Firewire. Recently, some writers have covered Apple’s stance on USB 2.0. I won’t bother to go into a technical debate as to which technology is better, or holds a better edge in future generations of the spec — let’s put that aside. I want to focus on whether Apple is following the best long-term strategy. In my opinion, it is not.” Read the editorial at MacSlash. On another editorial, on PCMag this time, Dvorak predicts: “Apple will switch to Intel processors within the next 12 to 18 months“. Dvorak advocates the Itanium as the main candidate.
Why Apple Needs To Support USB 2.0; Apple on Itanium?
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
apple will never adopt usb 2.0 (not in the short term at least, til usb 1.1 is out-phased for good), it a direct competition to firewire *especially* since they just released the faster 800 Mbit version
Well, Firewire and USB are competitors (as in “their market roads are merging”) but they are not competitors on all things. Most scanners/hdd/printers work with USB, while other kinds of devices mostly work with FireWire. In a way, they complement each other instead of directly competitng (while they still do). In other words, people NEED USB 2 as much OTHER people NEED Firewire.
But it does not matter if they compete or not, it would be a STRATEGIC mistake of Apple to not include USB 2.0 support on MacOSX 10.3, when USB 2 is so popular and many people have such devices. It is all about strategy, not about what we like or dislike.
Dvorak has a long history of writing things that make him sound like he’s on crack, and has on rare occasion admitted he chooses subjects based on their ability to provoke response rather than their technical merit.
Having said that, he makes a better case than most I’ve seen for this. He recognizes that a dual-processor solution would be the only way to really pull such a switch off, and recognizes further that Apple has too much of a stake in being an innovation leader to move to something truly commodity. On the other hand, he doesn’t address the possibility of a move to the PPC 970 chip at all–something analysts with a less “let’s say something to provoke response” attitude tend to place a much higher probability on.
Moving to the Itanium would be quite ridiculous considering that the PowerPC 970 is a few months from being ready. If we consider only the megahertz rating, Apple would be in a more pitty situation with the Itanium 2 1 ghz. Beside, wasn’t it Intel who said that the Itanium wasn’t ready for the desktop. I would imagine it’s much more expensive than both Motorola 7457 and the upcoming IBM 970.
Pixar processor choices are completely irevelant to Apple roadmap. I know Steve work for both companies, but they are completely free to use non Apple solutions. They used Sun render farms before and Apple never had anything to do with the Sparc.
>Dvorak has a long history of writing things that make him sound like he’s on crack
Haha… I actually like him. He writes whatever is on his mind really, without caring what other people think of him. I like that.
>He recognizes that a dual-processor solution would be the only way to really pull such a switch off
Xeons can do SMP, they have hyperthreading and they are faster than Itaniums, and are way *cheaper* than Itaniums. IMO, Xeons and the P4 architecture (compiling the OS directly for P4) would make more sense.
*However*, by choosing Itaniums, you get other strategic advantages: You don’t directly compete with MS, Intel is not pissed at you for competing with MS, Intel gets to push Itanium to the masses so they will like you, and most importantly, Apple would be the first to use the power of Itaniums for commodity machines in full 64bit. However, the Itaniums also don’t run in many Mhz (however the Itanium2s *are* fast), so Apple would face the same problem of the “Mhz don’t matter” slogan. Additionally, Itaniums don’t really run on laptops (or ‘don’t run as well’), which is the main market strength for Apple.
In other words, Itanium makes strategic sense, but it does not make market, price and product sense.
apple doesn’t exactly think in terms of “users need this”, usb-2 is very unlikely to make an appearance on macs because apple wants us to use firewire instead. I agree they are not used for all the same things, but apple is pushing it’s own not-so-proprietary-anymore technology. if you look 10 years back when macs had 3 i/o port types (adb, serial, scsi), it was organised in the way of “slow”, “mid-range” and “fast”; apple will keep usb 1.1 for the low-bandwidth devices and push its own firewire for the faster types. why would they provide 2 high-bandwidth implementations on one machine? instead, they’re just thinking “you really want usb 2? buy a pci expansion/pcmcia card”. and you need an expansion slot to put that card in it, which means you either need a g4 tower or a powerbook (convenient for them, isnt it?).
>apple doesn’t exactly think in terms of “users need this”,
Are you realizing what you are saying here? If Apple doesn’t ‘exactly think in these terms’, then they are dead on the water. Sorry. People buy what they need, not what Apple or Eugenia,Inc. dictates.
>but apple is pushing it’s own not-so-proprietary-anymore technology
As they did in the past with other types of technology. And they killed off all of them with time, in order to confront with the PC standards, just because that’s what the market reality dictates.
I believe that Apple will bring USB 2 soon. It only makes sense to do so.
From my observation, the people who seem to want USB 2 are switchers. So if Apple is trying to get switchers, they ought to listen up.
On the other hand, I’ve been a long time Mac user, plus all my PCs have Firewire, so I’ve never bothered to get anything other than Firewire devices. So my guess is that non-switchers are saying Apple doesn’t need it.
USB and USB 2 are designed to do one thing. They are designed as an interface for peripheral interconnect from some accessory to a PC. That’s it, that’s all the protocol covers, that’s what it’s designed to do.
