They are pissed because people are finally realizing that Mhz isn't everything. And because Apple is catching up (and will pass) them in Mhz, so they have to make Apple look bad somehow. What do you bet they'll have some guy doing a speech about the Mhz myth just like our very own Jon Rubinstein.
If they gain marketshare vs. the "desktop processors in laptops" that's a good thing because it drives the better processors' costs down. Meanwhile, Joe Consumer might realize that the best laptop is not necessarily the one with most MHz for least dollars. That might get Joe to consider Apple as well.
We all knew this would happen. As soon as intel can't inch any more mhz. our of their processors without creating unmanageable heat problems, they are left with a few options. Completely redesign their pentium lines, maximizing the processors efficiency/speed/heat ratio OR back down from the mhz. jargon and start to convince people it's not that important.
one of these things would cost them millions and millions of dollars.
Posted 03/13/2004 @ 11:59 AM, by Ken "Caesar" Fisher
Could it be that the megahertz myth is really about to die? The phrase megahertz myth points, of course, to the long standing awareness that mere clock ratings do not convey comparative performance between CPU architectures. The myth was commonly cited as a detriment to competition by Apple, whose offerings have trailed Intel in overall clock speed for years. In more recent years we've heard a bit of rumbling about this from AMD, too. Indeed, AMD even went so far as to mount the Soap Box and wax philosophical about the matter. But fate would have it that Intel, the market leader, would rise (or some might say shrink) to the task. Why? We spotted it in the Fall of 2002: their mobile processor line confuses the heck out of customers.
Quote:
The chip giant is expected to begin the practice with the launch of its latest Pentium M processor, dubbed Dothan, which is due in the second quarter. Pentium 4 and Celeron chips will also get model numbers, as Intel aims to get the system in place by summer, the source said. Under the model number system, processors will be given numbers to describe their performance, in addition to being described as running at 2GHz or another speed.
Another "why" could be the fact that Intel is feeling heat like never before. While the company has struggled to climb out of its 3 GHz slump, the Athlon 64 and Opteron are gaining market recognition, and many still expect to see the PowerPC 970 hit 3 GHz by the Fall. Still, at this stage I think it's safe to say that Intel's Pentium M is near and dear to the company, and that this is the central impetus behind this move. The reasons should be obvious: the architecture in question is looking to have better potential over the long haul than Intel's NetBurst architecture in the Pentium 4. One can't discount the opportunity for clarity, either: with more than a half-dozen products in their CPU line, several of which are nearly identical to one another, this could help to company map its own product line more clearly for the consumer. One thing is for certain, however: you won't see a score along the lines of 2800+ (aka, AMD's scheme).
One very interesting aspect of this development will be watching how Intel deals with their own speed steps. If the company has to back off megahertz as a crutch for it's products, it's going to have to develop a performance rating system that "markets" their steps as being considerable. Many consumers think that there is a considerable difference between, say, a 2.4 GHz P4 and a 2.6 GHz P4 (200MHz wow!! wow!!). If Intel wishes to keep such granularity, the company may need to develop a scheme that artificially expresses a deeper gap between these two CPUs than there actually is. But then there's also the question of power consumption, too, a question that should be central to any scheme aimed at promoting the Pentium M... be it in laptops (for now), or desktops later.
A lot of Intel's de-emphasis of the MHz myth stems from this: class action lawsuit. They touted the P4 as the end all be all performer, when the reality was the P3 out performed the early P4's. People were sold (and paid mucho $$) for early P4's for what? It was a waste of money because AMD was right on Intel's heals as far as MHz (late 02, early 03 AMD's were allegedly superior in all aspects when compared to P3's and P4's except in raw number crunching) so Intel took the gamble and released a "faster" processor. People began to wake up soon there after because those AMD's and PPC's STILL outperformed the "superior" Intel processors, so obviously clock speed != processor performance.
The beautiful thing is that Intel is having to do this because Centrino notebooks aren't selling. It's nice to see them have to confront what Apple's been battling for years now: The marketing landscape is so completely divorced from reality that when Intel rolls out a superior portable architecture it flops!
I can't say I feel sorry for them. I'm glad it's Intel that will have to spend the money and effort to reeducate people on the merits of architectures like the Pentium M. It will only help Apple, and it may allow the portable space to really take off once consumers warm to an architecture that can actually be used in a credible portable.
It'll be interesting to see how Intel approaches this. If they do it the AMD way, they won't be de-emphasizing the MHz Myth. Instead they'll be promoting it.
What do you think of when you see 3200+ rating trailing an AMD Athlon 64? It's basically means "equivalent MHz." I thought we were debunking the myth here. I really hope Intel takes a much better route to this. I also think it will be interesting to see how AMD counters. Imagine if Intel all of a sudden introduced the Pentium 4 "8000." Would AMD have to come back with an Athlon 64 "9000+"?? Ridiculous.
It'll be interesting to see how Intel approaches this. If they do it the AMD way, they won't be de-emphasizing the MHz Myth. Instead they'll be promoting it.
What do you think of when you see 3200+ rating trailing an AMD Athlon 64? It's basically means "equivalent MHz." I thought we were debunking the myth here. I really hope Intel takes a much better route to this. I also think it will be interesting to see how AMD counters. Imagine if Intel all of a sudden introduced the Pentium 4 "8000." Would AMD have to come back with an Athlon 64 "9000+"?? Ridiculous.
This is why I hate AMD.
If I understood correctly, you'll not have a PR rating that implies MHz, but rather a model number scheme like the Athlon 64-FX and Opteron, for example "Opteron 246", "Athlon 64-FX 53", which says nothing about it's performance. It just fools customers to think that every step up in model number implies a significant jump up in performance.
