Is It Time to Wait for Intel?
by Chris Adamson
So, there are new PowerBooks and new dual-core PowerMacs. Nice.
But I don't think I'd buy one right now.
OK, sure, the politically correct thing to say is that "if you need a computer now, go get one", "all products are vapor until they ship", etc. But given the Intel transition, is that really a good idea?
The key is probably: how soon will it be until new stuff that you want won't run on your PowerPC Mac? Here's how I'm teasing out this logic: at WWDC 2005, we were promised a look at Leopard (Mac OS X 10.5) at WWDC 2006, which is next Spring or Summer. There's usually a lag between unveiling an OS to developers and shipping it, so the developers can learn and use the new features. Add 6-12 months to the Leopard preview and we probably don't expect Leopard until, what, late 2006 or (more likely) sometime in 2007. Intel Macs are supposed to ship in mid 2006, so clearly some future version of Tiger will support them, as will Leopard.
Apple said with the release of Tiger that the timeframe between major OS releases would slow down, so when do we expect to see Leopard's successor? If Tiger shipped in 2004 and Leopard is in 2007, then maybe Mac OS X 10.6 ships around 2010 or so. That's four years after the Intel switch.
And therein is the big question: will the PowerPC's be sufficiently old by that point to justify not supporting them in OS X 10.6? There's a big cost to having dual-platform code: everything has to be compiled for and tested on two different architectures. At an Apple BoF at JavaOne 2005, the Apple guys said they were a little concerned about how they'd now need to support six Java runtimes: Java 1.3.1, 1.4.2 and 5.0 for PowerPC, and 1.4.2, and two different HotSpot VM's for 5.0 on Intel. Oh, and Mustang (Java 6.0) ships in Summer 2006, so add two more, one each for PowerPC and Intel. Multiply this line of thinking across all the major libraries and frameworks -- Quartz, OpenGL, QuickTime, Core *, Cocoa, Carbon, AppKit, etc. -- and supporting those four-year-old boxes may start to look like something of a luxury.
So, I suspect that Leopard is the end of the line for PowerPC, and that 10.6 will be Intel-only. That means you are buying into a four-year dead-end on PowerPC. Is that OK? A lot of professionals and business types figure that computers have a three- to four-year expectancy of usefulness, so maybe that's OK. In the home realm, I think people are more careful with their money and expect their computers to last longer (even if they do upgrade this frequently, I'm not sure people really realize that they do).
My brother asked me recently if he should upgrade his ancient iBook now. Knowing the terrible performance of the Apple laptop line -- notice that the new PowerBooks didn't get any faster in the latest rev -- I said he'd be much better off toughing it out until the laptops go Intel. The crazy fast desktops may be a different story, but I still think you have to admit to yourself that you're buying four years of Mac, and amortize appropriately.
This is where you flame me for telling people that they shouldn't buy Macs now.
Your ridiculous advice.
I cannot believe you actually seem to believe your advice. Have you fooled yourself? ... Rhetorical question. ... Of course the ansewer is, "Yes, I have."
When the intel machine is released, will your advice be to wait for the second generation, and when the second gen intel machine is released will your advice be to wait for the next fastest machine?
What is the lifespan of your current machines? I use my machines between 3, and 5 years, because I have to pay for them myself. Any machine on the market is good for 3 to 5 years.
My advice is, if you need a machine, do your research, and get a machine.
I think your assumption is wrong
You seem to be assuming that within 4 years, the number of Intel Macs will outnumber the installed base of PowerPC Macs to such an extent that it would make sense for Apple to stop development of the powerpc version, despite the fact that the hard part is pretty much all done. I think that's a pretty big assumption.
Even if Apple sells so many Intel Macs in the next 4 years that they equal the installed base of PowerPC macs, why would Apple give up half the potential sales of 10.6, when most of the code that has to be written isn't processor specific and will work fine on both? Let alone all the loss of good will.
Of course, you also fail to take into account that the transition won't START until 2006, and they will probably be selling at least 1 or 2 PowerPC Macs until 2007.