Firewire and Firewire 800 are designed to do much more than that, and the protocol shows this quite clearly. It’s designed as an interconnect mechanism for any “smart” device. For connecting pieces of an audio rack together, for interfacing a SAN to some intelligent video equipment, etc. It’s designed to be more flexible and to go beyond the scope just connecting an accessory to a computer.
The endless winner take all comparisons are just silly. They don’t make any sense because even if USB becomes the only connection method for accessories to computers that still won’t affect the other places in other markets where Firewire is making many in-roads.
-Nathan
I find it unlikely that Apple will move to Intel, Dropping the PowerPC seems a very foolish thing to do with IBM pushing development so strongly now.
I don’t think apple will ever choose PC chips because of VMWare or MaconLinux type applications, it’s too dangerous. People with cheap PC’s will be able to run OSX at near native speeds, the demand for apple computers will fall. Apple cannot permit that.
Dvorak is a complete idiot. Apple move to Itanium??? You have got to be kidding! The Itanium is a complete and utter piece of crap. Intel has their head up their ass on this one. If Apple were going to move away from PowerPC, it would most likely be to AMD and their 64 bit architecture. But even so, IBM’s Power 970 is going to be awesome and not require Apple to do a whole lot in the way of rewriting software. There are rumor sites out there with stories already about a new motherboard designed to accomodate the 970. My prediction is that this chip is going to start a new revolution with the Mac. It will blow away anything Intel has to offer including it’s crapola Itanium. Yes this is probably going to start a flame war with the Windows turd burgulars but this chip will open doors for Apple like nothing else even with their paltry marketshare compared to Winblows. A new day is dawning for Apple!
Well, if MAcOSX on a P4 has stripped out completely the support for traditional 2D drivers and only support Quartz Extreme/3D/2D driver combinations, then it will be very difficult for VMWare to make OSX run on its products, even with a plain 2D Vmware driver… And remember our editorial here at osnews, where we said that Apple PCs would have a *custom* BIOS, so you won’t be able to run OSX on anything but Apple PCs. With this in place, VMWare will have no chance on running OSX. 😉
We see articles forseing the next big thing – Intel Macs.. every few months.
but imagine it really happened.. what would you like to see? Here’s mine
OpenFirmware BIOS
64-bit x86 processor
Ultra-SCSI HDD
RDRAM or equiv
OSX with PPC emulation for Classic apps. Possibly the option of a PPC daughtercard CPU for heavy classic users.
And just imagine.. run a Mac on x86 and you can use WINE for Windows application compatibility. No need for Microsoft to be on board when CodeWeavers Crossover allows the cheaper x86 Office to run and Apple/KDE Safari is default browser :-). No need to rely on Microsoft Connectix providing a port of VPC to the new architecture.
Yes I know I’m dreaming.
Apple created firewire for video, storage and peripheral devices (Scanners and commercial printers) that need dependable high bandwidth pipes, are hot-plugable, and I forget the other feature need (simple?).
Anyway, USB comes from a variable low-bandwidth pipe foundation, so it is ideal for mice/tablets & keyboards, consumer printers, and a few pictures from the digital camera.
I have experienced the wonders of USB and how it stutters audio output from an external CD player through the USB bus to the computer, not impressive. Especially, when there is a delay from the mouse, or the printer seems to stop to let the audio thru while I am trying to back-up data onto a CD-R (internal CD_R is faster than external because of USB bus limitations), this is frustrating. USB 2 ain’t gonna do for me what I need to do in a way that I don’t find annoying, frustrating, or stressful. The only benefit to using usb 2 over usb 1.1 is that pictures from the digital camera download almost, not quite, in half the time on a good day without me trying to print something out at the same time.
I wish I had time to tweak my system and spend hours trying to get things to work when I need them to, not when they manage to get around to finishing.
Therefore, USB 2 might be a good idea to adopt iff
1) digital camera and digital video camera makers stop putting Firewire ports on their devices
2)USB 1.1 is no longer available.
There are, sometimes, excellent reasons to upgrade to a new technology, but USB 2 only provides a little speed boost, at best for me. And Firewire is supperior to USB 2 for what I need it do (video storage and play back.
btw have you ever seen a filmscanner upload image data thru a firewire 800 pipe? It is fast.
Also, lets not forget SCSI and its varients. Still a superior option for Video RAID systems, USB doesn’t cut it, and won’t for some time. I try to use the tool that works best for what I am doing, whether or not its popular for moving digital photos to the computer almost twice as fast as USB 1.1.
“Xeons can do SMP, they have hyperthreading and they are faster than Itaniums, and are way *cheaper* than Itaniums. IMO, Xeons and the P4 architecture (compiling the OS directly for P4) would make more sense.”
The impression I got from reading the Dvorak article was that Apple could make a dual CPU system that uses one PPC chip and one Itanium chip. After that, modify the part of the kernel that loads executables so that it would first examine the program that’s about to run to determine what CPU it was compiled for and then launch that process and its threads on that particular CPU.
Itaniums aren’t exactly cheap last I heard though. We have people complaining about price as it is, these ‘theoritical’ Macs wouldn’t be exactly a bargain on the price arena, probably worse than they are now to some.