If I understood correctly, you'll not have a PR rating that implies MHz, but rather a model number scheme like the Athlon 64-FX and Opteron, for example "Opteron 246", "Athlon 64-FX 53", which says nothing about it's performance. It just fools customers to think that every step up in model number implies a significant jump up in performance.
The Athlon 64 FX and Opterons aren't geared toward the average buyer though. And the FX line will probably fade away after the move to Socket 939.
According to Anandtech, AMD aiming for an Athlon 64 4200+ in '05, but because the numbers are arbitrary, I'm sure the release date for whatever ends up being the "4200+" is flexible...and probably heavily dependent on Intel's performance rating scheme.
Comments
If they gain marketshare vs. the "desktop processors in laptops" that's a good thing because it drives the better processors' costs down. Meanwhile, Joe Consumer might realize that the best laptop is not necessarily the one with most MHz for least dollars. That might get Joe to consider Apple as well.
one of these things would cost them millions and millions of dollars.
Intel plans to dispel the megahertz myth which served so well
Posted 03/13/2004 @ 11:59 AM, by Ken "Caesar" Fisher
Could it be that the megahertz myth is really about to die? The phrase megahertz myth points, of course, to the long standing awareness that mere clock ratings do not convey comparative performance between CPU architectures. The myth was commonly cited as a detriment to competition by Apple, whose offerings have trailed Intel in overall clock speed for years. In more recent years we've heard a bit of rumbling about this from AMD, too. Indeed, AMD even went so far as to mount the Soap Box and wax philosophical about the matter. But fate would have it that Intel, the market leader, would rise (or some might say shrink) to the task. Why? We spotted it in the Fall of 2002: their mobile processor line confuses the heck out of customers.
The chip giant is expected to begin the practice with the launch of its latest Pentium M processor, dubbed Dothan, which is due in the second quarter. Pentium 4 and Celeron chips will also get model numbers, as Intel aims to get the system in place by summer, the source said. Under the model number system, processors will be given numbers to describe their performance, in addition to being described as running at 2GHz or another speed.
Another "why" could be the fact that Intel is feeling heat like never before. While the company has struggled to climb out of its 3 GHz slump, the Athlon 64 and Opteron are gaining market recognition, and many still expect to see the PowerPC 970 hit 3 GHz by the Fall. Still, at this stage I think it's safe to say that Intel's Pentium M is near and dear to the company, and that this is the central impetus behind this move. The reasons should be obvious: the architecture in question is looking to have better potential over the long haul than Intel's NetBurst architecture in the Pentium 4. One can't discount the opportunity for clarity, either: with more than a half-dozen products in their CPU line, several of which are nearly identical to one another, this could help to company map its own product line more clearly for the consumer. One thing is for certain, however: you won't see a score along the lines of 2800+ (aka, AMD's scheme).
One very interesting aspect of this development will be watching how Intel deals with their own speed steps. If the company has to back off megahertz as a crutch for it's products, it's going to have to develop a performance rating system that "markets" their steps as being considerable. Many consumers think that there is a considerable difference between, say, a 2.4 GHz P4 and a 2.6 GHz P4 (200MHz wow!! wow!!). If Intel wishes to keep such granularity, the company may need to develop a scheme that artificially expresses a deeper gap between these two CPUs than there actually is. But then there's also the question of power consumption, too, a question that should be central to any scheme aimed at promoting the Pentium M... be it in laptops (for now), or desktops later.
I can't say I feel sorry for them. I'm glad it's Intel that will have to spend the money and effort to reeducate people on the merits of architectures like the Pentium M. It will only help Apple, and it may allow the portable space to really take off once consumers warm to an architecture that can actually be used in a credible portable.
chickens
home
roost
I'm still trying to wrap my mind around how computer companies can get a P4 into a laptop.
What do you think of when you see 3200+ rating trailing an AMD Athlon 64? It's basically means "equivalent MHz." I thought we were debunking the myth here. I really hope Intel takes a much better route to this. I also think it will be interesting to see how AMD counters. Imagine if Intel all of a sudden introduced the Pentium 4 "8000." Would AMD have to come back with an Athlon 64 "9000+"?? Ridiculous.
This is why I hate AMD.
Originally posted by Eugene
It'll be interesting to see how Intel approaches this. If they do it the AMD way, they won't be de-emphasizing the MHz Myth. Instead they'll be promoting it.
What do you think of when you see 3200+ rating trailing an AMD Athlon 64? It's basically means "equivalent MHz." I thought we were debunking the myth here. I really hope Intel takes a much better route to this. I also think it will be interesting to see how AMD counters. Imagine if Intel all of a sudden introduced the Pentium 4 "8000." Would AMD have to come back with an Athlon 64 "9000+"?? Ridiculous.
This is why I hate AMD.
If I understood correctly, you'll not have a PR rating that implies MHz, but rather a model number scheme like the Athlon 64-FX and Opteron, for example "Opteron 246", "Athlon 64-FX 53", which says nothing about it's performance. It just fools customers to think that every step up in model number implies a significant jump up in performance.
Originally posted by Zapchud
If I understood correctly, you'll not have a PR rating that implies MHz, but rather a model number scheme like the Athlon 64-FX and Opteron, for example "Opteron 246", "Athlon 64-FX 53", which says nothing about it's performance. It just fools customers to think that every step up in model number implies a significant jump up in performance.
The Athlon 64 FX and Opterons aren't geared toward the average buyer though. And the FX line will probably fade away after the move to Socket 939.
According to Anandtech, AMD aiming for an Athlon 64 4200+ in '05, but because the numbers are arbitrary, I'm sure the release date for whatever ends up being the "4200+" is flexible...and probably heavily dependent on Intel's performance rating scheme.