changes in amd pr system to come?

[draggoon01]

http://news.com.com/2100-1006_3-1021832.html?tag=fd_top

In addition, AMD is apparently looking at changes to its existing Athlon XP line of processors. It is considering a plan where it would not raise the performance rating on these chips, but instead would change their underlying architecture--especially cache--to reduce manufacturing costs.
...
In the Athlon64 line, for instance, the 3700+, 3400+ and 3100+ chips will initially come with 1MB of cache. In the fourth quarter, however, the underlying structure of the 3100+ will change: Its clock speed will substantially increase, but its cache will be reduced to 256KB, or one-quarter the original size.

that would be interesting

 

[DaveSimmons]

I'm curious to see whether AMD will use the A64 rollout as a chance to make the PR ratings more honest than the inflated "3000+" and "3200+" Bartons, but I'm not holding my breath.

 

[Lord Evermore]

Eh, just sounds like a complicated plan that ends up being exactly what they do now. Barton is a lower clock speed but a higher cache so it gets the same rating as a T-bred of higher clock with lower cache. I think it's asinine myself to use these stupid model numbers to name CPUs that are so vastly different. Who wants to have to wonder if they're getting a "good" processor or a crappy low-cache version that's clocked higher specifically because of the lower cache?

 

[ObiDon]

Originally posted by: Lord Evermore
I think it's asinine myself to use these stupid model numbers to name CPUs that are so vastly different.

No doubt. I thought the whole point of the dumb model numbers was to make it easier for Joe Schmoe to understand the differences performance-wise.

The problem is...the people who previously didn't know still don't know because it's not THAT important to them. It's still about a high MHz number to them regardless of actual performance. Sort of like how someone sees a car with a 200HP engine and a car with a 220HP engine and automatically thinks the 220HP car is faster. I know plenty of people who still think this way about both things.

This would just make things overly complicated for the people who DO care about the differences. :|

 

[AgaBoogaBoo]

Originally posted by: Lord Evermore
Eh, just sounds like a complicated plan that ends up being exactly what they do now.

Thats what I was thinking at first.

I think AMD will do something though because lots of their sales come from the enthusiast market and so they have to be very careful with it, until they can compete more with Intel on a worldwide basis in retail stores.

 

[Lord Evermore]

It'll be interesting to see what happens with the Opteron model numbers. Right now there's just one core at different speeds, but when cache and bus speed differences come into play, is a 244 going to be faster than a 252, or the other way around, or is a 264 faster than a 255? Or will they do nothing but give us incremental numbers based on their own ideas of which CPU is "faster" without any indication of what makes it faster? So we might never see a "264" but only a 266 to indicate it's faster, even though it may have a lower cache and just be higher clocked, or have a wider bus, or something like that.

Or, since 240 to 242 is a difference of 200MHz, will a 1.6GHz Opteron with half the cache also have a 2x2 number like 222 to indicate the smaller cache? Does that end up meaning it's "slower" than a 240 then even though it's a faster speed? And would a 262 at 1.6GHz with lots of cache be "faster" than a 242 which has a higher speed?

Of course, most of us won't be buying Opterons, but it's not like most of corporate America has ever heard of AMD.

As far as I'm concerned, AMD is blowing away any advantages they had, and the "root for the underdog" mentality gets kind of old when the underdog is punching itself in the face.