I agree with Eugenia totally on USB 2.0. I personally use firewire devices, but I DO think it should/will be supported for compatibility reasons. USB and Firewire aren’t really directly competing for the same markets even though they do overlap on a few things there. Firewire is targeting for media and high bandwidth tasks (Video, RAID, LAN) without the need for a computer. Where USB is more targeted towards peripherals (CD, DVD, Mouse, Keyboard). The two complement each other so I don’t see them as a threat against each other in such a critical way.
everyone else is providing value at commodity prices.
mac is sitting on that borderline right now, prices are high…high enough to drive away most of the mac “curious”.
but low enough to marginally keep the faithful.
a move to itanium? suicidal and stupid.
opteron makes more sense, but the new ibm chip makes the most sense.
Apple is actually DEFINITELY going to be including USB2.0 soon… What people need to realize is that Apple is going through a major architecture change RIGHT NOW–the significance of USB2 ranks low as a result, but the inclusion of USB2 will come with these changes. I do not know to what extent they are playing a game–trying to force FW adoption, but I do know that it’s a very sensible decision to push it off another 4-8 months if they are redesigning the entire Mobo for “most” of their products.
When you get to see a 970 PM, you will also see that it is equipped with USB2.0.
Dvorak is just completely useless. It’s one thing to say he’s provocative–it’s a whole ‘nother story to even suggest he’s credible, informed, sensible, has any understanding of the Mac community or Apple at all.
(Is this guy so out of the loop now that he doesn’t even know about the 970? Doesn’t he know that WWDC in May is a much better time to unleash his dual-processor prototype on the dev community? Does he know the web addresses of ANY mac sites (rumor or not) and does he check them or talk to anyone before writing this twaddle? etc…)
I bet they will do something crazy like put it on an Intel. I mean this always happens to me (I just bought an iBook). It’s everything I buy, Buy the Sega Dreamcast, only to have it drop from 99.99 to 49.99 the next week, then buy an Xbox for 299 only to have it drop to 199 the next week. Now of course I have to live with something like a new processor, and I waited out for 8 months before getting my iBook because I was so scared they would change. You can’t trust Apple, they do that sometimes, just change this or that.Then you have to buy this or that.
Apple is one of the only companies I can think of that can actually pull this off. And they have from time to time.
Just look at how far they are behind in the FSB speed games. Surely this would increase overall speed quite a bit, and it is really rather inexcusable as to why their ‘pro’ machines lacked DDR for so long, and even now without a decent fsb ddr doesn’t mean a whole lot.
Another case is when Apple decided to abandon the floppy completely with the iMac. I remember I was a freshman in college and LOTS of students had brand new iMacs in their dorm rooms. Guess what? Every single one of them had a usb floppy or zip drive hooked up. Apple’s decision was pretty forward looking, and they often make decisions like this based on what seems principal, or in an effort to shape the industry.
Why can they pull off being so techinically behind the hardware game and at times overlooking user needs? Because Apple users are just so damn loyal, and most aren’t really techies. Apple users by apple hardware for the OS.
(note: I own an ibook running debian gnu/linux exclusively)
I would be really surprised if Apple went with Itanium. Itanium 2 isn’t a bad processor (it is actually an excellent server/workstation chip). It just isn’t targeted for the desktop in its current incarnation (slow clock, monster cache, expensive!!) and I don’t think Intel could hide a desktop oriented version that would be ready this year. I suspect that Apple will adopt the 970. It makes an easy transition and IBM should be a fairly reliable supplier.
As for USB 2.0, I don’t really see a downside to Apple supporting it. It is just a backward compatible upgrade to USB 1.1. Most modern southbridges support it without the need for additional chips, so it is really cheap. Apple will probably support USB 2.0 when they get a new chipset (with the 970). Apple will still push Firewire (2) as the favored solution (which will be easy since even Firewire 1 provides better performance than USB 2.0), and I doubt many Apple products (such as iPod) will support USB 2.0.
My USB 2.0 experience has been fairly positive. My Creative Nomad Zen USB 2.0 has decent transfer times (~25% slower than what I’ve seen people get for the Firewire version), and I didn’t have to get a Firewire card (USB 2.0 is supported on my southbridge). Plus, the USB 2.0 Zen was cheaper than the Firewire/USB 1.1 Zen.
eugenia:
“I won’t bother to go into a technical debate as to which technology is better, or holds a better edge in future generations of the spec — let’s put that aside.”
Right, because firewire is infinitely better. And just flat out acknowledging something is better is against your religion.
Why are you quoting my name over there? It wasn’t me who wrote that! Why every time people have to shoot the postman?
“””The Itanium is a complete and utter piece of crap. Intel has their head up their ass on this one. “””
What is the problem that everyone has with the Itanium? It’s a well designed ISA/chip, powerful, fast, and runs existing x86 code. Cost is most likely a non-issue since it’s an extremely low run chip (that’s like complaining that a power4 or an ultrasparc isn’t cheap in it’s current form).
Additionally, Intel is giving it a real chance by developing the chip at a reasonable rate allowing the necessary software infrastructure to develop instead of slamming it out the door only to find it ill-equipped to meet the market’s needs.
Let’s take a look at his logic:
“The choice of the Itanium is suggested by four factors.”
