Article: AMD's Mobile Strategy
By: rwessel (robertwessel.delete@this.yahoo.com), January 10, 2012 5:30 pm
Room: Moderated Discussions
Tom J (afraid@of.spam) on 1/10/12 wrote:
---------------------------
>Here are some examples of people who need more CPU performance: gamers, engineers,
>scientists and automated financial traders. People doing special effects for video
>and film is another example. Video analysis for security and law enforcement is
>another example. Almost any kind of simulation gets more accurate with more CPU
>performance (cloth simulation, weather prediction, climate modeling, computational
>chemistry, computational fluid dynamics, crash simulation, ...). Voice recognition
>should become more accurate with increased CPU performance.
>
>Doug Siebert on 12/31/11 wrote:
>> If Intel announced tomorrow that it had reached the limit of Moore's Law, and
>> could not make future generations of CPUs any faster than
>> the current
>> generation, I doubt it would have all that large of an effect of PC sales.
>
>Do you think corporations would replace PCs as often if there was no improvement
>in CPU performance? I don't. I know people whose companies used to replace their
>PCs every 3 years and today it's every 5 years. That's because the rate of improvement
>in performance has slowed down. If the rate of improvement slows down further,
>those purchases will get delayed even more.
True, but much beyond that and the reliability of the systems starts taking a nosedive. A fair part of that trend is being expressed by the acquisition of less-and-less expensive systems, particularly for desktops, where the display, which is often the one major cost item worth splurging on for most employees, has a life cycle basically independent of the system its attached too.
---------------------------
>Here are some examples of people who need more CPU performance: gamers, engineers,
>scientists and automated financial traders. People doing special effects for video
>and film is another example. Video analysis for security and law enforcement is
>another example. Almost any kind of simulation gets more accurate with more CPU
>performance (cloth simulation, weather prediction, climate modeling, computational
>chemistry, computational fluid dynamics, crash simulation, ...). Voice recognition
>should become more accurate with increased CPU performance.
>
>Doug Siebert on 12/31/11 wrote:
>> If Intel announced tomorrow that it had reached the limit of Moore's Law, and
>> could not make future generations of CPUs any faster than
>> the current
>> generation, I doubt it would have all that large of an effect of PC sales.
>
>Do you think corporations would replace PCs as often if there was no improvement
>in CPU performance? I don't. I know people whose companies used to replace their
>PCs every 3 years and today it's every 5 years. That's because the rate of improvement
>in performance has slowed down. If the rate of improvement slows down further,
>those purchases will get delayed even more.
True, but much beyond that and the reliability of the systems starts taking a nosedive. A fair part of that trend is being expressed by the acquisition of less-and-less expensive systems, particularly for desktops, where the display, which is often the one major cost item worth splurging on for most employees, has a life cycle basically independent of the system its attached too.