Basically, I think you're flat wrong, and Apple's sales for the last quarter would indicate that a lot of people agree with me. :)
I agree about waiting on laptops, and unless something urgent is due in the next quarter, waiting on desktops as well.
The current release had to fulfill three goals, driven mostly by the marketplace:
- Prove the PPC is still worth spending money on.
- Don't exceed the coming Intel boxes. When Intel hits, it has to be fantastic.
- Don't look pathetic compared to the coming Intel boxes. When Intel hits, these machines can't be so embarrassing they make recent customers feel ripped off.
I think the towers largely did that. The G5 quad ships with a ton of horsepower, it almost seems silly how much they crammed into the case. If these are setting the bar for Intel's introduction, then I'm even more excited than I was after WWDC.
The PowerBooks however are fairly disappointing. While I'm glad to see better screens, GPU and battery life; the chip will have been shipping for a year come January. The G4 seems to have been maxed out for portables. Intel's roadmap shows Pentium M chips at 2.26ghz for the first half of 2006. As a PowerBook user who's due for an upgrade, that big a speed jump is worth waiting a few more months. Intel's portable chips also claim significantly improved energy efficiency, so these batteries might also be a middle step to even more impressive battery life.
Looking longer, Intel machines have to ship a few months before WWDC 2006. Developers need time to get situated with their new machines before the conference, which will undoubtably be very Intel focused.
Well here's hoping anyway.
Something of nothing
I am at a loss to the logic in the blog.
On a pure hardware development level, I'm excited by the Intel change. That's technology lust. But that's not linked to my immediate processing needs.
Is having a faster processor going make my iTunes music play faster? Will it make my word processor typing faster? Answer, No it will not. I don't need it to.
I think as an average user, the even present portable machines are fine, some writing, spreadsheets, some music, some film, bit of CAD- I might be missing something, but what are you guys using yours for?
I think if you are hanging waiting for any machine that will not date - you will never buy one.
Just another voice echoing what's already been said.
I'm in the market for a laptop that I'll have to pay for myself (I'm self-employed). I currently use a mini for a desktop but I need something portable as well. My usage is mostly word processing and email, with occassional graphics. A speed boost isn't going to be a huge deal for me. And the move to Intel isn't going to stop me from picking up one of the new Powerbooks.
A couple of things occured to me reading your article. There's no guarantee that an Intel laptop will be released first, so how long do you expect someone to wait. And would you really suggest that someone buty the first rev of the Intel machines? You also assume that everyone immediately jumps on a new release of an OS, and that's not the case, especially for home users.
I think you should reevaluate what you've written. You're offering dumb advice.
From a strictly hardware standpoint, it might be worth it to tough it out. Maybe. Depending on whether the first Intels out the gate are within your brother's price range, and they work well with all the recompiled software etc.
From a software standpoint, the situation is murkier. How much software will be available in Intel format on the day your brother decides to buy his new machine? If he needs to pay for upgrades to get Intel-native version, will that put the overall cost of the package out of reach? I know that a lot of developers have been quick to port to Intel, which is great. But we won't know until Macintels start appearing whether everything we need is available in that format.
And if it isn't, then much of the hardware advantage will be negated by emulation.
I bought a new Mac about a week after the Big Announcement. I figure this will last me at least until the Macintels have been through a round of revisions to shake the kinks out.
Unsupported is not Obsolete
I agree with most of the other comments so far, but I think one additional point hasn't been made yet:
Just because new software stops supporting a particular piece of hardware, that doesn't mean that the existing software suddenly stops working. Maybe you won't be able to use iTunes 13 (the version that plays dreams directly into your brain), but iTunes 12.8 will still work. If you can't use 10.7, 10.6 will still serve you pretty well.
How many creative folks are still on OS 9? It's unsupported, new software generally dried up years ago, yet the machines and their users keep humming along.
Unless your drive for the latest software constantly requires hardware upgrades (in which case you aren't buying for the long haul anyway), there's no good reason to fear obsolescense with a current Mac for at least five years.
What about your software investment?