1. “First, there is zero evidence that Apple is talking to AMD—and it would if it were staying with the x86 legacy chips.” Actually there is evidence. Is there any evidence Apple is talking to Intel? Far less than the evidence that they have talked to AMD. Purchasing hardware for Pixar doesn’t count as Apple talking to Intel about a strategic anything. And does not talking to AMD really point to using Intel–couldn’t not talking just indicate they aren’t switching?
2. “Second, Apple likes to make jazzy announcements in which it claims to be the first or the most aggressive in a market.” Okay, this is just a pathetic, stupid statement. But how does it indicate Intel at all? Nevermind Itanium. Wouldn’t picking Alpha or SPARC be even “jazzier” for the insanity of it all?
3. “Third, if Apple optimizes the OS X kernel for the Itanium, the likelihood of the Apple OS being ripped off by normal PC users is nil.” And this is also true of any processor in the world.
4. “And finally, by choosing the Itanium, Apple will have an ally in Intel, who will put its design team to work for Apple and perhaps even invest in the company, knowing AMD is not in the picture.” Apple would be willing to invest in any chip partner it had–I don’t even get what the hell he is talking about when he says, “knowing AMD is not in the picture.” Aren’t Moto (even if they are useless) and IBM partners NOW? Does Intel ally with ANYONE besides MS? Isn’t it AMD that has been doing substantial work to create strategic alliances (the GPU makers, IBM, Toshiba, etc..)?
Dumb@ss, dumb@ss, dumb@ss.
Itanium2 is a great CPU (as long your program doesn’t run on x86 compatibility mode). End of story.
Right, because firewire is infinitely better. And just flat out acknowledging something is better is against your religion.
Do you not get it? The article gave reasons why Apple should support USB 2.0. You come back with a personal attack (taken out of context) aimed at the wrong person? Did you even read the article before mashiing your keyboard?
Let’s see:
it can’t run workstation software at close to the speeds of equivalent x86 PCs.
it is expensive–low volumes are a factor, but like the Power4, the cost does come from cache and in the case of the Itanium the RAM as well.
it’s x86 emulation is poor.
no one is in a rush to create 64 bit Itanium only code.
Do we need more reasons?
“Apple has a unique ability to get away with changing processors radically. It has used the 6502, then the 68K, and now the PowerPC. Each transition happened almost flawlessly.”
If my memory serves me, the 6502 processors were used in Apple 2 machines like my old 2gs which used a 65816. The 68000s were used in early Macs. There was no transition at all between the Apple 2 series and the Macintosh other than a marketing one. No software compatibility. No upgrades. About the only thing which could be seen as part of a “transition strategy” was the use of ADB ports on the 2gs. Apple alienated a large part of its existing customer base by dropping the Apple 2. What Apple got away with in 1990 was quite different from what Apple got away with during the introduction of the PPC.
Apple would be willing to invest in any chip partner it had–I don’t even get what the hell he is talking about when he says, “knowing AMD is not in the picture.”
I think you misunderstand. He is saying that Intel would invest in Apple because on Apple machines they would have no competition from AMD.
As far as USB 2.0 goes, most people worry about the issue that 2.0 is now at 480 Kbps, while Firewire is only 400 Kbps. You do need to understand a little bit of the technical side here. USB, both iterations, relies on the CPU to function. Therefore, if you have other stuff happening (moving the mouse, running other programs, etc.) the USB bus slows down. The keyword is that the 480 Kbps is “theoretical.” Now, Firewire, IEEE1394, is processor INdependent, which practically speaking gives it a higher actual throughput than the USB by quite a wide margin. When you are talking about digital video and/or audio, Firewire 400 is in no danger at all from USB. So I would imagine that Apple has not included USB2.0 for purely economic reasons.
As far as Dvorak’s comments, ignore him. He has a long history of throwing out outrageous articles about Apple just to get his hit rate up. Apple would not do a another major architecture change after only 2 years. They are just now getting all the software vendors switched over to OSX. All you people that have been patiently waiting for an OSX version of Quark, can you imagine the wait for an OSX on Intel version?
Apple should do BOTH. I agree keeping the desktop running PPC 970. However,it would make sense to support several hardware platforms for XServe (Intel,AMD,SPARC,Alpha et cetra.). If you are running a web site, clustering, or a database, you want the best cost permoformance ratio you can find. Pixar could have picked Xserve if they were had Intel or AMD running inside! You only have to re-compile/port the server apps, not photoshop or Quark. Apple has the source code to those things. Microsoft was developing a PPC version of Windows when the chip first came out. Be like microsoft, ready to *switch* at any moment.
