Myth or Reality #1: TIMING

In the last few weeks I have heard mention of every machine that Apple produces bar the PowerMac as a possible candidate for early Intel introduction. That includes mention of the XServe by a senior industry analyst. On the other hand, Paul Murphy thinks they'll be rushing something out the door because the chips that they need aren't ready (so they'll use Pentium M). There appears to be a consensus that Apple will introduce the first Intel machines, but little consensus on what they might be and how they might be powered? So, let's dig a bit deeper.

There's has been plenty written about Intel's Yonah in the last couple of weeks. This is Intel's first 65nM chip and the first dual core design based around the Pentium M architecture - a lower power architecture. The rumours are that this will formally be introduced by Intel this week at CES (preceeding MacWorld). There are also credible reports now of both NEC and Dell products based on this chip available imminently. So, I think Paul is wrong on this one - the chips that Apple needs are ready approximately ontime (or perhaps earlier than Apple's worst case projections). But what machines might Yonah be suitable for?

Most projections have been about new Mac minis and new iBooks. And I'd go along with this. Paul makes a comment that the cost of such chips would be way more than the current PowerPC G4 that Apple uses in today's mini. But I'm not so sure. I suspect Apple will use a mid-range Yonah in the mini. And I suspect at the quantity they buy, they will have ensured a good price from day 1. Furthermore, Paul uses older Intel pricing which was for chips on a larger die size and smaller wafer. It is worth remembering that the Yonah is a 65nM chip AND produced on a 300mm wafer. Yields will be much higher for this, allowing lower pricing. AMD is not producing 65nM chips yet, and this does give Intel at least a yield advantage.

What about iMacs, Powerbooks and XServes? Well, there will NOT be new iMacs announced at MacWorld based on Intel. Why? because any Mac using the G5 chip today is a 64-bit computer. The Yonah is a 32-bit chip. I just can't see Apple taking a step backwards in this. Apple will wait for the next design to come on stream - the Conroe/Merom family. Conroe is for desktops and Merom is for lower-power use. Like the Yonah they are dual core, but they will offer a level of performance up from the Yonah. Rumours this week were that Conroe has been brought forward. So, we might get to see a Conroe-based iMac (or Merom if lower power is key) perhaps by MacWorld Paris in September. My guess, is that Merom will also find it's way into the top-of-the-line Powerbooks too, so I'm expecting the Intel-based Powerbooks to arrive around the same time. As for XServes, I think these will be in the same time-frame as the PowerMac series updates - sometime in 2007. These will probably look to use the replacement for the Xeon chips that Intel has code-named Woodcrest. However, it might be conceivable for Apple to use Conroe or even Merom chips in the XServe. In the meantime, perhaps the XServe will get updated with newer PowerPC G5 variants to tide it over (dual-core chips for instance).

However, while sounding certain about some of this, I have to say there are holes in this argument. For instance, if performance under Intel is drastically improved as has been suggested, is this not going to mean the lower-price machines are better performing than the current higher price machines (eg Intel Mac mini beats iMac G5; Intel iBook beats Powerbook G4)? Usually when a technology company makes a step change, they hit the high end first. So, there are risks to this strategy and I don't fully understand how they will be resolved yet. We will consider some of these items in due time under the other myths/realities. But I've got a little suggestion here.

It strikes me that just about everyone is considering one model today with a direct replacement for that model. A G4 Mac mini replaced by an Intel Mac mini; a G4 iBook by an Intel iBook. Apple doesn't have to do this of course. It may want to be very careful with names, and it may not want to proliferate models too much, but one strategy that would blur the performance distinctions is not to replace like-with-like but to introduce a new range or more variation in the line-up. Do I have anything in mind? Well, yes, I do as a matter of fact. The current iBook and Powerbook ranges are very close in performance, but there are many differences between the two which help to justify the (large) price differential. That includes screen resolutions and display quality, build quality, connectivity, etc. While many Powerbook afficionados would have liked a G5 for purely processing reasons, many would have been happy with a dual-core G4 and perhaps a slimmer design with better battery life. Yonah would give the latter lot a perfect machine, while the former wait for a Merom-based machine. It so happens that many of that group are also the ones who will be longest using the PowerPC apps such as Photoshop etc. Web, Office Productivity and iLife apps (with exception of MS Office) will be the first Universal binaries. So, I think a perfectly reasonable strategy would be either to introduce a new range of laptops - lets call them XBooks that have many of the advantages of the existing Powerbook but maybe slimmer and with dual core, using the fastest Yonah chips. The heavy media user would have to stick with the current PowerPC architecture until the Merom models come along and offer 64-bit computing at which time, more apps will be available as universal binaries without the penalty of PowerPC emulation.

The other model area that Apple will have a problem with is the iMac vs Mac mini situation. A much faster Mac mini at similar price points to today will start to hurt iMac sales quite badly. So, perhaps what we'll see is some more variation in the mini range. I've covered in an earlier blog about the mini not being entirely suitable for the media centre tag its being loaded with (disk performance + expandability as two key limitations). So, perhaps if Apple is to move in this direction, there will be a higher priced/higher specced mini - and indeed possibly a bigger mini! Or it could be that the higher-specced mini will be the dual core machine, and the lower-end ones will utilise a single core.

Conclusions: I am pretty much with the camp that predicts Intel Mac minis and Intel iBooks announced in January and deliverable soon after. I think these will be Yonah based. I also would not be surprised to see Apple make changes to the range so as to produce some of the new machines at price points that overlap the more expensive PowerPC-based machines of today, perhaps via a new range or change of name. Score 1 for reality.