For many people, software is the greatest cost of computing. If you use apps like Create or iMaginator for Stone Design the Intell switch won't be a problem - there will be no cost to upgrade.
If you use the current version of Photoshop though, that takes advantage of Altivec on the PPC, you may have some trouble. It's going to need to run in Rosetta - which doesn't handle the Altivec calls. It's probably going to mean purcha$ing new upgrades form Adobe.
I would rather take my chances with a mature kick a$$ hardware configuration and the software I currently have, rather than Intel v.1 whenever it actually appears.
Besides, Apple is going to switch to nitro-burning funny Macs after Intel:)
Wait for laptops, buy powermacs soon in my opinion
Given the disappointing powerbook upgrades, and the fairly accepted fact that the first line-up of Apple products to get Intel are going to be the laptops; you'd have to be desperate or crazy to buy an Apple laptop right now.
On the other hand, Apple's 1st gen products have typically had issues. Getting a 1st gen iBook or Powerbook utilizing Intel could be an unpleasant experience. Add in the likelihood that it'll be awhile before a lot of the Altivec code runs quickly on Intel, and even the recent powerbooks start to look appealing.
As far as the desktops though, lets assume the timelines are accurate from Apple and we see an Intel powermac sometime in late 2006, early 2007. This means a 2nd generation Powermac somewhere around 2008 and late 2008 to 2009 before the vast majority of software has Intel binaries available. That's a lot of waiting....
I have a 3.5 year old Powermac dual 1ghz G4 and am rather torn myself about upgrading to a Quad G5, or just getting a video card upgrade and dual 1.8ghz G4 upgrade. This brings me to something I'm not sure many have actually considered though.
With the switch to Intel, how easy is it going to be to slap in new processors? It's quite likely with the new Intel based powermacs that upgrading the CPU will be as easy as in any other PC (or they could cripple that with DRM). The G5's (especially the liquid cooled ones) have no such upgradability (yet).
Anyways, I think the one thing everyone has proven by now, is that its a very very tough question to answer, and the answer changes a lot depending on your particular circumstances.
I think your assumption is wrong
Your "within 4 years scenario" is correct. Even at that high rate of sales, half of the installed base will still be PPC based. But you aren't factoring in another fact about OS sales. Many users (probably the majority even) don't upgrade their operating system at all. Mac enthusiasts do but most people only upgrade when software requirements dictate or when they purchase a new machine. So while half of the Macs in 4 years will still be PPC, that doesn't meant he market for PPC version of 10.6 is going to be 50% of the market.
I do believe that Apple will sell a large number of Intel machines in the first 4 years and I think that PPC will be well on its way out by then. I think it is non-controversial that 10.5 will continue to support PPC but 10.6 is much more questionable. Apple has been pretty consistent throughout its history obsoleting old products sooner rather than later.
But, how long do you keep a machine? Even if it is as long as 5 years, you are going to be getting an Intel Mac just a few months after the release of 10.6. That doesn't sound particularly scary to me.
As always, depends on your needs...
My latest mac is a 400MHz Powerbook G3. I'm getting an iMac soon to act as my main machine and storage/backup server. Part of that is that I know Retrospect and some OS9 (classic) apps I still use will have a home for some time to come. When the Powerbooks are out (probably 2nd rev) I'll think about a new laptop.
go ahead an buy the PPC Mac
1) Current software: There is currently alot of software for OSX on PPC. Yes, you will be able to run this on an Intel Mac, but only in emulation mode, which will be slower. Some of this existing software may not be released in an Intel version for a long time (until there are enough Intel Macs out there to justify it).
2) NEW software will not be a problem for PPC: For several years, new software will be released:
a) As "fat" binaries, which will run on either PPC or Intel OR
b) In both Intel and PPC versions.
It will be years before anything is released just for the Intel Macs. I wouldn't expect to see Intel only versions until at least 2009.
In my opinion, you would be better off getting a PPC Mac and waiting on the Intel mac until at least 2008, when most of the software will be available in a native version or fat binary. In fact, if I was buying a Mac today and could choose PPC or Intel, I would buy the PPC.