********* Think diferent! *************
What the article is saying is that many low-cost consumer peripherals are available in USB versions only. If Apple doesn’t adopt USB 2.0, that limits the use of a potentially huge number of devices for use on the Mac. Computer manufacturers are unlikely to swith to FireWire for these consumer-orientated devices because FireWire is not a standard component on most PC motherboards (and the PC market is likely to be their primary target because of its size). It’s not about whether Firewire is superior to USB. You’d think Mac users would welcome more choice, but judging from some of the reactions here, maybe not!
http://www.flickerdown.com/phpBB2/viewtopic.php?t=2743
So what? Intel has long realised that as a marketing company, Apple is bar-none the best one out there. If Intel exec were in the audience, it was more to learn from Steve “Sno-blowing” Jobs than anything else. Anyone have a count of how many shares Intel has in Apple? In regards to Pixar, it was a logical step. Use of their entirely antiquated systems from sgi, et al. forced the move more than anything else. Ever heard of TCO? By moving to a commodity processing platform Pixar is able to cut operating losses in off-seasons and dynamically increase their performance…<BR>
Change is good. Apple has a unique ability to get away with changing processors radically. It has used the 6502, then the 68K, and now the PowerPC. Each transition happened almost flawlessly. On the PC side of the fence, no Z-80 maker survived even the transition to the 8080. Apple has also cultivated a fanatical following, who have long since accepted the fact that Apple eschews long-term backward compatibility. The legacy concept does not hold the power over Apple users that it does in the PC universe.<BR>
That’s more a testament to Jobs’, et al. stubborness in the face of ever-shrinking margins. A “price cut” to Apple amounts to a maximum of 5% off the top of their systems. Compare this to the whopping cuts you see from AMD and Intel and you get the picture.<BR>
Apple’s only concern is cannibalization. It cannot change architectures with a pipeline full of PowerPC products. So expect a slow transition that will start with the high-end workstations. Apple’s concern is that Motorola may muddy the situation, so Jobs will have to convince Motorola and customers that the PowerPC will not be phased out but will remain as part of a dual-processor architecture.<BR>
How the heck are they going to do that? You’re already building a pay-as-you-go OS on top of the FreeBSD/Unix operating system. Gee, how many x86 ports are out there? Motorola has already moved on. There is more money to be made in the consumer electronic market than there is in Apple’s kludgedom. And, I would shy away from calling anything that Apple makes “high-end”.<BR>
Scenario. Apple will announce its Intel initiative by showing a transition machine that uses both the Intel and Motorola processors. “So current Mac owners will not have to worry.” This will be a high-end machine optimized to run Photoshop. Apple is adept at creating dual-processor architectures, so this won’t be too radical. We’ve heard rumors of this kind of scenario for some time, under the code name Marklar.<BR>
Prove it. How are you going to saddle two distinct hardware architectures, buses, power requirements, etc. into one “proof of concept” machine? Tell you what, do a price search on an Itanium processor (either the first or second revisions will do). Tell me the cost. Add approximately 50% of that processor cost BACK to the system in addition to the original processor cost and you should have a rough estimate of what that machine will cost just for the box (remember: no monitor is included). Contrast that system cost (running in the mid $5K range by now) with a dual or quad processor based system from either Intel or AMD. See the difference in the “bang for your buck” level? You’re a moron if you want to buy that system and even a bigger one for suggesting it…<BR>
Itanium. What will be radical is the company’s choice of processor. Apple will announce its use of the Itanium chip, which can be used in such a multiprocessor design and will become the first desktop use of the chip. The choice of the Itanium is suggested by four factors. First, there is zero evidence that Apple is talking to AMD—and it would if it were staying with the x86 legacy chips. Second, Apple likes to make jazzy announcements in which it claims to be the first or the most aggressive in a market. The Itanium fills the bill perfectly, because Jobs can lord it over current PC makers with all sorts of performance claims.<BR>
ROFL! You’re kidding, right? The “average” user has already complained that AMD machines put off too much heat, are incompatible with x86 software, etc. And you want to strap Apple to a $2K processor designed for datacenters? You’re out of your bloody mind!!! Here’s why:
1.) Apple has 2% of the world market for PCs.
2.) Intel has about 15% of the world market for enterprise class (read: competitors to Alpha, Power4, SPARC, etc.) and is increasing every day.
3.) The typical profit margin for consumer level PC’s is ~5%.
4.) Apple would have to price their systems much more competitively than they are now (read: that’s a hilarious joke) and cut THEIR already stupidly high profit margin even lower.
