The pros and cons of an Apple-Intel divorce
#1
Posted 22 January 2013 - 02:00 PM
#2
Posted 22 January 2013 - 02:40 PM
Macworld, on 22 January 2013 - 02:00 PM, said:
One of Apple's platforms - the general purpose computer - is not a walled garden, so explaining the reason it is one is nonsensical.
The other Apple platform - the consumer appliance - is a walled garden not to avoid sharing profits but to increase the relability of devices that are supposed to "just work" and improve accountability when they don't.
Pity. Other than that sentence it wasn't really that bad a piece.
#3
Posted 22 January 2013 - 08:43 PM
#4
Posted 23 January 2013 - 03:22 AM
mretondo, on 22 January 2013 - 08:43 PM, said:
This is actually acknowledged in the "stupid" story. Anyone who knows a hoot about processors and programming knows that the raw performance gap between the two architectures only matters if the target audience is running compute-bound processes for any significant amount of time. For most consumers, that's really not an issue and hasn't been for a decade or more. Storage - the amount of RAM and the performance of disk - has been the limiting factor of late. Cheaper, higher-density RAM and dropping prices on SSD are starting to chip away at that.
#5
Posted 23 January 2013 - 07:46 AM
Quote
It's not the story that's stupid, it's the idea of using ARM over Intel that is. This idea has been getting tossed around, and the author is responding to it.
You're right in that ARM gives rudimentary performance at a huge energy savings (for mobile), but raw performance is another matter. I don't mind ARM too much for phones, tablets, or my fridge, but for my laptops or desktop. No way. Even if somehow ARM catches up in performance ("centicore" chip?), compatibility is not something that should be sacrificed "willy nilly", there would have to be compelling reasons.
#6
Posted 23 January 2013 - 07:52 AM
Quote
Anyone who knows a hoot about consumer behavior knows that people don't make rational purchasing decisions based on usage-weighted profiling of performance bottlenecks. They want the fastest CPU, period, end of story, and no amount of "oh, this tradeoff makes sense when you do the math, really!" will convince them otherwise.
This post has been edited by technologist: 23 January 2013 - 07:53 AM
#7
Posted 23 January 2013 - 08:23 AM
technologist, on 23 January 2013 - 07:52 AM, said:
Quote
Anyone who knows a hoot about consumer behavior knows that people don't make rational purchasing decisions based on usage-weighted profiling of performance bottlenecks. They want the fastest CPU, period, end of story, and no amount of "oh, this tradeoff makes sense when you do the math, really!" will convince them otherwise.
That's true for measurebators. Not the general consumer of this century. I've seen a small but growing number of people foregoing traditional PCs outright because their phone or tablet does everything they need from a computer. Implicitly that means they've come to the conclusion that it's "fast enough."
#8
Posted 23 January 2013 - 08:49 AM
Quote
Anyone who knows a hoot about processors and programming knows that the raw performance gap between the two architectures only matters if the target audience is running compute-bound processes for any significant amount of time. For most consumers, that's really not an issue and hasn't been for a decade or more.
Can you cite some references to support this claim?
#9
Posted 23 January 2013 - 10:30 AM
#10
Posted 23 January 2013 - 10:59 AM
Quote
Anyone who knows a hoot about processors and programming knows that the raw performance gap between the two architectures only matters if the target audience is running compute-bound processes for any significant amount of time. Anyone who knows a hoot about consumer behavior knows that people don't make rational purchasing decisions based on usage-weighted profiling of performance bottlenecks. They want the fastest CPU, period, end of story, and no amount of "oh, this tradeoff makes sense when you do the math, really!" will convince them otherwise. That's true for measurebators. Not the general consumer of this century. I've seen a small but growing number of people foregoing traditional PCs outright because their phone or tablet does everything they need from a computer. Implicitly that means they've come to the conclusion that it's "fast enough."
Indeed, but Apple already has a product line for these people.
#11
Posted 23 January 2013 - 11:09 AM
macmutt, on 23 January 2013 - 10:30 AM, said:
1. The platform you're describing is called a PC (or a hackintosh, if you will).
2. Been there, done that.
#12
Posted 23 January 2013 - 11:16 AM
technologist, on 23 January 2013 - 10:59 AM, said:
Quote
technologist said
Anyone who knows a hoot about processors and programming knows that the raw performance gap between the two architectures only matters if the target audience is running compute-bound processes for any significant amount of time. Anyone who knows a hoot about consumer behavior knows that people don't make rational purchasing decisions based on usage-weighted profiling of performance bottlenecks. They want the fastest CPU, period, end of story, and no amount of "oh, this tradeoff makes sense when you do the math, really!" will convince them otherwise. That's true for measurebators. Not the general consumer of this century. I've seen a small but growing number of people foregoing traditional PCs outright because their phone or tablet does everything they need from a computer. Implicitly that means they've come to the conclusion that it's "fast enough."
Indeed, but Apple already has a product line for these people.
Intel Macs have never, ever had the fastest processors Intel makes, yet they continue to sell well. Why? Because the number of people (i.e., professionals) who need the absolute best-performing computers is shrinking. For most computer consumers, the garden-variety Core i7s are fast enough.
This post has been edited by wardoggie: 23 January 2013 - 11:23 AM
#13
Posted 23 January 2013 - 11:54 AM
macmutt, on 23 January 2013 - 10:30 AM, said:
This is a fascinating theory. How do you reconcile these notions with the reality that the cooling profiles of future components will be very different from those contemporary with the "host" structure and that any physical interface you design is likely to be outmoded - by need, not by whimsy - in a few years?
#14
Posted 23 January 2013 - 12:01 PM
technologist, on 23 January 2013 - 10:59 AM, said:
Quote
technologist said
Anyone who knows a hoot about processors and programming knows that the raw performance gap between the two architectures only matters if the target audience is running compute-bound processes for any significant amount of time. Anyone who knows a hoot about consumer behavior knows that people don't make rational purchasing decisions based on usage-weighted profiling of performance bottlenecks. They want the fastest CPU, period, end of story, and no amount of "oh, this tradeoff makes sense when you do the math, really!" will convince them otherwise. That's true for measurebators. Not the general consumer of this century. I've seen a small but growing number of people foregoing traditional PCs outright because their phone or tablet does everything they need from a computer. Implicitly that means they've come to the conclusion that it's "fast enough."
Indeed, but Apple already has a product line for these people.
What product line is that, please?
Help