5.) Less profit margin = less inflow/outflow. Less outflow = less ability to compete. Less ability to compete = your ass is on the line with the shareholders…Get my drift?<BR>
Third, if Apple optimizes the OS X kernel for the Itanium, the likelihood of the Apple OS being ripped off by normal PC users is nil. And finally, by choosing the Itanium, Apple will have an ally in Intel, who will put its design team to work for Apple and perhaps even invest in the company, knowing AMD is not in the picture.<BR>
Not to bag on EPIC at all (well…..ok, i won’t) but this statement is retarded. Again, what software drives the Apple market? Graphics (P-shop, etc.), DAW (digital audio workstation) software, and Video (Final Cut pro, etc.) Why would all of these companies RE-CODE their software to run on an enterprise-class instruction set/processor? There would be NO advantage.<BR>
The Apple switch cannot be just a short-term fix for the megahertz dilemma. Jobs is part of the anti-Microsoft Silicon Valley clique, and despite the fact that Microsoft helped Apple financially, the favor was designed to benefit Microsoft more than Apple. Jobs is a peer of Bill Gates. He sees the numbers Microsoft has racked up. Apple has enough confidence in its hardware designs that it can again risk licensing the Mac OS to the Intel platform. The perfect ploy would be to make an Itanium-only Mac OS with some sort of backward compatibility with Microsoft code. The fact that Apple recently released Keynote as a standalone software product says the company is ready to go after the Microsoft cash cows: Office and Windows.<BR>
Actually, it just proves they’re dumber than they look. Don’t mistake confidence for stubborn stupidity (well, you’re already making that mistake, but…that’s another story). Apple is at the mercy of their suppliers. It’s not enough to change your case design every two years. You have to work on the internals. Faster processors, more memory, SuperDrives, etc. will only take you so far. Their software is NOT superior, they simply know how to market it better. The ironic thing is, Mac people are about as dumb as sheep in this regard (and the recent appointment of Gore to the board doesn’t heighten my opinion of them AT ALL). For example: Mac OSX 10.2 was nothing more than a “point-release” (equivalent to XP SP1 or so) and yet, Apple charged (as of the time that I write this) $120.00 or so for it. And people bought it???? geez…<BR>
Timing is everything. Announcing the new architecture in July at the next Macworld Expo would be ideal, since it takes place in the media center, New York City. Whether Apple can actually have a working unit by July is questionable, but Jobs has been known to drive his people hard. Waiting until 2004 is too risky, but that might be the reason Apple is upset about the 2004 Macworld show being moved back to Boston. And consider the fall 2003 possibility: Comdex. Now that would get some attention.<BR>
Hey Eugenia,
I have one question for you tho, I like the idea about custom BIOS, but how long do you think it will take some 14 yr old hacker to figure out how to bypass it?
Apple should take the plunge and move to x86 hardware totally, it can survive off of support and service and multimedia, Plus I feel that it Apple was to make the “switch” it will kill MS Windows, Geeks galore will love it because its desktop BSD and the home users will love it because of the application base, the stability and the ease-of-use are legendary on the Mac, yes PC has surpassed it, but look at the price. If Apple keeps the current price scheme ie. $129.00 for a copy of OS X on intel, I think people would buy it. I dont personally think another x86 hardware company is what we need, Hardware is pretty much a mute point IMHO, I think Software will start to drive this industry. What do you think?
hey, sorry, I did think eugenia said that.
in any event, firewire is better. we all know that. why bother with a debate on whether apple should pick inferior tech. apple doesn’t. that’s the point of apple. like it or leave it.
probably apple will eventually use usb 2. they just won’t be in a real hurry. I would say that’s just about got it right.
I think the basic point here is that if Apple keeps USB 1.1 rather than upgrading to USB 2, they are using inferior technology. Apple has pushed the industry along before in moving to new technologies, yes–but not by keeping an old one in place simultaneously.
But it would be the most popular workstation released in years.
IA64 has suffered from lack of a real desktop os.
RH Adv. Wrk & hpux are not good answers and Windows support is limited.
It will have the fastest stream numbers besides marvel. (mem bandwidth)
It will have more FPU muscle them anything else and integer won’t be horrible.
It has real system io. Dell is going to sell motherboards and chips so the price may come down. Deerfeild will be out soon but I am not sure it would work in a laptop.
Check out a hp rx2600 or the single processor version. They are pricy today but have real horse power. mulitple GBps of io.
I would be a real rendering and modeling power house.
It would give their attemp at high performance computing some clout.
There will be a lot more madisons shipped the the first I2.
the problem with the ppc970 is it will still be slow when it is delivered compared to p4 and amd.
I think Apple will eventually adopt USB2 but they will be in no hurry. Whats the point anyway?
Given that you can buy a scanner with FireWire and USB2 for your Mac why use USB2? Its not like its faster. Is there a compelling technology out there that given the choice of USB2 or FW400 would someone choose USB2? Apple is all about proliferating FW400, why support USB2? Also, when you have just dropped $1499+ on a new G4 its not like you can’t afford a $9 USB2 card.
Apple moving to Itanium? Its a ridiculous notion, either that or IBM is just kidding about introducing a processor that is exactly what Apple is looking for, even going so far as to using the Altivec name which is probably licensed from Moto.
Apple is interested in building a platform that they control totally. They are not interested in the possibility of any aspect of their software or hardware being reversed engineered. MS already did that and got what they needed, they don’t need 20 mobo manufacturers in Taiwan on China to do the same for hardware. Whats the point in producing hardware if MacOSX runs on a WalMart PC. Keep your WalMart PC and I’ll stick with Apple’s designs any day.
A custom bios is interesting but Mac and PC users know how to extract a BIOS and flash/hack it on another system. Mac users do it all the time with video cards not intended for the Mac platform. I don’t see this as a compelling deterrant.
I disagree with Eugenia on two points:
(a) Itanium is good — they haven’t even figured out how to right software for it to get it run fast yet (sure the FP tests look cool, but the integer scores are pathetic). This stems from the fact that the architecture is VLIW, and compilers just aren’t ready yet. It runs too hot, has a large die, and is too expensive for what you get. It can be useful for some stratospheric pie in the sky researchers, but that isn’t what made intel, and thats the big mistake of Itanium for Intel.
(b) USB 2 — I don’t see the big advantage to USB2. USB2 still suffers from the same issues that USB1 suffers from — dependency on the host, and bus arbitration nightmares. Sure, there are probably 6 or 7 USB2 products out there on the market now, but firewire is steadily gaining ground. For Apple to offer USB2, they would give up that hard won ground because they would be supporting the inferior protocol. Their goal is to have firewire to be as cheap as USB2 stuff, and the only way to do that is to make hw manufacturers who want mac business make firewire products or live with abysmally slow USB1.1.
Apple should get their OS running on x86 hardware and port wine for most of the apps they would need. If they market it right there could be a clear separation between the systems and OSs while they work identically and seamlessly. The x86 side would have some windows compatibility for anyone that wanted to write software for both platforms and the OSX side could be kept going strong with the PPC970 for pure OSX horsepower, etc. Though I have no idea if this could be done with wine and the darwin architecture and licenses.
The mac addicts might want to consult http://www.macbidouille.com (sorry it is in french) So far last year they have not made one false prediction or offered uncorrect information. Looks like the september PPC 970 mobos will include USB2.
Apple just has not been in a hurry to include USB2 until their next major mobo rev, and why should they ? Nearly everything so far is either backward compatible USB1 or offers firewire.
Even if Apple offers OSX for intel platform (might make sense, although driver support would be extremely limited) I do not see them abandonning the PPC anytime soon, after IBM stepping in the arena with competitive offerings, convergent interests and consistent roadmaps. That would be the death of the platform. I do not think the mac developers would accept once more to change their apps for a non existant base market share (OSX/INTEL).
“Itanium2 is a great CPU (as long your program doesn’t run on x86 compatibility mode). End of story.”
Is it much more energy efficent than the pentium? If not, than it is a poor cpu in my opinion. Intel and others need to focus on low power, efficient solutions – not power guzzlers.. Not many consumers need much more speed as it is…
No one seems to note the fact that in the PC world, FireWire and USB 2.x are established standards. They are complementary, and in fact, while Apple boasts Firewire development, the pc world is way down the path ahead of Apple be implementing TCP/IP over Firewire, something that is still only in the wings at Apple. That says to me that in the Pc world, Firewire is there to stay. They also can take advantage of add-ons that benefit from USB 2.0 – scads of scanners, printers, etc. Complementary, not either or. So why should they enjoy all the advantages and fun? I hate being limited to the paucity of add-ons available to me as a Mac user. Why can’t Apple avail itself of the industry-standard, just as it learned to embrace Bluetooth. Remember a little over a year ago, the anxiety whether 802.11 or Bluetooth was going to be the standard? How nervous the Mac community was? But Apple caught on to the complementarity there, and it should do so again.
Surely Apple will adopt USB 2. They have to in order to compete. They may be waiting until they, hopefully, start using the new IBM chips, get a faster bus…in other words, until they change the whole architecture of the Mac later this year.
According to Think Secret, Al Gore has been named to Apple’s Board Of Directors! I guess he’d better ditch that Blackberry 🙂
Will Apple adopt USB2? Yes, eventually…
Will Apple switch from PPC? Absolutely NOT. Apple will obviously switch to PPC970, which won’t harm backward compatibility with existing software. With the IBM PPC line looking to outperform x86 some time in the near future, of course this is the route they will take…
“.. The ironic thing is, Mac people are about as dumb as sheep in this regard (and the recent appointment of Gore to the board doesn’t heighten my opinion of them AT ALL). For example: Mac OSX 10.2 was nothing more than a “point-release” (equivalent to XP SP1 or so) and yet, Apple charged (as of the time that I write this) $120.00 or so for it. And people bought it???? geez…”
Windows 2k version number is 5.0 and Windows XP is 5.1. Looks like a point release there. Also I believe Windows 95-ME were all in the 4.x range. Wow.. Looks like all point releases! People who are tricked by version number increases are the ones ‘dumb as sheep’. E.G. Linux distros seem to go from 8.x to 8.2x then to 9.x, yet there are minor changes. Why do they do this? Marketing.
Most of the time, I have discovered Mac users aren’t dumb or stupid. They just buy what they like and what they are used to. Or they could be a *nix geek and want a new toy to play with.anyway… </rant>
If Apple moved onto Intel hardware would Microsoft move Windows XP onto PPC? Their development tools certainly support it, and the NT kernel is multi-platform. Would we get a UNIX (Linux, MacOS X,…) vs. Windows world? The bond between operating system and hardware could finally be being broken. This makes for interesting times as a spectator. Hardware becomes a pure commodity that users can then go out and install which ever OS and application flavours suit them. The OS goes from something that comes with your PC to something that enables your favourite software with the look and feel you most enjoy.
After much consideration, I have decied you are correct. Apple should recall all of their computers and downgrade to USB 1.0 so that no one is using an inferior technology. Or perhaps just remove USB support (hope your three button mouse isn’t USB) since that is supporting an inferior and peripherally competing technology (nih).
Masao, I half agree with you. 10.2 was much much more than a typical Windows Service Pack – there was so much introduced in it – but charging $129 for it was not fun. And, Apple says they will continue to do this. Their cycle is 12-18 months and, when 10.3 comes out, we can expect to pay through the nose again (and again 🙂
Apple pulled off the 68K to PPC crossover brilliantly with their embedded 68K emulator. The fact that they were able to get this system tight enough into the core of the OS to where they were running system level code (specifically things like the file system) written in 68K on a PPC machine is simply amazing. It was slow, but not THAT slow, and certainly not unusable.
It helped that the PPC was a much faster chip than the 68Ks, and also pretty much was a superset of the 68K architecture, and therefore had the bandwdith and capacity to perform the emulation and run “almost as fast” as the original CPUs it was replacing, and then run its own code that much faster.
This made a PPC Mac competetive in both speed and price with a 68K Mac and also gave the machine a far brighter future than awaited it should it have stayed with the 68K.
Apple is pulling off OS X because they’ve shrouded the Mac OS ROMS and OS into its own little pseudo Jail that runs within the architecture of the Unix system underlying OS X. The code runs natively, with all of the ROM traps faulting back into the Unix kernel as necessary. This works pretty well. This works so well that no doubt the 68K emulator is STILL THERE within the Mac OS running on top of Unix. Emulation upon emulation.
It is difficult enough to get 2 CPUs in a machine to cooperate together. Getting two distinctly different CPUs to cooperate is a nightmare. It’s one thing if it was simply a coprocessor card, with its own RAM etc that hooked in to the bus and played nice with the other hardware. But then, you have essentially a loose coupling of the systems. You’ll have Mac OS X It running on the host with Mac OS X PPC running on the card.
What a nightmare. What a kludge. What a vastly expensive way to pull it off.
I do not believe that the Itanium has the capacity and bandwidth to emulate the PPC effectively, but I may be wrong. The x86 certainly does not.
This is the biggest challenge as to why Apple can not simply jump CPUs at this time. Not Yet.
People think they’re going to be able to do another 68K -> PPC leap like they did before, and I don’t think this is true.
I think that if they were to change CPUs, they would go whole hog, damn the torpedoes, and switch. They would do this by following the NeXT strategy of making cross development essentially painless and having the Fat Binary format available.
Seed the cross compilers and machines early. Most high level apps should Just Work with a simple compile. Shareware authors should be able to easily compile their applications for other CPUs even if they don’t have access to them, so Shareware Should Work with minimal testing.
Java will work, “of course”.
OS X enables this because it is more portable than Mac OS was. Developers getting their apps to work on OS X makes them (ideally) more portable than they were before, simply because the more esoteric less than portable code had to be changed to make the leap to OS X.
I think that the Cluster XServer would be the perfect example of an early entry into this market. This gets an alternative architecture into the market in a limited space and makes it available to users and developers.
So, if they want to come out with an Itanium anything, it would be the back office servers first, then migrate the rest of the line.
How long has Dvorak been saying apple must go intel. No they don’t have to but apple and IBM really need to find some volume applications for non-embedded power pcs. I think G4s running linux for the enterprise is the killer app for linux and IBM power PCs and potentially the savior for apple. Apple’s strategy is very good. They are expanding their potential markets and seeking out lucrative high margin segments (ipod and xserve). Dvorak just can’t help writing like all the other drones in the PCs world. Must assimilate…must follow..must say the same thing everyone else does. Jobs has showed these guys up again and again and its not going to stop. Personally, i eagerly await the iphone or a purchase of Palm (palm needs the help).
Apple does not have to “keep up” with intel’s GHz obsession. They need to build good “systems” and optimize performance in other fashions. this use intel stuff is nonsense, the same nonsense that has gone on for years.
“Masao, I half agree with you. 10.2 was much much more than a typical Windows Service Pack – there was so much introduced in it – but charging $129 for it was not fun. And, Apple says they will continue to do this. Their cycle is 12-18 months and, when 10.3 comes out, we can expect to pay through the nose again (and again :-)”
Well, from what I have heard Apple isn’t make much profit off the OS and it is close to at cost. That is why they didn’t sell an ‘upgrade version’. Now, I do agree with you. $129 for an upgrade is too much, but since Apple is close to at cost on this, then it is understandable.
Now personally, if there is no real reason to upgrade then why upgrade? Personally, I only upgrade software for speed improvements, they have to be large ones, or I don’t bother upgrading at all. FYI, I don’t own a Mac, yet. 🙂
-Masao
Think about this for a second. What if Apple were to drop USB completely from its product lines. Make its keyboards/mice either bluetooth wireless or firewire. You would say this would piss people off and make switching more difficult, and you would be right. However what this would create is a boon in Firewire, just as the original iMac created a boon in USB. Soon, tons of devices would be appearing for new macs that are firewire only, and the market would grow exponentially. The real question is, is it worthwhile to do? Probably not, although it would advance computing and help Apple long-term, it would probably hurt Apple short term too much.
YAY! JOHN DVORAK SAID THAT APPLE SHOULD SWITCH TO ITANIUM! YAY IT MUST BE TRUE!
C’mon people, Dvorak also said that the best way for linux to take off as a desktop os was for Adobe to port Photoshop to it.
Yes, photoshop. The only thing missing from linux according to Dvorak is freakin’ photoshop.
More kernels of wisdom from Dvorak:
“Web service should be metered and users should pay for the amount of service they use.”
“Why can’t Apple come up with a fresh idea? The company continues to deliver form over substance.”
“Free OS/2! There’s no good reason not to make this code available to the programming public.”
I’ll say one thing… at least he’s not Robert Cringely.